Investing in Artificial Intelligence ventures is not for the faint of heart. In addition to the technology being complex and rapidly evolving, tech journalism routinely portrays AI in sensationalistic or grandiose terms, adding to the overall confusion and anxiety. This paper cuts through the hype, puts AI into broader perspective, and suggests a framework that will help investors decide which AI investments to back and which to avoid.
“Knee of the Curve”
For the bulk of human history, advances in technology were few and far between as daily survival consumed our attention. Early civilization inched forward as we slowly built a foundation for future knowledge. With the Renaissance came a burst of innovation that continued into the Industrial Age. As progress quickened, greater numbers of creative people had resources for intellectual pursuit, propelling technological progress even further.
In our lifetimes, we have seen this positive feedback loop go into hyper-speed with innovation surging forward like never before. One manifestation is the enormous growth in data volumes. A decade ago, human knowledge encompassed 160 exabytes of information. Today, it has increased by three orders of magnitude, to 44 zettabytes. We have reached the knee of the curve.
Looking to the 2020s, data volumes will continue to expand as SaaS offerings and cloud applications proliferate, 5G becomes widely available, cloud computing and storage continue to improve, and devices like wearables, IoT, surveillance devices, and even household appliances record ever more user data. From our vantage point today, exponential growth of unstructured data seems certain to continue with no end in sight.
AI Versus Data Chaos
Artificial Intelligence refers to a massive human effort to invent, develop, and deploy ever-more sophisticated data analysis tools to make sense of a data-saturated world. This effort is necessary because human knowledge has increased by eight orders of magnitude since our grandparents’ time.
AI represents a recent manifestation of an ancient human drive to create tools to transcend biological limits on our understanding. When limits in our visual acuity posed insuperable barriers to understanding the solar system, Galileo developed the telescope and discovered four moons of Jupiter. The same drive that fueled Galileo now pushes today’s AI pioneers to invent and refine tools and techniques that, with careful orchestration and refinement, will extend human cognition.
The first rule of creating AI systems is: Start at the End. Define the question that you want answered. Then identify the data set that contains the answers. Draw upon experience and ingenuity to identify the combination of AI tools (from thousands of options) that will deliver the best answers. Integrate data sources and requisition the computing power necessary to run sophisticated data operations. Tune the algorithm to optimize the results. Devise a user interface that can display extracted information in a fashion that is most helpful to the user. While this is demanding work, it does not have to be mysterious or intimidating. If an investor can grasp how a house is built, or how minerals are refined, understanding AI is no more difficult. Engineers even refer to it as “putting the Lego blocks together.”
In the last 2 or 3 years, data science has generated an avalanche of new and improved Lego blocks. For example, the market for bulk data feeds has matured and become extremely competitive. Major providers of Cloud Computing (e.g., AWS, Google, Microsoft, and IBM) offer resources to run computation-heavy algorithms, putting powerful new tools in the hands of data scientists. IBM Watson’s catalog alone has more than 1,500 unique AI tools, the attributes of which are not well known or publicized. These resources serve as raw materials that data scientists draw upon to build AI systems.
Questions to Ask Before Investing in AI
The Artificial Intelligence sector is projected to reach $1.2 trillion by 2020. Enormous public and private investment is being made in AI. The Trump Administration has issued an executive order aimed at keeping America at the forefront of AI. Emerging companies with strong tech teams, a compelling vision, and a demonstrated ability to deliver custom AI systems have a chance to become incredibly successful. By asking the right questions, an investor can take advantage of tremendous opportunities presented by Artificial Intelligence while mitigating risks.
How did you choose the particular AI tools incorporated into the product and what role does the AI play in improving the user experience?
For a venture to say it “has AI” is almost like someone claiming to “have a power tool.” Is that a good thing? Depends which tool and what you want to do with it. A chainsaw is great for clearing timber but not for cleaning windows. Before getting into an AI venture, investors should understand what tool(s) are being used, how those tools work, and what data problem the tool is being used to solve. Does the AI tool solve a problem that is central to the overall task? Does the AI accomplish something important that conventional technology could not? Is the technology performing a task that cannot be done with human cognition?
If the AI is being used for a problem that is not central to the product’s overall task, or if the problem could be solved without AI technology, an investor should be wary. If the AI doesn’t help the user in a big way, the AI venture can easily end up undercapitalized in a crowded market competing on price. This grim scenario can be avoided by investing in technologies that solve pressing problems that only AI can solve. The more fundamental the problem the better (as there will be more adjacent use cases).
Assuming your business plan comes to full fruition, how will the world be different?
No one starts an AI company without a compelling reason. Under the best of circumstances, the AI journey is long, hard, humbling, and requires patience and determination. As an investor, understanding why founders chose to embark on a challenge as difficult as AI is a key to understanding what you are investing in. If the founders are motivated only by the hope of getting funded and making money, they will give up when they figure out there are easier ways to make money. By asking the founders how the world will be different when their technology comes to fruition, an investor is going to learn what really motivates the founders and how deeply the founders have thought into the future.
Tell me about your engineering team and how they came to be involved in this project?
Every AI company claims to have game-changing technology. Unfortunately, most investors don’t have the background or frame of reference needed to evaluate these claims. A handful of AI ventures are doing truly groundbreaking work at the frontiers of technology. Most are not. How can an investor with limited technology knowledge tell the difference? An excellent idea is to speak directly with the tech team.
If the founding team does not include an experienced data scientist, you can be confident that the AI is not particularly innovative. On the other hand, if the venture has a team of experienced engineers who were drawn to the project because of the power of the technology and vision, that is a strong indication that the founders are doing something special. Engineers tend to be straight shooters. Ask them what they like about the tech and the company. They will tell you.
Why did you choose to build this particular product?
When starting an AI company, the founders can choose to build anything using any AI tool, any data sets, and customized for any use case. An investor should be interested to know why — out of the universe of possibilities — the founders chose to build this specific product. The founders may have personal experience in a market or industry that will enable them to design for and sell to that market. The founders may have developed an amazing technology and then sought out a use case that seemed well-suited to the tech. Or perhaps the founders had an existing non-AI product they wanted to spice up with something they could market as AI. Whatever the answer, an AI venture’s choice of use case is the most important decision a startup makes. As an investor, the answer to this question will give insight into what the founders consider their core advantages and how the founders think strategically.
What will stop your competitors from replicating your product?
An investor should understand whether the underlying tech is open-source, third-party proprietary, or developed in-house. Most likely it will be a combination. Ask whether any particular segment of the workflow requires special knowledge or skill to create. Is there any component that is obscure or hard to source? By understanding what aspects of the product were most difficult to develop, an investor can gain a good sense of whether competitors will replicate the technology.
Of course, a patent is the gold standard. Relatively few early stage ventures patent their technology. If an AI venture has filed a patent, that is a positive sign in three respects. First, it demonstrates that the company believes in the technology strongly enough to spend scarce time and resources protecting it. Second, it shows that the founders are behaving as if they plan on being long-term successful. Finally, it demonstrates that the founders are sufficiently “on the ball” to engage IP counsel and articulate their concept in a rigorous and methodical way.
Who is on your board of directors and advisors and what criteria did you use to select them?
Investors can learn a lot about AI founders by seeing who is serving on the company’s board. As Bob Dylan said, “you gotta to serve somebody.” Company founders have the privilege of choosing the board to whom they will answer. How the founders go about recruiting a board will tell an investor a lot about the founders’ approach to business and corporate responsibility. Serving on a board is a major commitment, and if the founders have recruited people of substance, wisdom, and experience to serve on their board, that is an excellent sign. On the other hand, if the founders have chosen their board carelessly, it is a red flag that the founders are not ready to handle other people’s money. Given that AI founders are confident and headstrong people (almost by definition), an investor should know whether the board is capable of acting independently.
Describe your vision of what the world will look like in five years and explain how that vision influenced your design decisions?
An AI entrepreneur must be a futurist. We know technology is advancing at an astonishing and accelerating rate. This is having an enormous impact on our culture and our relationship with machines. While nobody can know the future, an investor should assess whether the founders are capable of thinking and strategizing into the future. This requires a different set of mental processes than we’re accustomed to using. If the founders have a well-reasoned vision of the future, and if that vision is reflected in the design of the product, that is a positive sign.
If the business strategy calls for industry disruption, does the CEO have the personal characteristics to be a disruptor?
People often talk about “disruptive technology.” In reality, technology doesn’t disrupt. People disrupt. It’s an important distinction for an investor to remember. If the business plan calls for disruption of a major industry, the founders had better be tough as nails; industry leaders do not die quietly. As an investor, ask the founders how their past experience has prepared them to take the slings and arrows that go with being a disruptor.
Artificial Intelligence is the most exciting and fastest-moving sector of the tech world. Exponential data growth means the demand for AI systems will grow, together with the sophistication of the AI tools. Investors who are equipped to ask the right questions have a unique opportunity to be part of re-making civilization for the massive data era.
John H. Snyder, is founder of Agnes Intelligence, NYC trial lawyer, technologist, entrepreneur, and designer of AI ecosystems. Snyder started his career at Proskauer Rose as a commercial litigator. In 2010, Snyder founded JHS PLLC, a boutique litigation firm known as the “AI Law Firm,” where Snyder built a reputation for aggressive advocacy and willingness to take cases to trial. Snyder has been honored by his peers as a New York “SuperLawyer” since 2013 and has been named a “Top 100 Litigator” in New York State since 2017. Snyder has advised many tech entrepreneurs, engineers, and startups concerning a wide variety of business and personal issues that arise in technology ventures. A graduate of Brown University (Phi Beta Kappa) and Harvard Law School, Snyder was born and raised in eastern Washington state; in his youth, Snyder performed in a circus as a high wire walker and trapeze artist.