The Foundation of Entrepreneurship: Large Market Opportunities that have Repeated for 40,000 Years

“So in the future, ideas will be the real scarce inputs in the world — scarcer than both labor and capital — and the few who provide good ideas will reap huge reward”

“The Second Machine Age” by Brynjolfsson and McAfee.

“To do the impossible, you have to see the invisible.” Michel Foucault


I have been teaching traditional entrepreneurship and social entrepreneurship for about twelve years at several universities including each January for seven years at MIT Sloan. Three years at One Laptop per Child gave me first hand experience in social entrepreneurship and a billion dollar company I built in Indonesia some think qualifies me to teach entrepreneurship.

Awhile back I saw some research from Harvard Business School on why serial entrepreneurs were more successful than first timers. Somewhat counter intuitive, it was not experience nor perseverance alone that explained their success. Serial entrepreneurs do a lot more detailed research and analysis before they pick their opportunity. This factoid got me thinking about the bigger question of how entrepreneurs even find an opportunity. Further research revealed that this area of entrepreneurship — how to find the opportunity — is largely unexplored by either practitioners or academics. Many write that one should follow a passion or solve the biggest problem one has in [fill in the blank] as the basis to start a new venture. However, such approaches do not look consistent with the style of serious research of the most accomplished serial entrepreneurs.

Troubled by this lack of insight into opportunities, I began investigating. The most accomplished writers on entrepreneurship today — Steve Blank, Brad Feld, Eric Ries, Alexander Osterwalder — really have little to say. In fairness, they assume the prospective entrepreneur is already motivated by an opportunity or will pivot until they find traction in a market. When the leading practitioners and writers on entrepreneurship had little insight on the subject, I looked to the academics for guidance. However, I skipped the entrepreneurship writings that have left me generally unimpressed over the years and instead focused on economics, neuroscience, psychology and complexity science as the sources of insight about the fundamental nature of opportunity. What these sciences revealed is that certain types of opportunities have repeated over and over in the last 40,000 years and always created large new entrepreneurial opportunities.

What this article attempts to do is to describe ten frameworks to identify large market opportunities.


“…evolution can perform its tricks not just in the ‘substrate’ of DNA but in any system that has the right information processing and information storage characteristics.” [1] Eric Beinhocker

After twenty years living and working in Asia, I moved to Miami in 1999. While Miami claims to be “the capital of Latin America”, for sure it is the distribution hub for Central and South America and the Caribbean. A modern airport and a constantly upgraded port provide the infrastructure for much of the distribution and many businesses have capitalized on this investment. In Miami west of the airport in an area called Doral there are literally thousands of companies that provide logistics services, warehousing and distribution. According to the Miami Herald newspaper, Miami is reported to have more small businesses per capita than any city in the U.S.

These companies in Doral typically are privately owned by immigrants and have annual revenues of $1–30 million. A few have branch offices but most operate exclusively from their Miami headquarters. There is one notable exception to this small company landscape — Brightstar International, now a subsidiary of Sprint. Brightstar was founded in 1997 by Bolivian immigrant Marcelo Claure and has annual revenues of several billion dollars and worldwide operations. Starting from retail stores selling cellphones, today Brightstar is one of the largest distributors of cell phones in the world with clients such as America Movil and Vodaphone. The question to ask is why Marcello was able to grow his company so successfully while so many other entrepreneurs in Miami topped out at ten or thirty million dollars in annual revenues.

Many would argue that Marcelo executed better than the other Doral entrepreneurs, but this is only half the answer. Very large successful companies have two common characteristics:

1. Excellent execution

2. A big market opportunity

Execution determines the speed at which the company grows and probably its capital efficiency, but it is the market opportunity that determines the potential size of the company. Marcello identified a large market opportunity whereas most of his neighbors picked or saw much smaller opportunities. However, Marcello’s insight was not so much the impending popularity of cell phones but the more astute realization that cell phones would become popular in the developing countries of Latin America. Early on he teamed up with another legendary entrepreneur, Carlos Slim, who realized the likely popularity of cell phones in Mexico and established America Movil. Brightstar became the exclusive distributor of handsets to America Movil and later followed that company as it expanded across Latin America.

When we look at large market opportunities, there are three generic types of opportunities:

1. Do something better

2. Do something in a new way

3. Do something new

Do Something Better. In this approach, a large market is emerging or already exists. A new entrant to the market offers a far superior product or service that may be at a lower cost. Google used this approach to become the market leader in online search. Yahoo, Lycos and many others had proven that a market for search existed and was large. The superior offering from Google not only improved search results but also contributed to the growth of a now very large market.

Do Something in a New Way. In this approach, again a large market is emerging or already exists. A new entrant to the market offers a new technology, method of distribution or a new pricing strategy, to cite just a few examples of different business models, to become the market leader. For example, books had been sold since Gutenberg invented the printing press and for perhaps the last two or three hundred years had been sold through retail bookstores. Through this distribution channel book sales became a multi-billion dollar market. Amazon provided at least two new ways for the sale of books. First, Amazon offered books for sale online. Secondly, Amazon only took ownership of the book at the time of sale, thereby eliminating the cost of warehousing, the markdowns for unsold stock and the cost to finance working capital.

Do Something New. In this approach there is no established market, which indicates that the risk in this approach is much higher than the first two approaches. While there is no market, there is ample evidence of a large problem to be addressed. For example, bacterial infection has killed millions of people over the centuries. The discovery of penicillin by Alexander Fleming and its commercialization is the classic example of doing something new that created a large market. Other examples might be the steam engine, the airplane or the elevator.

One could argue about which of the three approaches are illustrated by which companies and probably cite evidence to, for example, show that Google is an example of “doing something in a new way” rather than “doing something better”. We should perhaps not be so concerned about how we classify a large, successful company but rather recognize that there are three different opportunity types.

While I believe this three-part framework is complete, it has perhaps only limited application in helping the aspiring entrepreneur to identify a large market opportunity. This framework describes the types of opportunities but does little to help in the identification process. In fact, in the academic literature and the popular press there is little available to guide an entrepreneur to identify a large market opportunity. For example, the popular Eric Ries’ Lean Startup methodology and Alexander Osterwalder‘s Business Model Canvas start with an assumed market opportunity. Perhaps not fair, but all of Michael Porter’s work on strategy development also assumes a known market opportunity to be analyzed. While all of these writers have made great contributions to the study and practice of entrepreneurship, I think they all “beg the question” of identifying the large market opportunity.

At the twenty-fifth anniversary celebration of the MIT Media Lab in 2010, many people spoke about the cutting edge technologies currently under development at the Lab. The smartest person at the celebration that day was arguably the legendary MIT professor, Marvin Minsky. Minsky founded the Artificial Intelligence Lab at MIT in 1963, trained many of the students who pioneered the Internet and has many other world class accomplishments besides his contributions at the Media Lab. As is the case with many notable thinkers, he is extremely knowledgeable in multiple disciplines, including mathematics, psychology and computing. What Minsky talked about that day was how to find great ideas. He said that when you hear a great idea, do not spend time trying to better understand the idea. Rather, one should ask the person how they came up with the idea. Minsky believes that the thought processes of great thinkers can be applied to new problems. Minsky’s fascination with the thinking process of world-class thinkers is actually a theme that dates back to Plato and has been written about by many “geniuses” including Descartes, Shannon and Einstein.

Using Minsky’s logic, if one wishes to identify large market opportunities, we should be able to apply certain ways of thinking to the subject and that some of these processes come from great thinkers of the past. What this article attempts to do is to describe ten frameworks derived in part from great thinkers of the past and present.

While many would cite Steve Jobs or Jeff Bezos as great business thinkers, I believe it may be easier and more useful for the general population and aspiring entrepreneurs to look at standalone frameworks that are easily articulated. One source of such frameworks is certain Nobel Prize winners in economics (sections 1–3 of the article). Before, you throw this article in the virtual wastebasket, I would remind you that my objective is to explain each framework in a practical way with several examples, such that an aspiring entrepreneur can easily apply it. The “economic” theory is explained briefly with no math and only to set the stage for the practical application.

While economics is challenging to some, the next section of the article (sections 4–6) is a presentation of certain findings from cognition that illustrate large market opportunities. The last section of the article (sections 7–10) presents large market opportunities understood through complexity science. What the sections on economics, cognition and complexity have in common is that they each describe very fundamental characteristics of human behavior. For example, economics explains the human reliance on trust, cognition explains the role of assumptions and complexity explains networks. Of course, all ten of the concepts could have been explained in different sections — they are fundamentals after all. However, these ten are fundamental building blocks of human behavior and the foundation for many, if not all, large market opportunities.

Two cautionary notes are required at this point.

1. Large market opportunities do not necessarily lead to large new entrepreneurial ventures. Examining Dropbox illustrates this point. Dropbox provides backup storage of computer content and the synching of such content across multiple devices. Dropbox could have been just a feature in a product, as evidenced by Apple’s discussions of acquiring the company to add it to iOS. Dropbox is similar in many ways to Microsoft’s [Sharepoint], which shows that Dropbox could have been just a product. It was the vision and drive of the founders of Dropbox that enabled the company to surpass feature and product options to create a company that today has a reported multi-billion dollar valuation. After one identifies a large market one still has to ascertain whether a company, rather than a feature or a product, can be built.

2. The second cautionary note is about market timing. Just because the large market opportunity exists that does not mean that there are customers willing to pay. Ignoring this point is the origin of the popular saying “build it and they will come”. This is a cautionary saying because the customers may not come. This point is illustrated by a personal investment experience. In 1982 I invested in a company that offered videotext. Videotext allowed remote access to mainframe computers and served up alphanumeric information in four colors. Videotext never became widely accepted by businesses, but it was the first technology to offer the benefits of the Internet that became so popular around 1995. My investment in videotext was premature by 10–15 years. Before investing significant capital one needs to confirm the presence of a large number of customers willing to pay for the product or service such that positive cash flow and preferably profitability can be achieved. There are documented cases of companies that pursued markets for ten years or more before they found significant product-market fit, but it is generally advisable to pursue market opportunities with much shorter time frames.

One might ask why I focus on techniques to identify only large market opportunities and do not offer guidance to small businesses. There is no profound reasoning here, no judgment on which type of company is more beneficial to society or any logic based on economic theories. The simple fact is that creating large companies is what interests me (and most venture capitalists, private equity investors and stock market analysts). My interest in large market opportunities began when I worked in Indonesia. In seven years I led a team that built a publicly traded company with annual revenues of $1 billion. Later I worked at one of the largest social entrepreneurship organizations, perhaps at the time second in annual revenues only to Mohammed Yunus’ Grameen Bank. I would hope that this article provides some practical advice so that others can find the large market opportunities and build large companies.

1— Hayek’s Habits

“The only real revolution is in the enlightenment of the mind and the improvement of character, the only real emancipation is individual, and the only real revolutionists are philosophers and saints. “ Will and Arial Durant[2]

Many have said that history repeats itself, but I know of no history of entrepreneurship. We have histories of technology and business and the later is a growing academic field. A history of entrepreneurship, to have real value and intellectual standing, would have to establish a domain separate from technology and business. So what would one write about?

To paraphrase Friedrich Hayek the Nobel Laureate economist, society is the evolution and imitation of organizations and habits. By “habits”, Hayek meant “customs,” “norms,” “practices,” “traditions” or “rules of conduct”. Perhaps the evolution of habits would provide insight into the history of entrepreneurship and help us to understand the first way to identify large market opportunities. Habits relate to individuals and groups of individuals, what some would call customers. Understanding customers, their problems and needs has always been a fundamental concept in entrepreneurship. Perhaps exploring habits is a worthwhile path to understand market opportunity.

About 40,000 years ago mankind transitioned from hunter gathers to community-based residents. What was the most fundamental change in a habit that was required to permit this change? “Trust”. Man went from only trusting his immediate family and tribe to trusting a much larger population. (One might call this behavior the beginning of networking (which is described in detail in Chapter IX.) At the same time technology permitted more abundant resources, which permitted a change in “sharing” (the other requirement for networking) and a scarcity that for the first time was manipulated by man. The expanded sharing led to benefits from division of labor, specialization and barter. The increase in specialized craftsman and barter eventually led to the creation of firms and money, which provided efficiencies that were required for the new scale of commercial activities. This scale and the related networking also led to the creation of marketplaces for the first time where a wide selection of merchandise proved an attractive draw to a larger number of purchasers than individual merchants could attract.

I do not pretend to be an anthropologist or to even present the history in the last paragraph in correct chronological order. However, I am quite confident that these fundamental habits, as Hayek would term them, are the foundation of modern commerce. To avoid confusion those habits are:

· Trust

· Sharing

· Exchange

· Markets

Each of these habits is a foundational element in society and in commerce. As each individual habit evolved and was imitated there were significant new market opportunities created. If these changes have created market opportunities over the last forty thousand years as we will now show, perhaps changes in these habits in the future will create new, significant market opportunities.


Advances in neuroscience show that we are biologically programmed to cooperate. For cooperation to provide meaningful benefits an individual must develop the ability to trust, which they do starting at about age two. As the individual matures the concept of trust finds more and more applications. Trust becomes a foundational skill to manage the more complex life that comes with age. The more complex the environment, the greater is the need for trust. Trust allows one to lower transaction costs, achieve economies of scale, delegate responsibility and share (described in the next section), all benefits to “better” manage complexity. Trust, as defined here, is also required in order to scale any venture or organization.

It is the continuing evolution of trust that permits the benefits to be derived from the increased complexity in society. These new forms of trust are the large market opportunities. While many associate advances in society with new technologies, these technologies will not be adopted without trust in the provider and/or the benefits of the product or service. Individual decision-making is guided by self-interest and trust permits one to properly quantify the utility. An example illustrates the point. EBay was one of the first online marketplaces. However, many have documented that their success came only after they added a “peer review of seller” feature to the site. This feature enabled buyers to have trust that merchandise would arrive from this new online service. Yelp, the current market leader in local business search, quickly jumped ahead of the older CitySearch by using a strategy similar to EBay. Yelp used identified individuals as the reviewers of local businesses, which created credibility and trust for users.

While EBay and Yelp are excellent examples of how trust can be applied in a new way, other examples of large market opportunities based on innovative applications of trust include:

· Money

· Banks

· Stock (equity)

· Insurance

· Delivery services

· Lay away

· Licensing such as doctors, lawyers, etc.

As may be apparent form the list above, every new financial asset class (cash, loans, equity, CMOs, derivatives, etc.) is a new form of trust in a counterparty or partner. And of course all these changes in money fostered efficiency in capital and capital mobility that stimulated growth in the economy and reinforced the importance of trust.

As trust permitted increased social and economic complexity, one of the forms of trust that emerged was central government. Presumably for providing the public good of a required shared service (the trusting) at a lower transaction cost, people gave up certain rights to government control. Military protection, which had always resided in the controlling entity, is probably the basis for central government to emerge. Obviously the notion of lower transaction costs has been lost over time, but the question one might ponder is whether there is a new trusted provider of services to replace government. Such a trusted provider would be a large market opportunity. Just for fun, if we polled the U.S. population and asked whether they would prefer [trust] Google, Johnson & Johnson or the elected officials and Washington bureaucracy to run the country, who would come out on top?


Sharing is the granting of certain rights to use physical goods for a pre-determined period of time. Sharing also applies to information, which takes many forms including prose, music and video. The sharing of objects and information frequently improves human relationships. (Market opportunities related to information are discussed in Chapter III.) The frequency of sharing is inversely proportional to the value of the good and the marginal cost to the recipient to use what is shared. To initiate sharing there must be some sufficient amount of trust in the other party by the owner of the good or information.

Sharing applied to goods and information has created some of the largest market opportunities:

· Real estate (shopping centers, hotels, office buildings)

· Books

· Newspapers

· Rental cars

· Cruise ships

· Blogging

· Social media (Twitter, Pinterest, Instagram, etc.)

· AirBnb, Uber, etc.

· Cloud computing

· Open source software

· Outsourcing

If we look at the recent trend for sharing assets, we notice an increasing number of different types of assets being shared. AirBnb and Uber would be noteworthy examples. The next step in sharing may well be time. I think the idea that different groups operate different businesses from the same asset may become popular. For example most dinner restaurants are empty at breakfast time, so a different group could use the restaurant from 6–11am to offer breakfast or 6am-3pm to offer breakfast and lunch. Maybe office space could be used the same way.


If we combine the concepts of sharing and self-interest, we derive “exchange”. Initially man probably shared as a means to build relationships. As soon as man realized that he had a scarce resource desired by others, his self-interest prompted him to be compensated for his “sharing” and “exchange” was born. These early exchanges of goods such as fruits and vegetables, weapons and tools were the first barter transactions. Barter introduced the notion of “value in exchange” as opposed to “value in use”, which may have been the first example of value being perceived. Value being perceived is an important concept for two reasons. First, the notion of perception opens the door for an exchange to have an emotional component which we show later on explains the popularity of money. Secondly, the concept of value as a perception is the underlying concept behind utility theory and explains how two people can both view the same exchange as in their self-interest.

Barter was successful as an exchange mechanism but it had shortcomings. Basically barter did not scale, which interfered with early man exploiting a specialization strategy based on his advantage from local natural resources. For example, the specialist furniture maker wants to make as much furniture as is in his self-interest. The same was true for every other craftsman and agricultural producer. However, to barter a kitchen table and chairs for ducks might result in the furniture maker accepting one hundred ducks, which results in the furniture maker becoming a large duck raiser. Struggling to be a successful specialist in furniture, unwanted diversification into ducks looks very unattractive. More problematic, a long position in ducks involves additional “transaction costs” such as feeding, housing and slaughtering the ducks. To solve this transaction cost problem we created a store of value called money to replace barter.

Money has proven successful as a means of exchange because of its low transaction costs and emotional qualities. Money is comparatively easy to accumulate with little or no marginal cost. Such accumulation creates wealth. Wealth is attractive because it speaks to several basic emotional needs, such as self-esteem, security and the welfare of future generations. Ducks and most physical “goods” typically offer no emotional fulfillment. The low transaction cost and easy to store qualities of money facilitate its emotional attractiveness and explain why money has not been replaced to date.

Money has changed many times over the last forty thousand years from its original form in high value metals such as gold and silver. In something of a chronological order, the changes include:

· Paper notes

· Bank accounts

· Wired remittances

· Credit cards

· Pay Pal

· Google Wallet

· Bitcoin

For each of these changes, a large opportunity was created. We can perhaps argue about whether Google Wallet and Bitcoin have proven themselves yet sufficiently to document a large market opportunity, but all of the others are documented large market opportunities. Every time that money changes form (metal coins, notes, credit cards, Pay Pal) or is combined with additional services (metal coins, bank accounts, remittances, Google Wallet) a large market opportunity is created. Note: if we consider all the different forms of money and why they arose, we realize that money is a technology. Money evolves in order to solve new problems, much the same as the Internet continues to evolve to solve new problems. Money is the fundamental technology to manage risk and time.[3]

I think that Bitcoin or another crypto-currency is the next exciting change in money that will create a new, large market opportunity. Bitcoin might denominate multiple asset classes. If futurists are correct that countries may no longer be viable, replaced by a return to mega-cities or city-states, then a worldwide medium of exchange to replace the U.S. Dollar, such as Bitcoin, will be required. That would be a large market opportunity derived from money and exchange.

Hayek’s concept of a habit shows us three ways to find new, large market opportunities:

1. Apply trust in a new form of transaction

2. Find a new way to share something

3. Transform stored value (money) in exchange

Markets are the last of the four Hayek habits that explain large market opportunities, but I think they are better understood in the examination of the writings of Ronald Coase in the next chapter.

Note: Sharing and exchange were possible in part because certain men advanced to the point where they had an abundance of goods that they could exchange with others. Whenever there is an abundance of a good or service, look for the scarcity to find a large market opportunity. For example, CarMax was birthed from the abundance of used cars and the perception of a scarcity of honest used car dealers. Food shows the abundance scarcity opportunity multiple times. Regular food was abundant, which lead to organic food and then vegan organic food and then gluten free, vegan organic food. One might even cite the first Internet search engines as an example of abundance scarcity — plentiful information but how do you find what you want.

2— The Firm

“Markets are an emergent property of networks of people, organizations and resources involving the exchange of information, goods and services. “ W. Bryan Arthur

Adam Smith legitimately claims the title of “father of modern economics” for his book Wealth of Nations written in 1776. This original work lead economists to consider three players in an economic system — the individual, the firm and the market. For almost one hundred sixty years no one answered the question, “why do firms exist”. In considering the allocation of resources in markets why does a “firm” offer benefits not found in other means of organization. In 1937 Ronald Coase wrote a paper, “The Nature of the Firm”, for which he was awarded the Nobel Prize some fifty-four years later in 1991.

Coase posited that the purpose of a firm was to reduce transaction costs. Transaction costs are the costs to secure for the firm resources of all forms in a market. At some point a firm realizes that by bringing an activity in house they can lower costs by converting transaction costs to administrative costs. The natural outgrowth of such a realization is economies of scale and economies of scope. In the first case one amortizes the administrative costs over a larger number of units of the same product. In the later case, economies of scope, one amortizes the administrative costs through the same production process but producing different products. This realization about scale or scope becomes one of the motivators for the owner or manager to seek a comparative advantage of lower product cost by expanding the production and revenue of the business. The desire for a comparative advantage in cost, self-interest, becomes a prime motivator for the expansion of the business and the realization of opportunity.

If we look back at the last 200 years of business history, we realize that changing views on what should be administrative costs and what remains transaction costs explain many of the large business opportunities in that period. For example, the development of railroads proved to be a cost effective alternative to in-house transportation. Later, large company-owned fleets of diesel-powered trucks showed many firms a way to replace railroad transportation at lower cost. Eventually another view prevailed with local transportation of goods handled by in-house trucks and long distance transportation outsourced to long haul trucking companies and airlines. Company operated drones may replace trucks for local transportation of goods and eventually we may have drone leasing and third party drone operator companies.

The history of the computer industry also shows the many opportunities developed as a result of different views on what should be a transaction cost and what should be an administrative cost. A few examples, not necessarily in chronological order, are shown below.

· Mainframe computing (administrative)

· Electronic data processing services (transactional)

· Personal computers (administrative)

· SAAS (transactional)

· Cloud computing (transactional)

· Resident mobile apps (administrative)

As shown above, a large market opportunity was created every time that there was an opportunity to opt for a transactional cost alternative in computing. Every time that the market adopted a new technology, the administrative cost model was used. As Clayton Christensen has described it, large corporations easily justify adopting new technologies when cost savings are evident. When cost savings are not obvious to large corporations, the technology becomes disruptive and the adoption is driven by the transaction cost approach.

In earlier times opportunity was realized through in-house specialization that lead to cost reduction, competitive advantage and greater production scale, as Coase posited administrative cost. However, in more current times the trend has been to convert administrative costs into transactional costs. This generic opportunity to convert administrative costs to transactional costs consistently represents a large market opportunity because the need and the market are already proven and documented by the administrative cost solution. Providing a lower cost alternative by shifting to a transaction cost to satisfy a current need is almost always compelling to the customer provided quality remains comparable. The worldwide phenomenon of outsourcing, whether it is logistics, manufacturing, sourcing, staffing or shipping, to name a few examples, all document the change from administrative to transactional cost. The low cost, ease with which information is exchanged over the Internet and the low cost of computing have facilitated this transition.

This shift between administrative and transaction costs does not only explain opportunity in commercial markets. Providing a transaction cost approach can be attractive for government services, with education one of the more obvious examples. Charter schools are nothing more than a switch from administrative costs to a transaction cost. As described earlier, outsourcing is always a switch from administrative to transactional costs. Other examples of the switch to transaction costs in education are online schools for home schoolers through publicly traded K12, Inc. or the increasingly popular MOOCs (massive online open courses) such as Coursera and Kahn Academy. With the exception of the executive branch, defense and the judiciary, I see little reason why the remainder of government could not be outsourced to a transaction cost model, which would return the range of government services to that anticipated by the founders of the United States.

The economist Michael Munger writing in “Forever Contemporary: The Economics of Ronald Coase” makes an interesting point about future opportunities related to transaction costs.

“What if an entrepreneur could sell reductions in the transactions costs of renting, using a combination of delivery services and software platform, such as Uber? The third entrepreneurial revolution will be based on innovations that reduce transactions costs, rather than reducing the costs of the products themselves. An unimaginable number and variety of transactions will be made possible by software innovations that solve three problems: (a) information, (b) transaction-clearing, and © trust. The result will be that the quality and durability of the items being used (in effect, rented) will increase, but the quantity of items actually in circulation will plummet.”

This Munger view of the future sets the stage for the popularity of the fourth of what might be called Hayek’s habits — the marketplace — that emerged early on in human history. Marketplaces are easily understood if we look at the model of transaction and administrative costs.

The merchant makes use of a marketplace, whether a bazaar, a shopping center, eBay or, because the cost of the selling location and sales promotion are typically lower than if the merchant operated in a stand alone way and brought certain costs in-house. The other big attraction of a marketplace is that the scope of the merchandise assortment from all the merchants is an attractive feature to draw in customers but the merchant bears the cost only for their own inventory. As Amazon and eBay continue to prosper, more and more companies have started to offer online marketplaces. Some marketplaces offer a single brand of merchandise, now viable because of the pure scale of online shoppers. Others follow the more traditional multiple brand, multiple product approach of a traditional department or discount store. Regardless of the approach, the proliferation of marketplaces suggests that distributors, middlemen and other service providers to retail will be reduced or eliminated as companies look to lower overall cost by switching to administrative costs. Amazon’s drones would be an example of a switch to administrative costs from the transaction costs of FedEx or UPS. In-house management of social media replaces the more costly transaction cost of an ad agency. In-house coding replaces the former outsourced Indians (time to market has an opportunity cost).

I see two areas that are prime to provide new opportunities surrounding markets or marketplaces:

1. Marketplaces where transaction costs are still high

2. Marketplaces that lack transparency

One marketplace where transaction costs are still high and perhaps increasing are bank borrowings. One might generalize to say that any market controlled by government regulation has high transaction costs, but I will restrict myself to discussing commercial banking. While the cost of borrowed money has declined recently compared to historic levels, the due diligence, documentation and analysis of new borrower requests has increased to unseen levels in response to regulator scrutiny. Preparation cost and management time create high transaction costs for prospective borrowers. The recent entry of hedge funds and insurance companies into traditional commercial lending is competitive and compelling because of the simpler transaction processing models of the new entrants. I think the increasing popularity of Bitcoin and similar crypto-currencies is another marketplace example that is explained in part by a desire to manage financial affairs with less government regulation and a lower transaction cost.

A large marketplace that is characterized by a lack of transparency is government services. Whether for historical reasons related to distrust in government, a lack of IT infrastructure that is citizen facing or perhaps the recognition that the citizenry does not demand better information, government services are not transparent even though many are provided through marketplaces such as those operated by the GSA (General Services Administration). Opportunity exists to foster transparency by the government. One area is in disaster monitoring and reporting. Every level of government has a responsibility to update their citizenry about current conditions, whether it is a tornado, hurricane, flood or earthquake. Yet few government agencies at any level have any ability to provide real time information. Given the popularity of Twitter and other social media, the government needs to provide real time information to satisfy current standards for transparency. Given the difficulties of bringing up the Obamacare website(s), a private sector solution may be called for, for every government in the U.S. This solution could easily be extended beyond just information to public and private sector service provider availability and costs for such services as contractors, tree trimmers, sources of building materials, etc., wherein lies the marketplace.

As we consider the likely large future opportunities that may arise from the switch to administrative or transaction costs by firms, we may actually face a third type of opportunity of enormous potential related to costs. We all remember in microeconomics the Law of Supply and Demand. (I will refrain from asking if you remember anything else about microeconomics.) Amongst other things, the law says that as supply becomes abundant the price of a product approaches zero. Therefore, the marginal cost of an additional production unit would be zero. As soon as a marginal cost of zero is achieved one would expect that the switching from transactional to administrative and vice versa would stop, assuming quality of solution being a constant. Such a trend in marginal cost is appearing in the cost of Internet connectivity (e.g. Google’s Kansas City Project), the cost of computing (e.g. Amazon AWS cloud-based services) and legal documents (e.g. Y Combinator’s document archive). The large opportunity is to control a cost factor by achieving near zero marginal costs and control the product/service as a transactional cost. (It is assumed that any firm that had achieved zero marginal cost as an administrative cost would find a way to commercialize that product or service, thereby making it a transaction cost or resource available in the market.) An example of such an approach is Uber, the ride share service, where the cost to add an additional car is nearly zero. Regulated taxi companies have to go through a timely and costly process to secure an additional license to put another car in service.

Google’s citywide free Wi-Fi program in Kansas City is a classic example of this logic to commercialize where a marginal cost is near zero. Broadband connectivity in the U.S. is a highly regulated industry that originally served to protect the economic interests of the telephone companies and more recently the cellular providers and cable television companies. These protections have historically restricted local governments and other alternative providers from offering Internet connectivity. However, in Kansas City Google has succeeded in being licensed citywide to provide Wi-Fi services for free. The additional cost for Google to provide such connectivity is minimal, if any, given its existing connectivity infrastructure. With little or no marginal cost, search related revenues from first time Internet users will likely exceed the marginal costs of Google fiber in Kansas City.

Another interesting case that relates marginal cost to opportunity is Dropbox, a leading provider of synched online storage. When Dropbox launched, synching files across multiple machines was quite novel and Dropbox swept up the market with its easy to install, easy to use, freemium pricing for synching and storage. As I write this article Dropbox is in a price war with Google, Amazon and probably Microsoft and Box. Everyone has synching for multiple devices and ease of use is common in all the providers. However, Google, Amazon and Microsoft have huge storage requirements related to other businesses and can shift the cost accounting from one business to another or achieve marginal costs for additional storage that Dropbox cannot match. Alternatively, with such low marginal cost, any of these giants could give away the synching and storage even to corporate users if the user agreed to use another more profitable services. Dropbox has no alternative services where it could earn a meaningful profit. Effectively, Google, Amazon and Microsoft have taken their marginal cost that approaches zero and commercialized it, albeit by deriving revenue from another business.

3— Asymmetry of Information

“Real innovation in technology involves a leap ahead, anticipating needs that no one really knew they had.”[4] David Yoffie

To set the stage for this chapter we need to go back 200,000 years and not the customary 40,000 to a time before technology emerged in any meaningful way. Evolutionary biologist Mark Pagel in “Wired for Culture: Origins of the Human Social Mind” describes the situation for early humans:

“So, beginning about 200,000 years ago, our fledgling species, newly equipped with the capacity for social learning had to confront two options for managing the conflicts of interest social learning would bring. One is that these new human societies could have fragmented into small family groups so that the benefits of any knowledge would flow only to one’s relatives. Had we adopted this solution we might still be living like the Neanderthals, and the world might not be so different from the way it was 40,000 years ago, when our species first entered Europe. This is because these smaller family groups would have produced fewer ideas to copy and they would have been more vulnerable to chance and bad luck. The other option was for our species to acquire a system of cooperation that could make our knowledge available to other members of our tribe or society even though they might be people we are not closely related to — in short, to work out the rules that made it possible for us to share goods and ideas cooperatively. Taking this option would mean that a vastly greater fund of accumulated wisdom and talent would become available than any one individual or even family could ever hope to produce.“

Sharing emerged and it included not just goods but also information. The ability to create and share information beyond mere genetic transfer became the salient feature that defined humans.

In 1945 Hayek, trying to point out the failure of central planning [Communism], writes one of his most famous papers “The Use of Knowledge in Society” (American Economic Review). In the paper Hayek states:

“The knowledge of the circumstances of which we must make use never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess. The economic problem of society is thus not merely a problem of how to allocate “given” resources…it is a problem of the utilization of knowledge which is not given to anyone in its totality.”

As was often the case with Hayek, one of his thoughts became the basis for a whole new school of economics — Information Economics. This quote also spawned the notion of an asymmetry of information, an important concept in economics and, as I hope to show here, a fruitful source of new business opportunities.

The concept of asymmetry of information is based on the notion, as Hayek points out, that not everyone has all the information.[5] This can be explained by three factors:

1. Not everyone has the same search skills

2. Not everyone can bear the cost to complete the necessary search

3. Not everyone defines the optimal search the same way

Therefore, asymmetry of information can be attributed to factors of quantity, quality or both.

Any unequal distribution of information gives rise to pricing premiums and above normal returns related to the transparency of information or the lack thereof. The transparency related premiums create the opportunity for arbitrage, which drives the market for resource allocation. In most markets sellers have better information, although stock markets may provide evidence to contradict that view. The legendary Charlie Munger of Berkshire Hathaway fame relates a lesson from the equally legendary investor and professor Benjamin Graham about stock market asymmetry of information. Munger states:

“Graham didn’t want to ever talk to management. And his reason was that, like the best sort of professor aiming his teaching at a mass audience, he was trying to invent a system that anybody could use. And he didn’t feel that the man in the street could run around and talk to managements and learn things. He also had a concept that the management would often couch the information very shrewdly to mislead. Therefore, it was very difficult. And that is still true, of course human nature being what it is.” “Warren [Buffett] trained under Ben Graham, who said, ‘Just look at the facts. You might lose an occasional valuable insight, but you won’t get misled.’”[6]

Asymmetry of information explains two important social and economic concepts:

1. Poverty

2. Entrepreneurship

Michael Spence is the former Dean of the Graduate School of Business at Stanford University and a Nobel Laureate in Economics for his work in Information Economics. In his 2011 book “The Next Convergence: The Future of Economic Growth in a Multispeed World” Spence makes the interesting observation that poverty is caused by an asymmetry of information. Without adequate access to information (the asymmetry), the poor are taken advantage of in terms of the prices they pay for goods and services. This overpaying helps to insure that they stay poor. These consequences are characterized as a “negative” asymmetry of information. (I gave a TEDx talk in 2017 on this topic — .)

A positive asymmetry of information is how Israel Kirzner, a retired professor of economics at NYU and perhaps the most highly regarded writer on entrepreneurship, describes entrepreneurship. In his 1973 classic “Competition and Entrepreneurship”, Kirzner stated that the entrepreneur has a “unique insight” or positive asymmetry upon which a new business is organized. Some say that Kirzner thinks of all entrepreneurs as arbitrageurs, clearing markets by reducing or eliminating premiums for the lack of transparency. This arbitrage is the “opportunity” or unique insight, which when combined with execution, provides the means to entrepreneurship.

Much of this discussion about asymmetry and entrepreneurship brings us back to markets and exchange, which we discussed in Chapter II. Markets first and foremost serve as a means for information exchange, which we know is always based on a certain lack of transparency or asymmetry on the part of one or both of the parties to a transaction. When we examine markets through the lense of asymmetric information we realize there are two strategies:

1. Exploit the asymmetry

2. Reduce or eliminate the asymmetry

These two strategies demonstrate the new business opportunities in asymmetric information.

Exploit the Asymmetry

Exploiting the asymmetry of information basically means that one garners an above normal return from their better knowledge or insight. When these above normal returns involve immoral or illegal behavior economists call the problem “moral hazard”. When we look at the positive opportunities, we see very large, well-established industries such as healthcare and hospitals. Hospitals serve as a sort of marketplace where doctors, specialized facilities and other medical services are coordinated for a naïve, inexperienced buyer — otherwise known as the patient. The patients routinely have no knowledge of the costs involved in hospital care and almost no way to find out, even today. Of note, the medical insurance companies negotiate informed contracts for services and reduce the hospital premium related to the lack of transparency (an example of 2 above). Other industries where asymmetric information is particularly noteworthy are used cars, health insurance and jewelry and precious stones. All of these industries are characterized by a high value but infrequent transactions where the search for information has a high cost or is difficult.

Information by its nature is usually time sensitive in terms of value. Old information has much less or no value compared with new or current information. One way to understand the constant improvement in communications is to look at it from the perspective of asymmetry of information. Almost all advances in communications, whether we look at telegraph, telephone or Internet (or the Rothschild carrier pigeons), were originally adopted by business in order to achieve a speed advantage in the acquisition of information. Such advantage allowed businesses to exploit the asymmetry of information for commercial purposes, until the particular technology became widespread.

Reduce or eliminate the asymmetry

The alternative new business strategy when faced with an asymmetry is to eliminate the asymmetry or in other words to create transparency. One creates transparency by providing the required, correct or accurate information in the format required by the user in a timely manner. The proper combination of all three requirements leads to large new business opportunities as demonstrated by examples in finance such as stock markets, Bloomberg Business, arbitrage and derivatives. For non-financial examples we need only consider CARFAX and Google. We could have also included ten or twenty or thirty searchable travel companies such as Expedia. We could perhaps have included EBay or Amazon, who brought transparency to used and new consumer goods purchases. The list of examples does just go on and on and is much larger in terms of both examples and dollars than the list of companies exploiting an asymmetry. This just goes to show that “it is better to give [information] than receive”.

Google reminds us how human history has been highlighted by technologies that eliminate asymmetry and bring transparency to ordinary people. Maybe it is not correct to start with the printing press, but we can include newspapers, radio and television [news], film [documentaries] and Twitter to name some important examples of such technology. I do not pretend that these technologies flourished only because they provided transparency by eliminating the asymmetry of information. The technologies satisfied many human needs, but creating transparency may have been the most noble for those of us who believe in democracy, human rights and public education. For each of these three to flourish transparency is required.

In the future I see many new opportunities based on the asymmetry of information. One might think that in this age of ubiquitous information we would not find opportunity exploiting an asymmetry. In fact, this is a key part of the model for many of the social media sites such as Facebook. Have you ever tried to export old posts, photos or videos from Facebook to another site or repository? Facebook has created an asymmetry wherein one can only effectively see certain of your personal information through Facebook. You never see Facebook results in Google searches. If information is only available one way [Facebook], that is an asymmetry and Facebook protects/exploits this asymmetry. I by no means wish to criticize Facebook but merely to cite them as a modern example of a company that exploits an asymmetry of information.

The next three examples for future opportunities and new business are based on an increase in transparency. As Edward Snowden made abundantly clear, governments collect a lot of information. As we have always known, much of the information is useful and not confidential or secret. However, with the exception of the IRS and a few other agencies that deal directly with the public, the U.S. government has limited experience sharing information with the public. The fiasco of the Obama healthcare website might be some form of proof of this point. I believe that there is a big opportunity to facilitate the sharing of government information and in the best case the data will just be publicly available for anyone who wants to hack a database or provide a service using the data. Much of the opportunity may simply arise by presenting the information “in the format required by the user”. This example also makes clear the large opportunity available to those who advance information presentation beyond SAP, Excel and infographics formats.

A second opportunity by creating transparency is what I call “bridging marketplaces”. Suppose you run a large unnamed consumer goods company based in Minneapolis. You need artwork, logos and graphic design for a new dog food. One choice would be to call your advertising agency in Chicago or New York and request their help. Another choice for the braver soul would be to look for the best free lance talent in the U.S., which should be cheaper for many reasons. Probably not in a million years do you consider hiring a design firm in Warsaw or Johannesburg or Karachi that just finished their third dog food project in the last 18 months. Why do you not seek such a firm out? First, the time and expense of the search is perceived as large. Second, how do we confirm the information the firm provides is true and correct? Third, how do we manage this remote firm? Points 1 and 2 demonstrate the asymmetry of information that arises for professional services from third world providers to first world consumers of such services. The example is the same whether we talk about architects, engineers or graphic designers. The example is the same if we talk about third world providers of handicrafts, textiles and clothing, engineering services or interior designers. Developed world consumers do not know about the talent, creativity and quality of product available at attractive prices in the third world due to a lack of transparency that reduces demand. By bridging the producers in the third world with the consumers in the developed world (the “bridging marketplace”), we create marketplaces that create transparency.

In a 2013 article in Business Insider a Cisco forecast for the year 2017 is cited. Cisco forecasts that:

“there will be about 2.8 billion machines on the Internet, representing 30% of the devices connected to the internet worldwide, up from 960 million devices and 17% in 2012”

This enormous effort to capture and collect information is what we refer to as the Internet of Things. Today there are not good estimates of the amount of data that will be stored. Part of the reason for the lack of an estimate is that we may currently lack the technology to cheaply transfer, store and filter the data. Improving each of these functions would be an example of a large opportunity related to an asymmetry of information created by the sheer scale of the available information from the Internet of Things. This negative asymmetry is our inability to easily draw insight from the data because we lack the tools to work at the required scale.


I had the opportunity to confirm the thinking of both Michael Spence and Israel Kirzner. In 2009 I joined One Laptop per Child (OLPC), a worldwide project started at MIT to provide free laptop computers to the world’s poorest children. One collection of stories from the project piqued my interest. Many of the poor families of children that had the OLPC laptops were starting new businesses or doing business differently. Initially this entrepreneurship professor could not explain the startup phenomenon. What did computers have to do with entrepreneurship? Answer — nothing, but laptops connected to the Internet provided many of the parents with their first low cost (free) access to information where the search cost was insignificant (a very low opportunity cost on their evening time). From poverty to entrepreneurship through access to information is exactly what both Spence and Kirzner would have predicted.

4— Functional Fixedness

“And therein lies the way forward. The future does not belong to an ultimate form of intelligence, but the ultimate mix of skills.” [7] Manoj Saxena

If I give you a cup, a Dixie cup, filled with water, most people have no problem to use the cup to take a drink. If I ask you to re-purpose the cup, to develop a new use case, many people will empty the water from the cup and use it to store screws or nails. A few people will empty the cup of water, fill it with soil, plant a seed and create a planter. If I ask again to re-purpose the cup and create a new use case, a few people will be able to imagine a game trapping a cricket in the cup and a smaller number still would use the cup to shape circular cookies from dough.

The difficulty in re-purposing the Dixie cup is called a cognitive bias, and the previous example illustrates what is called “functional fixedness”. The Gestalt School of psychology coined the term in the early 20th century. Humans come hard-wired to accept facts that are easy to process such as previous use cases of high utility. Daniel Kahneman in his book, Thinking, Fast and Slow, calls a cognitive bias such as functional fixedness a System 1 decision, characterized by intuitive, effortless quickness. The individual in the example above who “sees” the planter, the cricket game or the cookie cutter is using what Kahneman calls System 2 thinking. Here the individual is able to re-arrange information in a new context despite the cognitive bias to take the easy, more energy efficient approach.

Now in the example above I have illustrated functional fixedness in the simplest way, repurposing the physical object while maintaining it in its entirety. However, there are two other types of functional fixedness. One case is where you change the physical characteristics but not the purpose of the device. For example, for much of history cups where ceramic. Plastic cups that were unbreakable and lighter might be an example of the second case of functional fixedness — “change the assumption” — that is explained in the next chapter. The third case of functional fixedness is where one rearranges the components of a device. Adding a camera to a cellular phone would be an example. Decoupling is a variation on the combinatorial case where one separates components rather than add them. This combinatorial thinking is explained in Chapter V.

Hopefully, the examples of the three cases of functional fixedness have awakened your mind to the business opportunities merely by overcoming a cognitive bias. History illustrates the scale of the opportunity available from overcoming functional fixedness. A simple example illustrates. Since soap was invented I think it has been a solid “bar” of soap. However, today supermarket shelves are filled with liquid soap for dishes, hair and hands. Liquid soap still cleans but it is easier to handle and it does not leave a yucky film in the dish in the kitchen or bathroom. Another simple example demonstrates functional fixedness as an opportunity. Car seats for children are of sturdy construction and weigh about forty pounds. Not a problem in terms of daily activities when the seat is fixed to the car back seat, but what about when you travel by air. Now the seat is unwieldy, heavy and generally a huge pain to travel with. Enter the inflatable baby seat, which the UK safety board found safer in certain models than traditional car seats.

The best example, in my opinion, in the history of mankind for overcoming functional fixedness is the individual (or more likely a group) who realized that a dugout canoe (ship) could be constructed from pieces of wood rather than burning out an indentation. This re-purposing led to new boat designs that could safely serve international rather than just local commerce. These newly designed boats also served to support territorial conquest and the maintenance of such “colonies”. If the early Greeks had been able to overcome their cognitive bias, they might have placed their early stem engines[i] in boats, advanced the adoption of the technology and changed history even more[8]. Looking back we realize that sailboats, steamboats, diesel-powered boats, submarines, aircraft carriers and cruise ships were all examples of overcoming functional fixedness and creating large opportunities. All of this happened because someone realized that a boat made from pieces of wood could replace a dugout canoe.

Another great example of overcoming functional fixedness comes from ancient history and probably pre-dates the human species. This example is the ordinary rock or stone. Keep in mind the boat example as you consider re-purposing a stone. If we chip away or shape the stone, we have a spear tip or a knife. Overcoming cognitive bias again, we attach the spear tip to a stick to more safely hunt animals or men. From a chipped stone we bring technology to create weaponry and launch “modern” warfare, regrettably a very large market opportunity.

Another way to think about the cognitive bias of functional fixedness is to separate solutions from problems. The steam engine was originally developed to pump water out of mines (if we ignore the much earlier Greek example referenced above). Later on its application increased dramatically when it was used to power locomotives and ships. The introduction and widespread adoption of the steam engine is widely regarded as the start of the 2nd Industrial Revolution.

The radio was originally used in ship-to-shore communications and the Internet was designed to share research amongst academics.[9] When the Internet technology was re-purposed to support the individual’s appreciation for information, a huge opportunity was created. Throughout this article one will notice that huge market opportunities are created whenever the supply of information to the individual (consumer) is expanded. The fundamental role of information in human behavior is explained in Chapter VIII.

Another of my favorite examples of overcoming functional fixedness is Phonebloks, which is now supported by Google. Phonebloks is a cellular phone where the design anticipates that future components will be upgraded. Rather than buy a new phone, merely purchase a new processor or camera or Wi-Fi antennae, take out the old piece and snap in the new one. In a perfect world one could sell the old part to someone else. Why did we ever think of phones as static configurations?

Magne-Hinge designed by Nendo and EcoSwitch designed by Fuse Project both illustrate excellent examples of overcoming the cognitive bias. Magne-Hinge lets you snap your glasses apart so you can mix-and-match different color and shape temples and temple tips with the lenses. EcoSwitch from GE has two components: the base station, or hub, that uses induction heating and different vessels that function as kettle, blender, slow cooker and coffee maker. One heating source and multiple pots wherein the bias of one heating source per pot is overcome.

Another fantastic illustration of overcoming functional fixedness comes from urbanization and related innovation. Every major city (perhaps with exceptions in China) is physically challenged by limited space, expanding population and resource requirements. What is consistently one of the largest land areas unencumbered in any city? Parks. If we put the parks underground, we free up a significant land area without necessarily reducing quality of life. Such a project, called Lowline, is being developed in New York City. I think that innovation in urbanization will increasingly illustrate how a cognitive bias can be overcome.

All of these examples have hopefully helped you to understand the concept of the cognitive bias of functional fixedness. However, the really large opportunities from overcoming functional fixedness take us back to previous chapters where we discussed money and information. If we think of money as a “store of value” rather than a coin or a bill, we are poised to escape the bias of functional fixedness and innovate with bank accounts, checking accounts, credit cards, PayPal, loans and Bitcoin to name a few examples. In each example, we maintain the value but change the “form”, somewhat akin to switching from dugout canoes to constructed boats. Most of these examples themselves also spawn their own successors by further overcoming functional fixedness. For example, credit cards led to debit, prepaid and gift cards.

While changing the form of money always seems to create a large market opportunity, information may be an even bigger class of opportunities. From the time man first started writing about 3200 BC in Mesopotamia, information has been quick to change its form, casting aside the existing use case and overcoming functional fixedness with each new technology. Starting with the printing press, then the widespread acceptance of newspapers, followed by the Computer Age, now the Internet era and soon to be the Internet of Things (IOT), information is transformed by every major technological change of consequence. In fact, it may be a determinant of a major technological change that it involves a change in the concept of information. Lesser technological changes have also affected the form of information, where I would cite telegraph, telephone, radio and television. As was the case with money, every new information technology spawns multiple opportunities to overcome functional fixedness. If we look at broadcast television, we find cable television, Netflix, Chromecast and Apple TV. Speaking of television reminds me of another great example of overcoming functional fixedness. In the 1980s Nicholas Negroponte, Founder of the MIT Media Lab, developed the Negroponte Switch — the idea that everything with a cable should lose the cable and everything without a cable should have one. The cordless cell phone is an example of the first case and replacing broadcast television with cable is an example of the second case. In both cases the cognitive bias was overcome.

If we look to the future yet more examples of overcoming functional fixedness will undoubtedly emerge. At the simplest level perhaps we will have more products made of replaceable parts and components like Phonebloks. An environmentally motivated desire to reduce electronic waste from computers, phones and tablets might motivate such a move. Another easy to understand example might be the production of meat without animals. Our knowledge of biology and chemistry is advancing so fast that my daughter might see this change in her lifetime. Another interesting possibility is perhaps companies without humans. Artificial intelligence operated agents select a combination of programs and web-based resources and start trading the Dollar-Yen. Instead of investing in people one could invest in a new program for an FX trading company. Not so different from investing in a hedge fund, except no people. Might behave more morally or socially responsible than a company with humans…and maybe not.

If we look at the two big opportunities — money and information — -surely we will see further innovation. For sure physical money will disappear, replaced by a phone, a watch or glasses as the method to affect a payment. Perhaps even banks could be replaced if the Blockchain infrastructure that supports Bitcoin becomes sufficiently robust, trusted and used. Of course, if the money supply were no longer in banks controlled by government regulators, management of the economy would probably no longer be under government control. After such a dramatic change in the role of government, we might consider decoupling other government services which would again be a form of overcoming cognitive bias.

How information will evolve as we continue to overcome functional fixedness is a challenging question. For sure virtual reality (VR) will be a very popular new form of information. It will have far reaching implications for music, games, education/training and pornography to mention a few cases. Surely the voluminous data capture from sensors, with artificial intelligence applied to the data analysis, will lead to a more deterministic world. By this I mean that perhaps the role of information will be to dictate behavior rather than share knowledge. Within five minutes of searching Google for men’s shoes all of my social media sites start showing me ads for shoes. Eventually the algorithms will be able to tell if I am “window shopping” or seriously interested to buy shoes, at which time it starts to look like the new information provided to me is dictating my behavior rather than merely providing information about what shoes are available.

Perhaps a more interesting way to look at information and functional fixedness is to realize that for the first time in history knowledge is really no longer what one has memorized. Information now doubles every 13 months and that rate will accelerate dramatically with the onset of the Internet of Things (IOT). What this means is that organizing and curating information will become much more value-added. The change that we will see if we overcome the cognitive bias is that rather than think of information as unique individual data, we will think of information as “collections” or datasets or archives. A project by Erez Aiden and Jean-Baptiste Michel at Harvard illustrates the point. The researchers are analyzing all the books digitized by Google, over 30,000 volumes (representing 25 percent of all known books), to see if changes in word usage demonstrate shifts in economic, political and social values and beliefs.

Some are already cautioning that we run the risk of having correlation replace causality in such approaches to information analysis, which perhaps only proves that such a change in thinking about information is imminent (if critics have already surfaced). To encourage the critics, I relate the case of how IBM came up with an effective approach to translation. Their machine learning approach is the basis for the very successful Google Translate. Basically using a machine learning approach, the computer learned that certain words matched up in a sentence-by-sentence analysis in Canadian Parliament writings that had to be in both French and English. The computer was not able to give any consideration to linguistics or context in its determinations. Here again, the information was the collection of Canadian Parliament writings in two languages. As in the Harvard case earlier, the collection took on value because of the user rather than any inherent value such as is the custom with a traditional book.

S.I. Hayakawa was correct when he said, “the greatest of human achievements … the pooling of our experiences in great cooperative stores of knowledge, available to all …”. However, it may turn out that the achievement is not the stores or the creation of stores of information or knowledge. Historically much information has been proprietary or at least not shared. Corporate information would be an example. However, as the Harvard and IBM cases show, the value is in the heuristics and algorithms. Therefore, one could argue that making available the stores to the public is the first step in the value chain where algorithms and other analysis will determine the real value. This logic might lead one to share information stores rather than guard them as they are today. The first group that should consider overcoming the bias that information stores are private is the U.S. government and President Obama initiated legislation to make public more government information at Building algorithms, heuristics and other analytical processes to analyze the government stores of information would be a big opportunity.

Another consideration in this environment of ubiquitous information is that knowledge management will shift from a top-down to a bottom-up activity.[10] Simply put, rather than being given information by an authority (teacher, boss, government), the individual will naturally take responsibility to secure the information that they need or that interests them. Information services will change from providing information to organizing and curating it. The good news is that such a user-centric model is exactly the way that Jean Piaget’s seminal work said children learn. Maybe finally we will see that computers foster better learning in children. The principle here is that technology improves the user experience when it changes the user’s behavior. In this case the computer would enable the user to pull the information they need rather than having it pushed at them from an authority. Thinking about information in this way is perhaps the most profound way to overcome the cognitive bias.

5— Combinatorial Computation

“I have gathered a posy of other men’s flowers, and nothing but the thread that binds them is mine own,”


About 40,000 years ago through evolution humans developed the ability to use abstraction and representational art. This change in ability enabled man to move from a world of physical objects to symbolism and the intellectual development of ideas. Language, mathematics, music and art also were made possible from this change in the functionality of the human brain. To “create” these ideas required pattern recognition and a combinatorial faculty of the mind. At this point in history, not by coincidence, technology became a significant factor in the social and economic development of mankind. As Schumpeter stated, “new combinations of productive means” [technology] disrupt the economic equilibrium”. Man now had the ability to combine ideas, invent technology and live a richer life filled with the arts.

Since the time of Seneca, great thinkers have studied creativity and characterized it as a combinatorial process. For example, Einstein said, “Combinatory play seems to be the essential feature in productive thought.” In his new book, How Google Works, Eric Schmidt states “we are entering … a new period of combinatorial innovation.” This happens, he says, when “there is a great availability of different component parts that can be combined or recombined to create new inventions.” A new paper in the Journal of the Royal Society Interface by Santa Fe Institute (SFI) researchers Hyejin Youn, Luis Bettencourt, Jose Lobo, and Deborah Strumsky[11] shows that most inventions are combinations of existing ideas and have been for quite awhile. As many have characterized it, including Alfred Koestler (the seminal thinker on creativity), creativity is a “slot machine” where pattern recognition enables one to see value or “a means to human purpose”.[12]

Now if we think about the Schumpeter quote above, we can show how this combinatorial facility explains large market opportunities. W. Brian Arthur, one of the founders of the field of complexity economics, describes the invention of novel technologies as a process of linking solutions “until each problem and sub-problem resolves itself into one that can be physically dealt with — until the chain is fully in place”. And Arthur defines technology very widely, to include “industrial processes, machinery, medical procedures, algorithms, and business processes”. As Arthur states, ‘we now have a system where novel elements (technologies) constantly form from existing elements, whose existence may call forth yet further elements”. So technology begets more technology…and opportunities.

There is something of a debate in academic circles about whether technology precedes a problem or whether the problem comes first and then the technology is created to solve it. I think the problems appear first, then the technology is “created” and then the innovative solution is offered in the market. The “new” technology or invention has to go to market [for public scrutiny and evaluation] to be an innovation and fulfill the opportunity. Once the technology is created to solve a particular problem, it is not uncommon that the technology be applied to a new problem. John Chisholm — MIT grad, entrepreneur and Forbes columnist — wrote an interesting article[13] that inspired this chapter in the article (and my interest in the complexity studies at SFI). He and his senior technologist each pick a collection of distinct technologies that they understand and are knowledgeable about. Then they combine a technology from each of their collections and determine whether the combinatorial result can solve a problem.

Two examples clearly illustrate combinatorial thinking. The laser printer was constructed from the existing laser, digital processor, and xerography. Motorola built the company around radio technology and the technical adaptions through combinatorial thinking that allowed them to pioneer car radios, public safety radios, WWII walkie-talkies and cellular phones. When it was sold to Google in 2011 for approximately $12.5 billion, analysts said that Google only valued Motorola’s intellectual property and not the operating business.

Simpler examples show combinatorial thinking with fewer technologies combined. Take for example the electrical motor. Vacuum cleaners, refrigerators, blenders, blow dryers, washing machines, dishwashers, water heaters[14], golf carts and now automobiles all use the simple electrical motor. Another good way to show combinatorial thinking is to consider liquids. Think about all the new products and opportunities that were made possible by a liquid being mixed with an additive — — lead added to paint to prevent growth on boat bottoms, children’s drinks enriched with every vitamin, thickener added to bleach so it does not splash, citrus flavoring added to vodka, salt mixed with water to create saline solution, engine cleaning additives in gasoline and on and on. Need a new opportunity…just find an additive for a liquid that solves the customer problem. Speaking of liquids makes me think of cooking as an example of combinatorial process. We see the combinatorial process in sauces, main dishes, adding sides and then we add to each of them, e.g. raisins in cereal or yogurt.

By now I am sure you are all experts on applying combinatorial thinking and have come up with many, many of your own examples that are better opportunities than mine. Congratulations. Now please explain why it took 6,000 years to put wheels on a suitcase. Naseem Taleb asked this question in his book, “Anti-Fragile”, and there is still no obvious answer to the question.

As much as we humans enjoy combining things to be creative and identify opportunities, that which can be put together can also be taken apart. Humans are great at decoupling things in order to solve a problem or make things better. Almost all the great thinkers cite the combinatorial process as the source of creativity, but decoupling challenges combinatorial thinking, in my opinion, for the crown. Decoupling is the simple idea that one can create value by separating or eliminating the parts of a whole. Remember the Negroponte Switch from the last chapter. The Switch is not only an example of functional fixedness but also an example of decoupling, when the cords are eliminated from the appliances (such as phones). Stephen J. Gould and Stuart Kauffman, well-known complexity scholars, refer to decoupling as “exaptation” or “a novel function for a part of an existing entity”.

I think in my lifetime that money will no longer be physical paper and will just be bits and bytes of code somewhere. From its earliest days money has been a store of value, whether it be gold coins, bills, bank accounts or bitcoins. However, every time someone overcame their cognitive bias the store of value changed its “form” through decoupling. Users accepted these changes because they satisfied an economic and emotional need. Another store of value is debt or a promissory note. However, why should we consider the obligation in the note to repay principal and interest to be linked. If we separate the note into two separate obligations for principal and interest we create a derivative. We could further strip out the currency that the obligation is denominated in. With some idle time on our hands we could further break up the principal repayment obligation into multiple tenors. Now totally bored we could link the amount of principal to be repaid to the obligor’s stock price. Decouplings creates equity, commodity and foreign exchange derivatives for any financial instrument. The Bank for International Settlements estimates that at December 2012 the notional value of derivatives was $635 trillion, which has to be the largest dollar denominated example of an opportunity from decoupling. As soon as one holder hedges their derivative risk, we, of course, create a new derivative. As future holders of a derivative hedge their position, we start to see autocatalysis, which explains in part why the derivative market is so large. Note: autocatalysis is a characteristic of several very large market opportunities and is discussed in several places in this article. Autocatalysis is a situation whereby the product of the process initiates the process again.

Music has gone through almost the same decoupling as derivatives — first albums to songs, then from vinyl to tape to CDs to iTunes. We now have specialists who mix new lyrics with existing musical scores. It is not hard to see that books have gone through the same changes both in form and in forum. Originally books were only owned by the rich, then found their way to libraries and now are ubiquitous through Kindle and similar technology, all different locations.

Skype is an easy way to demonstrate decoupling. Take away the landline (decouple) and complete the telephone call or data exchange over the Internet. Other well-known examples of decoupling are Google’s driverless car, MOOC’s (massive open online courses) where we separate classes from schools, home entertainment where we separate the show from the theatre and outsourcing where we separate certain requirements such as manufacturing from the business. Retailing decoupled from the store with the first online shopping at Comp-U-Card in about 1980.

Computing offers countless examples of decoupling. First we might mention the cell phone where the entire computing function moved to another device or the iWatch or Google glasses. The combination of hardware and software, which might be a simple definition of the computer, changed when we move the apps to the web as “software as a service” or through Google Chromebooks that have no apps resident on the device. Cisco has been queried countless times about whether it would sell its software as a standalone product, separate from its network hardware. The cloud first decoupled data storage, then processing and the Bitcoin Blockchain is perhaps an example of decoupling where the distributed processing is organized for the first time voluntarily. We have even reached the point where we are decoupling the software stack, the combination of operating system and applications (screen management, browser, etc.), so companies develop only the particular app or suite of apps that creates the most value for the customer and all the other software is licensed or open source. GitHub claims 25 million code projects available for use, which perhaps proves the point that more and more people are using other people’s code to complete the stack.

If we look at general themes from these examples of decoupling, we see locations and content separating (e.g. libraries, stores, theatres). We see information in all forms, including music, decoupling. Money is constantly being decoupled, now through derivatives. We see human operators separated from their machinery (driverless cars and many other machines) and we see computing constantly decoupling down now to the code level.

If we try to look ahead, we will see AI added to everything from cars to planes to lawn mowers. Barney Pell, a well-known former NASA researcher and entrepreneur in residence at the Silicon Valley venture capital firm Mayfield, has coined the Pell Law that basically states that the first company to successfully introduce AI into a product changes the industry’s product development strategy to focus first and foremost on AI. (Pell also believes that as applications and APIs become more sophisticated and specific, AI has the advantage over humans in realizing the benefits.) This combinatorial process will also lead to a higher frequency of decoupling with, for example, humans being the “component” separated or no longer needed.

There will be interesting opportunities to re-purpose real estate such as schools and libraries and theatres as we become increasingly comfortable with digital content, virtual presenters and no live performers. I think we will see an increasing combination of features built into buildings as we look to reduce the carbon footprint, increase sustainability and look to manage better the ever-larger cities. For example, why is the roof of every office and apartment building basically empty? This space is perfect for solar panels or vegetable gardens or composting pits.

Another big opportunity involves new private sector providers for government services. I believe that cheaper computing power combined with big data and AI will lead to a transformation in society similar to that at the time of the first automobile. These technologies will lead to people being able to takeover more of the current role of government, wherein lies the decoupling. A reduction in the scale of government is a decoupling. One way that the government might survive at its current scale and scope is if it became a better provider of information. If the government could become a low cost provider of publicly available big data information, we might see a new value in government. Building products and services on top of this publicly available data may be a large combinatorial opportunity. I think the government may become such a big data provider, but the government may become jealous of the private sector companies making large profits or eventually want to charge for the data. Probably easier to just reduce the federal government to the military, the Justice Department, a small executive branch, an environmental policy and research group and maybe an organization that funds scientific research. An automated national sales tax, with certain people exempted, would fund these programs.

One other view of the combinatorial process in the future comes from combining the physical and the digital. For example, some car manufacturers are replacing physical dashboards with holographs. Another example is using augmented reality (AR) glasses to insert information into the physical reality. As you approach a person the glasses give you their name based on facial recognition. While these examples are rather limited today, in the near future they will be commonplace.

Note: the combinatorial process looks very similar to a property found in biology. An emergent property in biology is a property in a complex system that the individual agents do not have. For example, proteins beget cells that beget tissue that beget organs, a natural example of the emergent and the combinatorial. If all of biology and the natural world follow this concept of emergent properties, it is perhaps no surprise that combinatorial processes are such a source of market opportunity. If we look at mankind’s “development” as a collision between nature and technology, the force behind the collision on both sides is a combinatorial process — that we call creative when we talk about technology and evolution when we talk about biology.

6— Change the Assumption

“When everyone agrees on some fundamental assumption about how the industry works, the opposite point of view can lead toward disruption.” FarnamStreet

Henry Ford became famous in 1907 when he introduced the black Model T Ford, an outstanding example of combinatorial thinking based on the gas engine. Each semester I ask my entrepreneurship students to create a new business by changing just one assumption about the Model T. The first student responses produce car paint, custom parts and trucks. The second group of responses includes taxis, rental cars, car insurance and toll roads. The more thoughtful students mention motels and parking lots, where they change not the car but the driver’s behavior. The best response I have gotten was a student who proposed fast food restaurants…but he was a hospitality major. Each of these large business opportunities came from changing just one assumption about the product or the user behavior.

There is no literature that directly discusses the cognitive inability to change one assumption, but the concept looks consistent with Kahneman’s concept of ‘fast thinking”. Anchoring, a documented cognitive bias, comes close to explaining the inability to change a single assumption. Anchoring is the tendency to rely on the first piece of information as presented. We are programmed not to explore new assumptions. Tunnel vision is another cognitive bias that may explain why we cannot consider new evidence in decision-making or change an assumption. While cognitive science does not yet fully explain the inability to change assumptions, complexity science may provide an answer. Stuart Kauffman writes:

“When searching the space of possibilities, whether for a new product or a different way of doing things, it is not possible to explore all possibilities. It may, however, be possible to consider change one step away from what already exists. In this sense, exaptation may be considered an exploration of what is sometimes called the ‘adjacent possible’. That is exploring one step away, using ‘building blocks’ already available, but put together in a novel way…. the push into novelty in the molecular, morphological, behavioral, technological and organizational spheres, is persistent and happens through exploration”[15]

So a biological behavior or tendency from complexity science may explain the “cognitive bias” and the power and frequency of the innovation and new opportunities that come from changing one assumption.

Retailing demonstrates clearly the power of changing just one assumption. The first retailers sold from their homes (caves, tents, lean-tos?). Then they changed an assumption, abandoned their homes and sold in marketplaces. Then the retailers realized that stores in the newly emerging cities provided a better shopping experience and protection from the elements. With the advent of the computer, selling moved online with companies such as Comp-U-Card and then Amazon. Changing one assumption — to sell for immediate cash payment — rental and leasing of products and then services emerged. There are many more examples in retailing where a simple change in one assumption led to large opportunities — gift wrapping, lay-a-way, door-to-door selling, to name a few. It is well documented that about 70 percent of a purchase decision is emotional. Any change in assumption that further engages the shopper’s emotions is likely to represent a big market opportunity.

Many other examples demonstrate the large market opportunities created by changing the assumption. One category is the case where we separate the forum from the content. The first theaters were organized in Greece in the 5th century BC. A theater is a combination of a comfortable place, entertainment (the content) and (preferably) paying customers. Both television and radio took this entertaining content and made it available outside a theater. Movie theaters with recorded content would be another example. Continued innovation by separating “forum” and content leads to Netflix, You Tube and Chromecast. Other examples where we separate content and forum to create large opportunities would include taking books out of libraries (originally the only place to find books) to create bookstores and then Kindles. MOOCS, massive online open courses, is another example. The MOOC separates the course from the university. Now MOOCs are evolving to be considered a collection of lectures where the student picks just the lectures of interest, which is made possible by changing the assumption about what a course is.

Lycos and Yahoo commercialized Internet search. In the [last] section of the article I will explain the real reason why I think Google was so successful and took search to a whole new level of customer acceptance. However, from Google’s success came You Tube, image search, Wolfram Alpha and probably Pinterest and Instagram. We can debate whether the assumption changed in Pinterest and Instagram was the search content or the method of curation, but Google plus one new assumption created them. You Tube continues to evolve commercially through the change of single assumptions. In fact, in many ways the evolution of You Tube mirrors the history of cable TV in terms of content, commercialization and talent. A great example of change the assumption comes from the cable TV industry. The story goes that John Hendricks, the Founder of Discovery Channel, was watching CNN, the first cable news channel, and realized he could create a new channel by switching out the news and offering only documentaries. Thus was born Discovery Channel.

The management thinker and professor Peter Drucker noted the power of an assumption in developing an opportunity or business concept. In a 1994 article in Harvard Business Review, Drucker outlined his “Theory of Business”. Drucker stated:

“Basic assumptions about reality are the paradigms of a social science, such as management. They determine what it focuses on. Yet, despite their importance, they are rarely analyzed, rarely studied, rarely challenged — indeed rarely even made explicit. What matters most in a social discipline such as management are therefore the basic assumptions. And a change in the basic assumptions matters even more. “

An example from the European Innovation Center illustrates Drucker well. Consider Apple’s iPod, the first in a long list of successful Apple products that serve up music. In considering the product there were three key assumptions:

1. Will people pay

2. To download music

3. To listen to music in public

Napster had already shown that people would download music and Sony Walkman showed that people would listen to music in public. The key assumption was whether people would pay for music through downloads. Now Apple’s very successful history in music demonstrates one of the best cases in history for change the assumption. Downloads first appeared on the iPod and then the device was changed and iTunes was available on the iPhone, the iPad and then the iWatch. Apple took their amazing insight about music and transferred it across multiple devices, every time changing the device assumption. One insight (iTunes or App Store) spread across multiple devices or technologies is a great way to find a large opportunity. To digress for further clarification, Drew Huston’s insight at Dropbox was completely different from Apple. Huston realized that people were going to want the same information available on all their devices. The assumption that computer-based information should be kept in a single device was about to change. Apple, on the other hand, built an infrastructure that could serve multiple devices for music and then built those devices.

Drucker goes on to explain that there are three types of assumptions:

1. The environment of the organization — society and its structure, the market, the customer and technology

2. The mission of the organization

3. The core competencies needed to accomplish the organization’s mission

What I find most interesting are numbers 1 and 3. In 1 Drucker categorizes the environment in terms of society, market, customer and technology. These four categories are the four types of assumptions that can be changed when one examines any existing product for a new, large opportunity. In 3 Drucker is a bit obtuse in my opinion. I think what he really should have said is that an organization needs core competencies in management to match the three to four key assumptions in the business. As Professor Rita McGrath at Columbia University makes clear in her writings, the assumptions represent the risks in the business.

Another interesting example of changing an assumption comes from my friend Konstantin Gregoriou, who reviewed early drafts of this manuscript. He points out that by changing one assumption about the application of a business model one might find a large market opportunity. For example, the sharing business model of Uber, with one assumption changed, becomes Airbnb. As hardware and value chain become less important and software and value networks become more important, more and more large opportunities will come from applying existing business models to serve new customer needs.

If we look ahead to the future, one assumption I see changing is size. Rather than building the biggest homes, offices and buildings, I think we will reduce size in order to conserve resources and reduce the stress on the environment. I see smaller apartments, more office sharing and more construction where materials are re-used. Re-cycling and re-using materials is a simple change in assumption that might represent a big opportunity if we ever commit to conserve natural resources.

Another opportunity I see is if education continues its trend away from a hierarchical approach toward a bottom up methodology. In the early part of the twentieth century Swiss psychologist Jean Piaget put forth the idea that learning should be child-centric and individually directed. This notion conflicted with the then popular government mandated top down system (to train factory workers) that remains in place to this day in most countries. So if we flip the switch from top down to bottom up (the assumption), we might see a dramatic increase in peer-to-peer learning, more certificate programs and less diplomas, more formal education for people at a later stage in life, more curation technology to support individual learning and more products where purchasers can talk, chat or text with authors, artists and musicians in real time.

If we refer back to Chapter III where we discussed asymmetry of information, I think that many opportunities will exist to eliminate the absence of information. The assumption that will change is that a particular market, product or industry will continue to have a negative asymmetry. Healthcare immediately comes to mind, for example, when we consider how much more medical information will be archived from remote testing and diagnosis. Overcoming the HPPA issues (U.S. federal laws governing the privacy of individual health information) and combining various archives, the shopper could analyze and validate the actuarial assumptions for life insurance and the related pricing or perhaps buy the better priced product from an unregulated company in another state or country.

Today there are 28 mega-cities with populations over 10 million according to the United Nations[16]. In forty years most of these cities will have populations of 30 million or more, roughly equivalent to the 2015 population of Malaysia. We may see a return to the city-states of Phoenicians times. If one examines carefully the success of the most successful city-state in history — Singapore — one can put forth an argument that maybe Shanghai or Mumbai would provide a better quality of life to their citizens as city-states. Where is the opportunity? Answer: anticipating when the change to city-state will happen and providing the services that a new city-state might need. The economist Paul Romer believes that a variation on city-states could be used to develop parts of poor countries. Rohmer calls them charter cities and describes them as: “economic zones founded on the land of poor countries but governed with the legal and political system of, often, rich ones”[17].

Perhaps too obvious to mention, but solar power will replace fossil fuels within 30,40 or 50 years. We might consider such a change as a paradigm shift (discussed in Chapter XI), but I prefer to think of it here as a merely a change in the fuel assumption. The opportunities such a change will make possible include all means of devices and locations to recharge cars (that will have no human driver) and opportunities for all new types of batteries. A senior executive at Quanta Computer, the large Taiwanese electronics outsource manufacturer, once told me that the only breakthrough left in computing technology was in batteries. Batteries look like several large opportunities depending on your assumption about the underlying device.

Ray Kurzweil of MIT and Google fame predicts that by 2020 scientists will be able to turn off fat cells, preventing the body from adding fat and perhaps presenting a strategy for controlling diabetes. The assumption that we will be able to selectively turn off and on cells opens up many curative possibilities. As Kurzweil says, “thanks to the human genome project, medicine is now information technology and we are learning how to program this outdated software of our bodies exponentially.” [18]

Cameras are part of our fascination and need for visual stimulation and information and have changed countless times over the last 100 years. Many times just one assumption changed. One example is film changed to digital, then cameras moved to phones. In the future we will all make much greater use of the still and full motion cameras located in drones. Dronebase and other drone-based services might be an example of the large opportunity in this space.

At this point the reader might think that every new opportunity mentioned in this article can be explained by “change the assumption”. This is probably true but hopefully some of the other ways to find opportunities are more effective or faster ways to get there in certain cases.

7— Complexity

“…current risk management systems and other attempts to predict the future are based too much upon linear relationships derived from past experience. They fail to take into account our behavioral limitations in handling probabilities, and also the nature of complex non-linear systems that do not always have a definite or repeatable cause and effect relationship. We need, therefore, to consider a new way of looking at the world around us and a new way of thinking about issues.” [19] Lam Chuan Leong

A traditional approach to entrepreneurship by both scholars and practitioners is to start by identifying the potential customer problem or need. If one were to think of the taxonomy of problems, most people would respond that there are two types of problems:

1. Simple

2. Complicated

A “simple” problem is, for example, fixing a flat tire. If one jacks up the car, removes the tire and replaces the tire and nuts properly, one safely goes on their way time after time after time. A “complicated” problem might be producing a new vaccine.[20] We have a set of steps, processes and know-how that if we follow diligently in the end will produce a vaccine. We just do not know what the end product will be until we get there. This two-part division of problems characterized much of thought up until very recently. Writing on such intellectual tendencies, the noted Harvard education professor Howard Gardener writes about the innovative French scholar Michel Foucault

“Foucault argued that historical eras are characterized by certain underlying (and typically unconscious) assumptions about the nature of knowledge. Assuming such a structured stance vis-à-vis the seventeenth century, Foucault discerned the same taxonomic assumptions about knowledge operating in such diverse fields as biological classification, economics and linguistics.”[21]

Beginning around the middle of the 20th century a few intellectual luminaries such as FA Hayek and Herbert Simon began talking about a third class of problems. At first the term “complexity” was used in the vernacular but gradually the term took on an increasingly precise definition amongst scholars (although scholars in different fields may not use the same definition). Outstanding scholars such as Edward Lorenz, Stuart Kauffman, Murray Gell-Mann and W. Brian Arthur did much of the early groundwork to define complexity (working in different fields). Following the logic of Foucault, the 21st century will mark the recognition that complexity will join the taxonomy of knowledge as the third type of problem. As Stephen Hawkings said it so well, “Complexity is the science of the 21st century”.

Complexity will shape knowledge across many fields including entrepreneurship.

As we will shortly see, entrepreneurship is an example of complexity and we may draw many valuable insights about entrepreneurship from such a realization. However, this is an article about identifying large market opportunities and we will examine complexity here principally as it helps us to understand such large opportunities. One other preliminary note is required about the fundamental relationship between complexity, organizations and networks. I have chosen to separate out individually the discussion of organizations and networks in the following two chapters to hopefully more clearly explain where large opportunities may be. The other point to note is that even if the reader does not accept this approach to explaining complexity, networks and organizations as stand alone concepts are fertile areas to find large opportunities.

To begin the discussion, let us define complexity in general terms. Later we will examine the economic complexity inherent in entrepreneurship. Complex systems are non-linear (non-deterministic) with inter-related variables or agents [the network] where results are self-organizing and cannot be determined from the characteristics of the agents. The parts of complex systems are many, inter-connected and intricate in their connectedness and the nature of the relationships between the parts may be unknown (even though results may appear regular over certain time periods despite the absence of cause and effect). No single part of a complex system can perform the function of the system but properties may emerge for the system that are not properties of the variables or agents. Therefore, complex systems, given enough time, prove to be unpredictable both in terms of result and scale of result, what Naseem Taleb called a “black swan”. Also, of modern interest, in complex systems the important attributes, parts or sub-systems may currently be beyond modeling by computational processing. While AI will help us to understand complex systems, it will not wrestle this steer to the ground. Some scholars are probably insightful when they say that the only problems remaining after the popularization of AI will be complex problems.

The characteristics of a complex system, from P. Ferreira’s 2001 MIT teaching notes[22], are shown below.

Complex systems are found everywhere and in a wide range of domains. One type of complex systems are the complex adaptive systems, characterized by the ability of their agents to explore and exploit the environment. Perhaps the most well known example is the evolution of living things, which probably explains why almost all interesting man-made systems are complex. If a cell is a complex system then it stands to reason that everything made of a cell, such as a human, must be complex or at least exhibit complex behaviors. FA Hayek perhaps said it best:

“Not only are individuals themselves complex orders, but their perceptions, beliefs and motivations are also the result of a complex apparatus for cognition that interprets a complex order of actions as a basis for their own actions”

Complex adaptive systems include markets or marketplaces, firms and all economic activity, which is where we turn now in the exploration of complexity.

Will and Ariel Durant in their well known book, The Lessons of History, write:

“Evolution in man during recorded time has been social rather than biological: it has proceeded not by heritable variations in the species, but mostly by economic, political, intellectual, and moral innovation transmitted to individuals and generations by imitation, custom, or education… New situations, however, do arise, requiring novel, unstereotyped responses; hence development, in the higher organisms, requires a capacity for experiment and innovation — the social correlates of variation and mutation.” [23]

Melanie Mitchell, the Santa Fe Institute complexity scholar, builds upon the Durants’ point.

“As in all adaptive systems, maintaining a correct balance between these two modes of exploring is essential. Indeed, the optimal balance shifts over time. Early explorations, based on little or no information, are largely random and unfocused. As information is obtained and acted on, exploration gradually becomes more deterministic and focused in response to what has been perceived by the system. In short, the system both explores to obtain information and exploits that information to successfully adapt.[24]

(bold inserted by the author)

[The Nobel Laureate] Herbert Simon provides his own very lucid take on complexity and economic complexity.

“Evolutionary processes are significant not only for explaining organizational loyalty, but also for describing and explaining the historical development of economic institutions, including business firms. The simplest scheme of evolution depends on two processes: a generator and a test. The generator produces variety, new forms that have not existed previously, whereas the test culls out the generated forms so that only those that are well fitted to the environment will survive. In modern biological Darwinism, genetic mutation and crossover of chromosomes are the principal generators, and natural selection is the test.” [25]

Simon’s description of complex economic systems sounds eerily similar to my description of entrepreneurship in the Introduction to this article, as shown below:

“Execution determines the speed at which the company grows [the test] and probably its capital efficiency, but it is the market opportunity that determines the potential size of the company” [the generator].

Simon’s “generator”, what the Durant’s called “experiment”, determines the market opportunity and by deduction the size of the opportunity. Simon’s “test” “culls out the generated forms so that only those that are well fitted to the environment will survive”, which sounds like product-market fit. Durant’s use of the familiar term innovation or what other writers including Mitchell call exploitation looks to me like Simon’s “test”. I particularly like the Durant’s notion that innovation is tested in the “market” and not merely a synonym for invention. If perhaps you think that “test” is shorter than “exploitation”, I would point out that a test for Simon would naturally scale as long as self-interest is served. Such scale gives “test” an exploitive nature.

Mitchell explains the role of both exploration and exploitation in complexity in terms of information, as re-stated below:

“the [complex] system both explores to obtain information and exploits that information to successfully adapt”

Some would say that all of complexity in any form including social systems can be explained by Shannon’s seminal work at Bell Labs on processing information and entropy (the second law of thermodynamics), which explains how information is organized and conserved. For further discussion of information theory, entropy and complexity I recommend MIT Professor Cesar Hidalgo’s book, “Why Information Grows: The Evolution of Order, from Atoms to Economics“. To confirm Mitchell’s thinking on the role of information in economic complexity, I cite Hayek:

“The peculiar character of the problem of a rational economic order is determined precisely by the fact that the knowledge of the circumstances of which we must make use never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess.” [26]

Herbert Simon perhaps completes the analysis of complexity in economics by introducing the concepts of subjectivity, environment and processing power into the discussion of exploration and exploitation. Simon states:

“Complexity emerges from the richness of the outer environment, both the world apprehended through the senses and the information about the world stored in long-term memory. “[27]

The question that you might be asking is why this process to obtain information happens. Lesser organisms explore and exploit in order to survive. Humans act from a similar motive — the all-familiar self-interest — the modern day notion of survival. Simon references “long term memory”, by which he means both computational as well as archiving capacity. To quote Simon:

“Economic systems, just like all natural systems, have an ability to produce information that is constrained by the systems’ computational capacity. For information to grow in the economy, the computational capacity of the economy needs to grow as well. Increasing the computational capacity of economic systems, however, is not easy, since the growth of an economy’s computational capacity is constrained by the ability of people to embody knowledge and knowhow in networks of people.”

This quote is one citation where Simon presents his seminal concept of “bounded rationality”, the limitation in cognitive processes that enables an individual to make never more than an optimal decision based on the limited information available. The concept of bounded rationality is of course the basis for his award of the Nobel Prize in 1978.

The last idea that needs some mention is the notion of “environment”, which we interpret in the context of information. Complexity naturally involves the environment and much of human development can be seen as measures to simplify the environment in order to more easily apply the information processing of Simon’s bounded rationality and conserve energy. This natural instinct for energy conservation was explained well in Kahneman’s Thinking Fast and Thinking Slow. Humans look to bring simplicity and order to information in complex systems. (Organizations and networks are two tools to achieve simplicity in complex adaptive systems and are discussed more fully in the next two chapters.) “Framed in this way, information is not, as in a traditional computer, precisely or statically located in any particular place in the system. Instead, it takes the form of statistics and dynamics of patterns over the system’s components. …However, randomness must be balanced with determinism: self-regulation in complex adaptive systems continually adjusts probabilities of where the components should move, what actions they should take, and, as a result, how deeply to explore particular pathways in these large spaces”[28], according to Mitchell. This notion of “adjusting probabilities” introduces into a complex adaptive system the concept of risk, which is perhaps the one component from traditional economics (and survival) that we have not yet addressed. Probabilities also explain the combinatorial nature of creativity, which is required to explain how new ideas come from joining together existing information and the perception of such information during exploration.

By now the reader hopefully recognizes that many of the characteristics of entrepreneurship (and Lean Startup, design thinking, and economics) can be explained by complex adaptive systems. In simple terms, exploration and exploitation is nothing more than Israel Kirzner or Howard Stevenson’s definition of entrepreneurship, which I paraphrase [together] as… the entrepreneur has an asymmetry of information or insight (exploration) upon which they build a company to pursue [exploit] the opportunity. Lean startup and design thinking both rely on the concept of exploration as a fundamental part of their processes. Entrepreneurship behaves in a way consistent with the principles of complex adaptive systems, suggesting to me that entrepreneurship is the natural method for economic development. (In the next chapter on network we will explain why government should “naturally” be downsized.)

If we summarize the basic tenets of complex adaptive systems with respect to information, with the objective of potentially identifying large market opportunities, one might cite these factors:

1. Information is available

2. Information is organized (simplification as used in entropy)

3. Information is archived

4. Computational power is functioning

All of these factors are, by definition, present in other forms of complex system in nature, but that fact does nothing to help us understand large market opportunities.

If we look at large opportunities that have most of the four characteristics above, we see the representative large opportunities shown in the exhibit below.

What we see from the exhibit is that the largest opportunities in the information space, such as Facebook, Cad Cam and Wolfram Alpha, all combine organized, archived information and computational power. One might realize that books and libraries have lost “market share” to digital alternatives because while they combined organized and archived information, they lack the feature of providing computational power. In fact the popularity of the Amazon Kindle and other e-book readers may be that they insert a semblance of computational power into books. If one considers the loss of market share for local computer storage in favor of the cloud, I would offer that the cloud provides a better experience by combining all four factors whereas local storage generally lacks significant self-contained computational power to manage the storage. Note: This insight about the combination of information storage and computation is, of course, the insight of Alan Turing that lead to the design of the first computer. This combination of storage and computation is also the unique characteristic of the fundamental organic material, RNA, which some say explains all living things.

The continued integration of computational power and storage is one interesting opportunity suggested by complexity theory. For example, databases that change their factors based on the data collected is an interesting opportunity being researched. Another example is two companies, Cohesity and Coho Data, that insert computing power (operations) into storage (that was formerly done at another place in the software stack). These companies are creating databases that program their own APIs.[29]

Complexity suggests that the opportunities may arise in either exploration or exploitation of information. Exploration suggests that combining big data with AI will speed up exploration making the search for information and analytical findings more efficient and cost effective. The further integration of AI and big data will be a big opportunity, probably offered in a SAAS model. This integration is one way that Google reportedly plans to compete with Amazon and Microsoft in the SAAS market.

Another way that the complexity theory view of information explains large market opportunities is to look at miniaturization. Miniaturization represents the more and more efficient organization of information. The computer processor makes the point clear. Originally the processor was housed in a building and now we hold it in our hand. Another way to look at it is that originally cameras rested on large tri-pods. Today they are merely a small component in a phone. In the future screens will disappear, replaced by your arm or a virtual projector of the image. This miniaturization, the complete disappearance, would definitely demonstrate more efficient organization of information.

Note: The ever-popular Lean Startup intentionally or unintentionally models itself after “explore and exploit”, a fundamental characteristic of complex adaptive systems. Talk to customers and iterate (explore) until you reach product market fit, at which time invest heavily and grow (exploit) — a simple summary of Lean Startup.

8— Networks

“In considering the study of physical phenomena, not merely in its bearings on the material wants of life, but in its general influence on the intellectual advancement of mankind, we find its noblest and most important result to be a knowledge of the chain of connection, by which all natural forces are linked together, and made mutually dependent upon each other; and it is the perception of these relations that exalts our views and ennobles our enjoyments.” [30] Alexander von Humboldt

In the last chapter we learned that complex adaptive systems, such as social and economic systems, are non-deterministic, self-organizing systems that process and store information. The dynamic tension between exploration and exploitation makes a complex system adaptive. The behavior that cannot be ascribed to any individual part of the leaderless system is the emergent quality of complex systems, which is more easily understood in the context of networks, which are the subject of this chapter.

One characteristic of complex systems in both natural and social systems is networks. The term network refers to the framework of routes within a system of nodes. Nodes are the depositories of information (e.g. people, websites).

A route is a single link that can be tangible or intangible between two nodes. Networks can be physically constrained, such as transportation systems, or non-spatial, such as certain social and economic systems. Cities might be an example of the later.

Examples of networks and their role in the history of economic development is shown in this quote from Jean-Paul Rodrigue and Cesar Ducruet’s article, “The Geography of Transportation Networks”:

“Transportation networks have always been a tool for spatial cohesion and occupation. The Roman and Chinese empires relied on transportation networks to control their respective territories, mainly to collect taxes and move commodities and military forces. During the colonial era, maritime networks became a significant tool of trade, exploitation and political control, which was later on expanded by the development of modern transportation networks within colonies. In the 19th century, transportation networks also became a tool of nation building and political control. For instance, the extension of railways in the American hinterland had the purpose to organize the territory, extend settlements and distribute resources to new markets. In the 20th century, road and highways systems (such as the Interstate system in the United States and the autobahn in Germany) were built to reinforce this purpose. In the later part of the 20th century, air transportation networks played a significant role in weaving the global economy. For the early 21st century, telecommunication networks have become means of spatial cohesion and interactions abiding well to the requirements of global supply chains.”

Carlota Perez is a history of economics scholar who has devoted much of her research and analysis to understanding paradigm shifts or more simply put — technological revolutions. In a paper in 2004, “Finance and Technical Change: A-Neo-Schumpeterian Perspective”, she includes a graphic, shown below, that traces each of the major technological revolutions, starting with the Industrial Revolution in 1771.

If one examines each example of the “New or Redefined Infrastructures” (Column 3 above), in each paradigm shift there is a new, significant form of network that serves as the infrastructure to support and facilitate the shift. If one accepts Perez’s analysis, this graphic clearly demonstrates the role of networks in the history of economic development and by extension in entrepreneurship.

When one considers an explanation for the close link between paradigm shifts and networks, traditional economic considerations of production, distribution and consumption provide me with no real insight. However, if we return to the insights of Ronald Coase, we see economic activity in a less traditional way as a combination of property rights [information], arrangements for collective choice [collaboration/feedback] and contracts for motivating managers and employees [social exchange/signaling].[31] Stepping back, what one realizes about economic networks is the efficiency a network provides. Networks provide connectivity, communication, operations and management, all in a self-organizing mechanism for [pursuing] information. Networks are nature’s answer to the Swiss Army knife.

Networks are such a “popular” and versatile mechanism for four reasons:

  1. Networks lower the cost of searching for information
  2. 2. Networks lower the cost of verifying information

3. Networks lower the cost of processing and storing information

4. Networks lower the friction in exchanging information

Economic and social networks achieve these benefits in part through trust amongst participants, which we discussed in Chapter I. The further disclosure and transfer of information within the network builds the trust and fosters the organizing, processing and archiving of information. Trust also lowers transaction costs, thereby facilitating the construction of larger networks.

Herbert Simon writing on hierarchies [networks] cites three reasons why they are so common[32] :

1. Networks facilitate the formation of complex systems (see Metcalf’e’s Law below)

2. Networks have direct channels of communication (connectivity)

3. Networks are naturally redundant (lower transaction cost)

Strengthening the versatility of networks is Metcalfe’s Law, which says that networks follow a scale-free power-law distribution. (Every additional node in a network increases the value of the network.) As Albert Barabási explains it, “This feature [power law] was found to be a consequence of two generic mechanisms: (i) networks expand continuously by the addition of new vertices [nodes], and (ii) new vertices attach preferentially to sites that are already well connected. A model based on these two ingredients reproduces the observed stationary scale-free distributions, which indicates that the development of large networks is governed by robust self-organizing phenomena that go beyond the particulars of the individual systems.”[33] This “robust self-organizing phenomena…beyond the particulars of the individual systems” is the emergent property present in all complex systems.

Mitchell makes an interesting point about the size of networks:

“Self-regulation in complex adaptive systems continually adjusts probabilities of where the components should move, what actions they should take, and, as a result, how deeply to explore particular pathways in these large spaces.”[34]

Mitchell’s probabilities, what might in the vernacular be called uncertainties, are discussed in more detail by JR Galbraith:

“the greater the uncertainty of the task, the greater the amount of information that must be processed between decision makers during the execution of the task to get a given level of performance”. [35]

This rather simple observation explains the evolution of “organizations”, which is the subject of the next chapter, and leads to two observations:

1. In small, resource constrained networks there is usually a large node or organization that dominates

2. In large networks the need for a large, dominant node is reduced (because of the distributed information processing power)

This relationship between network and the number of organizations [node] explains why early U.S. colonies required a federal government. Conversely, with today’s large, global, interconnected networks, perhaps we can downsize federal government in the U.S.

The relationship between networks and entrepreneurship is only now emerging, mostly due to the growing fields of network science, information theory and complexity economics. However, in some ways the practitioners are ahead of the academics in their understanding of this relationship. Notable venture capitalist Fred Wilson of Union Square Ventures sees the establishment of a network as a competitive advantage that prevents competition from entering a market. Peter Thiel of PayPal fame recommends that startups go after small markets where dense networks can be created. (Giulio Tononi’s Integrated Information Theory uses “dense network” as a measure of how much more a system is than the union of its parts.) Thiel sees networks as a mechanism to achieve monopoly, his preferred position in any market. Facebook’s eclipse of the earlier MySpace shows, however, that networks are not a panacea or invincible business model. An even better example of the network model is Google. Google’s search algorithm targeted nodes with a large number of connecting links, just what Barabasi explained about networks when he said, “new vertices attach preferentially to sites that are already well connected”. The insight here is that the Google algorithm followed the pure theory of power laws and networks and the market opportunity proved to be quite large. Google’s approach also used autocatalysis, a characteristic of some complex systems, where the product of the search reinforced the importance of the information in future searches.

Academic research has shown that companies using the network business model create more shareholder value. In an HBR article, “What Airbnb, Uber and Alibaba have in Common” [36] the authors analyzed companies in the S&P 500 over a forty-year period starting in 1972. Companies were categorized as one of four types:

1. Asset Builders

2. Service Providers

3. Technology Creators

4. Network Orchestrators

Companies that were network orchestrators showed “higher valuations relative to their revenue, faster growth, and larger profit margins”. The researchers also discovered that only five percent of S&P companies are network orchestrators. The authors explain the value creation, “We believe this occurs because the value creation performed by the network on behalf of the organization reduces the company’s marginal cost, as described in Jeremy Rifkin’s The Zero Marginal Cost Society.” Looking for a more network-oriented explanation, I would think that the scarcity of network operators perhaps shows the challenge of successfully building and sustaining a network model. The efficiency of network value creation perhaps demonstrates Michael Porter’s findings that the competitive advantage [in successfully building and sustaining a network] is a key requirement for extraordinary value creation.

As we look to the future and the market opportunity offered from our understanding of networks, The Second Machine Age perhaps provides some guidance:

“The winners are no longer those able to compete solely based on cheap labor or ordinary capital, both of which are being squeezed by automation. … Fortune will instead favor a third group: those who can innovate and create new products, services, and business models. … So in the future, ideas will be the real scarce inputs in the world — scarcer than both labor and capital — and the few who provide good ideas will reap huge rewards.”

The Manifesto 15 Handbook discusses a new pedagogy utilizing networks which is very consistent with The Second Machine Age. I believe it can be generalized to show a type of market opportunity:

“Our traversals across networks are our pathways to learning, and as the network expands, so does our learning. In connectivist approaches to learning, we connect our individual knowledges together to create new understandings. We share our experiences, and create new (social) knowledge as a result. We must center on the ability of individuals to navigate this space and make connections on their own, discovering how their unique knowledge and talents can be contextualized to solve new problems.”

Scientists have long believed in the power of networks to foster research and learning. The Royal Society in England was founded in 1660 to support understanding in science. My point here is not to foster the further development of learned societies but rather to show that scientists have viewed the world in a similar way to The Second Machine Age since 1660. The opportunities suggested to me by this expanded networking include increased outsourcing (what some call the platform effect), more hands-on learning between masters and apprentices and more tools for curating information.

In “Query Strategies for Priced Information”, Fagin, Venkatesan, Guruswami and Kleinberg describe a networked economy:

“We envision software agents that autonomously purchase information from various sources, and use the information to support decisions.”

In this scenario we have agents creating their own networks to source and purchase information. This comes very close to having artificial intelligence ad hoc combine computation with information storage. This might be a repeat of Turing’s insight that lead to the first computer and again create a large market opportunity.

One of the most interesting and difficult to understand parts of complex systems is that all complex systems are emergent. Before proceeding further, we should heed Melanie Mitchell’s warning that what do not understand about a complex system is not necessarily a characteristic of emergence. Emergent properties are characteristics or behaviors that cannot be explained by the leaderless system of independent variables. Some scholars explain consciousness as an emergent property. Others explain sexual desire as an emergent property. Facebook perhaps demonstrates an interesting emergent property of some networks. A report by the international audit and consulting firm Deloitte[37] estimates that the economic impact of Facebook on a global basis in 2014 was $227 billion, of which $29 billion was attributable to “platform effect” — third party apps and services that attached to the Facebook infrastructure. I believe that “platform effect” is an emergent quality that enriches both the original network and the third party extension. Another example of an emergent characteristic might be many authors joining a network of book readers where they can interact directly with the readers. Readmill, acquired by Dropbox in 2014, offered this feature. (Perhaps a greater focus on the emergent characteristic would have enabled Readmill to survive as a standalone company.) Perhaps the advertising revenue model of Google search is another example of a successful network with an identifiable emergent characteristic. Building networks to foster symbiotic emergent characteristics such as platform effect may be a large market opportunity. The platform effect at both Google and Facebook was an after thought, as would be expected based on complexity theory, but in fact a key to success in both cases. At Google it provided the means to monetize search and at Facebook it accelerated the network effect for Facebook (and probably drew the world’s attention to social media). A business based on a network with a weak emergent characteristic is by definition a failure as a network. Fostering emergent qualities in networks should be a big opportunity given that the number of potential networks will only increase with the proliferation of digital technology. Such opportunities could involve network design or perhaps services to encourage the emergent characteristic. Note: looking at “brand” as an emergent property of networks might be an interesting area of academic research. An institution like Harvard, for example, has much greater brand value because of its alumni network.

Another interesting opportunity related to the network effect is Bitcoin and the underlying Blockchain infrastructure. Originally I was totally enamored of the idea that Bitcoin would replace government as the monetary authority by eliminating the need for government–issued currency. (The notion of eliminating government control of monetary affairs is almost irresistible.) With more thought on the subject I think Blockchain is a potentially bigger opportunity. Blockchain allows the members of a network to collectively authenticate data, replacing the role of a central authority. The MIT Media Lab Enigma project, according to Fast Company, uses the Blockchain technology to “enable a marketplace where users can sell the rights to use encrypted data in bulk computations and statistics without giving raw access to the underlying data itself”. For example, personal health record data could be shared without revealing individual identities. Effectively the Blockchain technology creates trust, verifies the data and reduces the cost to a network of processing information. With the increased size, versatility and resources of current networks and with support from Blockchain technology, perhaps the biggest opportunity should be to use the newfound power of networks to reduce or eliminate many of government’s current functions. Of course with just the connectivity of the Internet, self-organizing networks of individuals alone may be sufficient to bring about these types of changes.

The last opportunity that may emerge is in services to networks. For example, a university wants to start offering educational tours in Africa to alumni as a means to add value to the alumni network (and hopefully increase donations). The university will need a wide range of services to execute a strategy outside classroom education. Another example comes from Blackrock, the asset management behemoth. Any company that Blackrock invests in can purchase travel through Blackrock’s travel supplier[s] and take advantage of the volume discounts. An interesting example comes from my hometown Miami Marlins. They have created a network to share business between their corporate box holders. Both Blackrock and the Marlins need services for the network to exploit this additional opportunity to create value. As network becomes a better understood method to add value to an existing business model, the need for network services should increase.

9— Organizations

“A central element in the model presented is the concept of techno-economic paradigm as the set of generic technologies and organizational principles that emerge with each technological revolution and guide its diffusion, through being adopted as shared best-practice common sense by all the economic agents.” Carlota Perez [my emphasis]

In the last chapter we focused on the network and the benefits of connectivity. We looked at the direct benefits of bringing together the nodes, perhaps an “infrastructure” approach, and the emergent properties of the complex systems that were created. In this chapter we examine the nodes in the network [complex system] in more detail and focus particularly on “organizations” as network nodes.

“The structure of a complex system is determined by the interaction of the components (or nodes)”.[38] Given the non-deterministic nature of these complex systems, these systems can be volatile and constantly changing (although for certain periods they can appear stable and deterministic). If we think about what creates the volatility in the complex system of a network, it must be attributable to either the links or the nodes [organizations]. However, links are not agents and therefore are not creating the volatility. It is the nodes or the organizations in social networks that create the volatility. Once the environment affects the status quo (short as it may be), a node must make a choice. The choice is always to gather more information, which can be from another node in a network. John Horton Conway’s writings on games show that “the move is the road [link], the option is the destination [node]”.[39] What Conway makes clear is that the choice of the node (or any means to gather information) is an option or probabilistic. When we see probabilities as assumptions, we find that Conway provide a logic consistent with Peter Drucker’s ‘Theory of the Business”.[40] To paraphrase Drucker, a business is defined by its key assumptions and the purpose of a business is to verify and update its assumptions. Nodes or organizations choosing where to go for new information are following Drucker’s logic (although Conway’s math might be a more compelling explanation). Writing on the uncertainty of information, JR Galbraith states that “the greater the uncertainty of the task, the greater the amount of information that must be processed between decision makers during the execution of the task to get a given level of performance”. According to Galbraith, the ability to process larger amounts of information is why new organizations emerge during periods of monumental social, economic or technological change [environmental change]. For example, fiefdoms were the common form of organization from the 9th to 15th century, guilds that held the exclusive rights to produce a certain good then replaced them, followed shortly thereafter by stock companies that arose to support marine merchant capitalism in 16th century Europe. Publicly-traded companies then financed the capital intensive industrial revolution in the U.S. and Europe beginning in the 18th century and were followed by two other more recent forms of organization — conglomerates and then global corporations. Each new form of organization emerges because it is better able to process the new, larger amount of information than the precedent form of organization. This better processing reduces uncertainty, lowers transaction costs and attracts more capital to finance the social and economic change. As capital flows respond to the reduced uncertainty [and risk], the market attracts the necessary capital to finance the “innovation” and eventually reaches a risk-return equilibrium [for capital].

The notion of processing information in a social network deserves some explanation. An individual or organization — as agent — acquires information, computationally evaluates and archives it, produces new information [what in social systems we would label as knowledge] and communicates or shares the information. In summary, organizations evaluate probabilities triggered by environmental changes in order to secure and process information…to survive. Except for the concept of knowledge, organizations behave much the same as natural networks of ducks or ants.

Another feature of organizations in human networks is that the organization is more “valuable” the more networks it spawns. The spawning of networks increases the information processed, increases the likelihood of redundancy in the network and enables the organization to better adjust to environmental change. Take a university as an example. A university has individual networks of students, alumni, faculty, research institutes, colleges and funding partners (NSF, CDC, NIH, etc.). Some universities have additional networks of athletic boosters, community partners, parents and specialized networks based on expertise (e.g. autistic children). I tend to think of each of these networks as a two-dimensional plane, one stacked on top of the next.

Note: A bit farfetched, but it could be that emergent properties in complex systems are just planes [networks] so far away from the original organization that the relationship is no longer obvious. For example, football boosters at a university such as Michigan or Miami or Texas take on behaviors that easily could appear separate from the university itself. Organizations do well when their strategic focus supports their largest networks, where the greatest amount of information is being processed (even if it is the football boosters).

A fascinating article in the MIT Technology Review, “The Extraordinary Link Between Deep Neural Networks and the Nature of the Universe”[41], states:

“In the last couple of years, deep learning techniques have transformed the world of artificial intelligence…Deep neural networks are now better than humans at tasks such as face recognition and object recognition…But there is a problem. There is no mathematical reason why networks arranged in layers [neural networks] should be so good at these challenges…In the language of mathematics, neural networks work by approximating complex mathematical functions with simpler ones. When it comes to classifying images of cats and dogs, the neural network must implement a function that takes as an input a million grayscale pixels and outputs the probability distribution of what it might represent. The problem is that there are orders of magnitude more mathematical functions than possible networks to approximate them. And yet deep neural networks some how get the right answer. The answer is that the universe is governed by a tiny subset of all possible functions. In other words, when the laws of physics are written down mathematically, they can all be described by functions that have a remarkable set of simple properties.”

If the universe can be modeled through neural networks, which are an example of what I called “stacked networks”, then perhaps reality is a series of networks. Donald Hoffman’s argument that reality is determined by “conscious agents” and that these agents can create yet other conscious agents also appears to suggest that networks might be an ontological fundamental and that stacking these networks might be a powerful model to understand reality. My point here is not to argue epistemology but merely to show that stacked networks might be a fundamental concept around which large market opportunities are to be found, similar to explaining Google’s success in terms of auto catalysis.

In 1948 Claude Shannon published his groundbreaking paper on information theory, “A Mathematical Theory of Communication”, which showed that all information could be communicated through waves in a binary format of 0 and 1 (that became known as bits). Building on this idea, Stourton Steen, in his book “Mathematical Logic” states “Mathematics is the art of making vague intuitive ideas precise and then studying the result and inventing a method whereby our thoughts can be either communicated to others or stored for our own memory. “ [42]

Combining the two ideas, we gain an insight into why computers have become so important. Shannon’s insight changed the way information was stored (amongst many other benefits) into a mathematical means with an unheard of level of precision [Steen]. This breakthrough enabled organizations to capture and organize larger and larger amounts of information at lower transaction costs. The computer also provided processing of information and eventually connectivity. In summary, the popularity of computers is attributable to the perfect match with a node’s [or organization] requirements to handle all facets of information processing, as determined by complexity theory. Just like Google search, the computer perfectly followed the theory of complexity. This is the second example where a large business was created that closely followed the concepts of complexity theory. Perhaps new business concepts should be evaluated to determine whether they follow or conflict with complexity theory. (More on this in the Conclusion.) Union Square Ventures, a highly regarded venture capital firm in New York City, has an investment thesis where they only invest in business concepts based on networks or marketplaces. Marketplaces are just networks that have an additional characteristic — the network has two distinct groups of similar participants, i.e. suppliers and customers.

The continued miniaturization of computers, their increased processing power and the declining cost of computing all contributed to dramatic changes in the nature of organizations:

· Smaller organizations could successfully contribute or solve more complex problems at a lower cost

· As more organizations became “viable”, they attracted more interaction with other organizations [connectivity] and consequently became part of larger networks

· As part of larger networks, organizations were better able to tackle more complicated problems despite a more volatile environment

· As the results from the expanded network showed the value of cooperation, trust between organizations increased, verification costs declined and autonomy increased.

These changes provide much insight into what I see as large future market opportunities.

As we look at future opportunities for new businesses concepts related to [networked] organizations, the noted former IBM executive Irving Wladawsky-Berger sets the stage perfectly in a Pieria article:

“Significantly lower transaction and coordination costs are leading to organizational decomposition, with companies fragmenting to smaller and smaller entities, even down to individual providers. At the same time, economies of scale and network effects are leading to organizational consolidation and a winner-take-all world where only the largest survive.”

As I see it, decomposition or decoupling represents the greatest number of new, large opportunities. Properly done, this divestiture will focus on the lowest value-added activities of corporations. Corporations will most likely outsource everything except that which relates to intellectual property and new product development, customer experience (including sales), high level executive recruiting and some M&A/business development related activities. I am undecided about what happens to finance activities such as cash flow planning, which are critical to survival but generally very low value-added. (Maybe Goldman Sachs should offer an AI-based service to manage day-to-day finance functions in addition to raising such required capital.) Such widespread divestiture of activities will create large opportunities to become the F500 service provider or the software provider for low value-added activities, such as human resource administration. Advances in AI could perhaps eliminate the need to outsource to a company and perhaps allow companies to use a SAAS provider for a software only solution to achieve the same end. Such AI-based software could takeover logistics, manufacturing, advertising placement and a host of other functions now routinely done in-house. Perhaps the really large opportunity is in combining SAAS and AI and the software tools to make that happen.

Divestiture of functions will also happen at the individual level, where not only will organizations outsource functions or departments but also individuals. As Ronald Coase made clear, organizations operate in ways to reduce transaction costs. Modern technology theoretically permits the company financial analyst to be located in New York or Mumbai. The challenge, of course, is to find that analyst in Mumbai who is properly educated with documented credentials and the necessary experience. Only then can one benefit from the expected cost savings. Much more sophisticated marketplaces will develop that enable organizations to source reputable individuals around the world with credentials and experience. Such marketplaces may specialize based on skills or professions in order to better match user requirements.

Another change that will present big opportunities is the effect of artificial intelligence on organizations. In a very interesting article in BCG Perspectives, “Travel Innovated: Who Will Own the Customer? “[43], BCG paints a picture of the future where a customer for an overseas vacation trip merely tells the AI agent where and when they want to go. This agent, based on your web history, determines all of the “affordable” flights, hotels and activities for the vacation. It may be difficult to determine when such a day will arrive, but:

· Such an AI-based agent will have the only relationship with the client and the algorithm will have the “relationship” with the travel service providers

· Advertising, except for destination, will be designed to create the individual web history for the Ai-based researchers (although the ad money might be better spent to just be the first preference for airline or hotel in the AI algorithm)

· The concept of economic value may change where guest review management may become more important than hotel management

The BCG article presents an interesting scenario in which the whole nature of the customer value proposition changes. Being the infrastructure provider for such a change may be a Google type market opportunity.

10— Domains

“…. it is astounding how few of us bother to invest enough mental energy to learn the rules of even one of these domains, and live instead exclusively within the constraints of biological existence.” [44] Mihaly Csikszentmihalyi

This is the last of the four chapters where we look to complexity science to provide insight into large market opportunities. Identifying “paradigm shifts” or “domains” and understanding their characteristics is the key to seeing the opportunities. This task is particularly relevant given that we are at the point where many say we are experiencing a paradigm shift, but interestingly there is much difference of opinion on which technology is sufficiently “revolutionary” or disruptive to claim the title. The popular candidates are:

1. Artificial intelligence (AI)

2. Robotics

3. Virtual and Augmented Reality (VR/AR)

4. Internet of Things (IOT)

5. Nanotechnology

A recent write-in candidate is the blockchain technology, which I think would be a wonderful winner for a paradigm shift. I believe that the blockchain technology could fuel the creation of more self-organizing network solutions to problems by individuals.

Joseph Schumpeter’s phrase “creative destruction” is perhaps the most well-known and often quoted expression in economics. The recent fanaticism around innovation has only fueled the popularity. A phrase emerging as a rival in the usage polls to Schumpeter’s term is “paradigm shift”. What many may not appreciate is that the two popular phrases are directly related. Schumpeter’s “clusters” of radical innovation are, in fact, what Thomas Kuhn called “paradigm shifts” in The Structure of Scientific Revolutions (1962). Recently Carlotta Perez at the London School of Economics has done much work to show in greater detail the “technological revolution” of each paradigm shift. In order to make the link between paradigm shifts and complexity theory more apparent, I am going to refer to paradigm shifts using W. Brian Arthur’s terminology of “domains”.

Arthur develops his concept of the domain in his 2009 book, The Nature of Technology, by contrasting a technology and a domain. A technology is invented to do a job in order to achieve a purpose. A technology is a product or a purpose and can change to incorporate incremental improvement. A domain is a framework that emerges that includes technology, but also includes related methods, practices, a particular vocabulary and its own symbolic system. As a domain adds components, creates new combinations of components and re-uses the components in new ways, the reach or penetration of the domain increases significantly and becomes a paradigm shift. As the tipping point is reached for a domain to become a paradigm shift, these conceptual frameworks adopt rules that extend the influence of the framework into the social, economic, political and scientific fabric of society. The far-reaching influence of the framework and its application across a wide range of industries and problems is the key characteristic that leads to a “technology revolution” or paradigm shift.

Paradigm shifts have been tracked and analyzed since the late 18th century when the first technology revolution occurred. At that time industrialization began to replace the hand made, craft approach to productions of goods. Since that time analysts have disagreed on the number of new domains that have reached the status of paradigm shifts. Carlota Perez believes there have been four since the late 18th century, not counting the one emerging now. Perez lists the paradigm shifts as shown below:

1. Industrial Revolution 1771

2. Age of Steam 1829

3. Electricity, Steel 1875

4. Age of the Automobile 1908

5. Age of Information 1971

When we look at these paradigm shifts the first thing one notices is that 1–4 all involved new forms of energy and each subsequent type of energy is more sophisticated than the prior one. The Industrial Revolution involved the switch from human-powered work to mechanical forms of work. Then steam was substituted for the mechanical devices and then electricity and the gas combustion engine came along. Each alternative energy emerged as the previous form of energy reached a point of costly inefficiency at the new scale of economic development. If the next paradigm shift is artificial intelligence, virtual reality, IOT or nanotechnology, then paradigm shifts 5 and 6 or even 7 may be based on new applications of information. (Robotics probably qualifies as either energy or information.) What we now see is that every paradigm shift finds its beginnings in either energy or information, the two fundamental concepts in complexity science. This shift from energy to information driven paradigms matches the transition from economic value being largely created from land and labor to information. Whether such a transition began with electricity or digital information requires further study. It is not so obvious when one considers that electricity made possible radio, television and video — all major sources of new information at the time.

James Haywood Rolling, Jr. states in Swarm Intelligence “when new information cannot be integrated into the existing paradigm or when problems persist which cannot be resolved, a new paradigm is likely to arise to replace it.” So perhaps all paradigms can be explained in terms of information and all paradigms involve the integration of massive amounts of information in a new way. A technology by contrast involves the integration of a smaller amount of information, in part because of the comparatively smaller scale of adoption.

Another obvious parallel to complexity theory is that each paradigm shift follows a classic pattern from complexity theory — explore and exploit. A technology would be the exploration of the scientific principle and some early adoption. The penetration of the technology into multiple industries and applications demonstrates the concept of exploitation and the necessary scale to be a domain or paradigm shift. To quote Melanie Mitchell again, “the system both explores to obtain information and exploits that information to successfully adapt”. The exploitation manifests itself, in part, in the development of networks to share the information about the technology and its application. As we pointed out in Chapter VIII, all paradigm shifts are also characterized by a new form of networks.

In order to understand the future market opportunities from a paradigm shift, we must first identify the paradigm shift and then identify the particular opportunities. In order to identify the paradigm shift, we must be looking for a comparatively simple, low cost technology (or the potential to reach that point) that has application across a wide range of industries. Also, the application of the technology has to be where the value is created for the user. For example, producing information is relatively low value compared to processing information, which is relatively low value compared to analyzing the information and issuing warnings. The later would be a technology more likely to become a paradigm shift.

Of the five or six candidates (including blockchain) for the next paradigm shift, I believe that AI is the likely winner. I see robotics and IOT as being subsumed under AI (a further proof that AI is the paradigm shift) and either VR or nanotechnology as the next paradigm shift after AI. I think VR will eventually change all forms of education and training, particularly after AI develops the skills to define and develop individual, customized learning programs. Nanotechnology will replace much of what we now call modern medicine, as soon as we become comfortable inserting monitoring devices inside our bodies through a permanently embedded input-output opening.

I believe AI is the next paradigm shift for the following reasons:

o AI creates the value for the user by the sophisticated processing and analysis of the information (data)

o AI is low cost, particularly as cloud-based software and algorithm libraries are made widely available; the widely available, free IBM Watson platform also fosters low cost usage for small scale projects and experiments

o AI is already widely used in many different applications and industries

o AI is well on its way to changing how retailers and service providers sell to the consumer, focusing product offerings on the basis of more, sophisticated data about the individual. This might very well change the consumer relationship to more of a relationship with their preferred AI shopping bot instead of the retailer or the service provider, thereby redefining the consumer shopping network.

The new business opportunities for AI are plentiful. Extending AI to new industries and applications is one example. Developing specialized algorithms for a particular problem is another large opportunity, such as for education to better document learning styles. As was the case in the paradigm shift to digital information, there will be plentiful opportunities in AI hardware including CPUs and storage devices (for the large amounts of data required). Outsourcing the AI software stack will also be a large opportunity as companies look to reduce fixed and variable costs for AI. Perhaps we will even see a new operating system or a Linux fork better suited to the demands of AI. (One will recall that the MAC OS started from the Linux kernel.)

Another type of opportunity comes from Arthur’s writings in The Nature of Technology. To paraphrase Arthur, there is always opportunity at the boundaries of a domain. For example, to print on paper a computer generated text, the printing is the opportunity where the information leaves the computer (the boundary). Using a gantry to take a container off a ship and put it on a truck is another example. (Containerized shipping is the technology.) The ATM machine is an example at the entry boundary of the technology of online banking. To now look at AI, we see that information capture in new environments is an entry boundary opportunity, such as nanotechnology sensors in the human body or the method to retrieve such information. An exit boundary opportunity might be the vacation travel package designed by multiple uses of AI to identify the consumer vacation preferences.

A last point about paradigm shifts and large market opportunities must be noted. In the Sequoia investor pitch template that I prefer for my students to use, an important question addressed is “why now”. In other words, is the timing right for the opportunity to be realized. Paradigm shifts can take a long time to develop. For example, the first computing devices were used during WWII to crack German codes, but popular commercial applications did not appear until the 1960s and consumer usage did not take off till the 1980s. AI is itself another example. AI started in the 1960s shortly after the first commercial computers were developed. However, it was not until this century before computer scientists realized that it was not the algorithms but the amount of data available for “learning” that was holding back advances in the adoption of AI. All digital technology shows us that there are “signature” uses of the technology which drive their popularity. Excel is responsible for the popularity of personal computers and financial industry processing was probably responsible for computer mainframes. AI has probably not yet found its signature application, but it might be something related to determining product offerings to consumers. Finding this signature opportunity for AI might birth another Microsoft or Google.

So far in this chapter we have focused on the technology inherent in a paradigm shift. Taking a lesson from our own writing, perhaps now it is valuable to change the assumption and examine paradigm shifts from the social or societal perspective. Returning to the writings of Thomas Kuhn, Kuhn surmised that paradigms develop because of their success in representing the prevailing understandings, shared beliefs, and research solutions of a community, what might be called an intellectual framework.[45] From Kuhn’s book in 1962, a more contemporary definition of paradigm has emerged, which James Haywood Rolling Jr. summarizes simply as “a body of beliefs and values, laws, and practices which govern a community.” Whether you characterize a paradigm as an intellectual framework or a community consensus, paradigms are basic rules that govern all social behavior and scientific reasoning.

If we look at historical paradigm shifts from the perspective of society, we might point to the introduction of free public education in the U.S. in the late 18th century. This change provided a sufficiently educated workforce to operate the factory machinery to support the industrial revolution, produced the students that later filled a nationwide system of private and public universities, created a huge textbook industry to educate students and created huge bureaucracies with their own needs (to be filled) to operate schools, universities and school districts.

Female contraception might be an example of a paradigm shift brought about by technology but I prefer to think of it from the perspective of a societal change. Several possible explanations might be cited for the emergence of more forms of birth control. Women just wanted more choices in birth control, women could change their sexual behavior (maybe, maybe not) and women could better manage when they wanted to have children are all possible explanations. Regardless, this “freedom” in large part explains the very significant move of women into higher paying traditionally male jobs and professions, Many opportunities were re-scaled to a significantly larger size by more women in the professional workforce, including markets for women’s business clothing, day care and nannies, pre-school programs such as nursery schools and location-based services to track children. The movement of these better educated women away from traditional teaching jobs probably lead to a lower quality teaching corp and a resultant lowering of teaching quality. Concerns about student performance birthed service firms to help children at every stage of the education process. At the extreme, all the pressure on children to perform better in school probably lead to the popularity of psychotropic drugs to treat the children’s more frequent psychological problems.

The increased income of working women also probably lead to the increased number of single woman households. This redefinition of the family lead to a rethinking of “relationships” and more public discussion of sexuality. These trends lead to more transparency and acceptance of alternative relationships and enabled marketers to better segment their markets and target new products to these “new” segments.

Another significant change in the “social consensus” as an indicator of potential business opportunities is the possible redefinition of the role of government. The increasing access to information and real time data will put pressure on government to be more transparent. With transparency, opportunities for the private sector to substitute for government will be more obvious and frequent. Private sector substitution will reduce the role of government. For example, charter schools to replace all public schools in a state or city would be an example. (Think of the capital windfall to governments from selling their school property…or better yet converting it to parks and forgoing the capital infusion).

These examples of opportunities from changes in social consensus point out that opportunities related to a paradigm shift can be found by examining the technology or the social consequences. Regardless of how one approaches the paradigm shift, following these trends, or being a partner or supplier to such a market leader, will always provide large market opportunities.

11— Conclusions

When I write I generally research for a year and then write for a year and finish the work.. However, in this case I let the material sit for almost a year to see if any other large market opportunities emerged that repeated throughout history. Chapter X — Domains — came out of the gestation period. During this time I also discovered thanks to William N. Goetzmann’s Money Changes Everything: How Finance made Civilization Possible that money is best thought of as a technology, which required Chapter I — Habits — to be re-written. Courtesy of Brian Arthur I also learned that science is auto-catalytic — science produces technology which further advances the study of science that produces technology…

This concept of auto-catalysis, which describes both the Google search algorithm and the development of science, illustrates a point that is true in entrepreneurship and all other disciplines. There are frameworks or mental models that can be used to learn a discipline, analyze a problem or develop a solution regardless of the topic or specialty. This article has presented at least ten frameworks with which to understand large market opportunities that seem to always repeat throughout human history. While I have tended to approach each chapter and framework with a positive attitude — helping the reader to find their next opportunity — the opposite is also true. Any new opportunity should be examined to see which of the ten frameworks explain the nature of the opportunity or provide additional insight into the opportunity. Failing to find an application of one of the ten frameworks might suggest that either the opportunity is not yet sufficiently well understood or that there may not be an opportunity.

“Learn much, teach more” Heraclitean Fire by Erwin Chargaff

I would welcome the opportunity to publish this article as a book.

This work by Robert H. Hacker is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.



[1] Mitchell, M., Complexity: A Guided Tour (Oxford University Press, 2009)

[2] Durant, Will; Durant, Ariel, The Lessons of History Simon & Schuster (2012)

[3] Goetzmann, W.N., Money Changes Everything: How Finance Made Civilization Possible (Princeton University Press, 2016)

[4] Cahill, T., Heretics and Heroes (Anchor Books 2003)

[5] A fourth form of asymmetry, complementarities, involves the perception of information inside and outside an organization. Henry Chesbrough popularized the concept of “open innovation” to address the problem of complementarities. Because of the challenges in determining the information inside any organization, Chesbrough advocated aggressive sharing of information and partnering so that it was easier for companies, for example, in the same industry to cooperate more effectively on research and commercialization of new products.


[7] Manoj Saxena, Head of IBM’s Watson Project, from an interview on Creativity Post

[8] Weinberg, S., To Explain the World: The Discovery of Modern Science (Harper Collins Publisher 2015)

[9] FarnamStreet blog provided these examples

[10] Pollard, D., PKM: A bottom-up approach to knowledge management. In Knowledge Management in Practice: Connections and Context (Information Today 2008),


[12] Arthur, W.B., The Nature of Technology: What it is and How it Evolves (Free Press 2009)


[14] Hidalgo, C., Why Information Grows (Basic Books 2015)







[21] Gardner, H.E., Creating Minds… (Basic Books 2011)


[23] Durant, W., Durant, A. The Lessons of History (Simon & Schuster 2012)

[24] Mitchell, M., Complexity A Guided Tour (Oxford University Press 2009)

[25] Simon, H. A., The Sciences of the Artificial (The MIT Press 1996)

[26] Hayek, F.A., The Use of Knowledge in Society (Library of Economics and Liberty 1945 )

[27] Simon, H.A., The Sciences of the Artificial (The MIT Press 1996)

[28] Mitchell, M., Complexity: A Guided Tour (Oxford University Press 2009)


[30] Wulf, A., The Invention of Nature: Alexander Humboldt’s New World (Alfred A. Knopf 2015)

[31] Coase, R., “The Nature of the Firm” (Economica 1937)

[32] Simon, H.A., “The Architecture Of Complexity” (Proceedings Of The American Philosophical Society 1962)

[33] Barabási, A.L., Albert, R., “Emergence of Scaling in Random Networks”(Science 1999)

[34] Mitchell, M., Complexity: A Guided Tour (Oxford University Press 2009).

[35] Galbraith, J.R., “Organization Design: An Information Processing View” (Interfaces 1974)




[39] Roberts, S., Genius at Play: The Curious Mind of John Horton Conway (Bloomsbury 2015)


[41] HTTPS://]

[42]Roberts, S., Genius At Play: The Curious Mind of John Horton Conway (Bloomsbury 2015)


[44] Csikszentmihalyi, M., Creativity: Flow and the Psychology of Discovery and Invention (Harper Perennial Modern Classics 2007)

[45] Rolling, J.H., Swarm Intelligence (Palgrave MacMillan 2013)