Digital Transformation is on top of the investment agenda for most businesses across the globe. That’s mainly because consumers have dramatically changed their behaviour when it comes to forming opinions about anything - including who to vote for and what product or service to buy. We are nowadays spending an inordinate amount of our waking hours interacting with digital distribution channels, mostly via our smartphones. The number of mobile phone users in the world is expected to pass the five billion mark by 2019. That’s 65% of the world’s population! Having access to the internet while on the move, 24/7, coupled with considerable computing power and memory to run complex applications and leveraging a host of miniaturised gizmos such as camera, GPS, sensors, etc. are reshaping consumer behaviour and expectations. Brand loyalty is no longer a given, and the business playing field is being constantly flattened with every new “something-tech” startup launching an app.
Given this profound socio-economic disruption corporate behemoths look into “digital transformation” as the magical elixir that will transform them into lean, mean, innovation machines; to remake them, in other words, into the youthful startups they once were. “Customer experience” have become magic words in the age of digital. Data are they key to understanding and influencing customer experience, and therefore behaviour — and Artificial Intelligence the technology that uses data to deliver the influencing. But Data in combination with AI have serious by-effects: they automate processes and work, and consequently business organisation. That’s why Digital Transformation is much more than a program of software development and deployment. It is about business reinvention in a continuously changing market environment. So how must digital leaders strategise in order to make the most of their investments? How should they optimally combine systems, people and organisational redesign for agility, innovation and profitability? And what role can “decentralised AI marketplaces” play in supporting digital transformation?
From pipes to platforms
As consumers access services and products using digital distribution channels the “market opportunity” window is getting narrower and narrower. This means that consumer tastes change quickly. To achieve customer loyalty brands need to continuously innovate around their value offering. But “innovation” does not happen in a vacuum. It is a complex process that requires a chemistry of talent, systems, rewards and leadership to get it right -and to continue getting it right every time. Alas, traditional, hierarchical, silo organisational structures are not fit for innovation. They were designed and built for a culture of command and control; to stifle innovation and enforce compliance to process. Hierarchical silos are very effective when your company is like a pipe and success depends on standardisation, efficiency and compliance. But this industrial-era model of producing value is unfit for the digital era. Human imagination needs to be empowered and encouraged to think “outside the box”, experiment, allowed to fail. This happens only when talented and diverse teams are given the opportunity to collaborate unhindered. You can spend billions of dollars in rolling out shiny new apps, IT systems and data lakes, but unless you redesign your organisational structure and change your culture, you are more than likely to end up in the dustbin of business history. As Geoffrey Parker, Marshall Van Alstyne and Sangeet Choudary argued in their book “Platform Revolution”, in order to innovate and compete in the digital economy business need to change from “pipes” to “platforms”. But what does “platform” mean?
A simple definition is to think of a platform as a marketplace that facilitates exchanges between various independent groups. So a company can be a platform that matches car drivers with passengers, like Uber. But a company can also be a platform of itself: a marketplace where different groups of employees can collaborate between them as well as with external talent. Indeed you can think of businesses transforming into “talent platforms”, like Topcoder or Upwork. Someone in the company has an idea; she describes the idea on an internal talent platform; the talent platform matches her idea with the skills, aspirations, profile, rate and availability of talented employees or contractors; and a winning team is formed to deliver on the project. By transforming from pipes to platforms employees become more and more like “gig workers”. Increasingly, we will see more and more “full time employees” to become gig works in all but name. Which calls for another type of business leadership style: one that looks more like an orchestrator of work rather than a manager. Successful business leaders will be the ones who lead like coaches, while having a deep understanding of how AI and data are transforming work.
Automating work with AI
Work automation is one of the most important debates of our times. Several economists predict the near obliteration of human work in the future. Recent advances in Artificial Intelligence seem to fuel the direst predictions. Computers can recognise language, translate between languages, understand the content of images, navigate in previously uncharted terrain, write, paint, sing. As Alpha Zero demonstrated computers can also learn, all by themselves, how to strategise. But do all these advances really lead to the elimination of human work?
A recent report from PwC argues that AI would create slightly more jobs (7.2M) than it displaced (7M) by boosting economic growth. The firm estimated about 20% of jobs would be automated over the next 20 years and no sector would be unaffected. The hardest hit industries in terms of job losses are expected to be manufacturing (+22%), transport and storage (-22%) and public administration (-18%). However, AI will lead to an increase in the number of jobs in health (+22%), scientific and technical services (+16%) and education (+6%). If one is to look across industries what decides the net effect of AI on jobs is how susceptible are core tasks to automation. If a job has many tasks that are well-defined and repetitive then this job is very likely to be taken over by a machine; and the opposite is also true; variability and creativity resist automation.
This realization is not particularly new. Over the past two hundred years we have seen technology continuously automating work tasks. The reason why we have not seen the end of human work till now is because technology is a multiplier not an obliterator; it makes us more productive. Artificial Intelligence should be viewed as a cognitive multiplier. The fact that Alpha Zero can strategise does not mean that human strategists are no longer needed and that we should relinquish the keys of our economies and businesses to the machines. It means that we can now use this technology to significantly advance our civilisation and augment our potential as human beings. But how can we do that in practice?
For a business the answer is job redesign. AI can automate work tasks. To augment human work one must begin with identifying which work tasks can and should be given to intelligent machines, which tasks can be done in collaboration with intelligent machines, and which tasks shall remain human only. For example, a lot of administrative tasks are fully automatable. Wouldn’t it be nice if I had an AI optimise my calendar and meetings, do all my bookings and generally help my time management? Work that involves gaining insights from data analysis can be significantly augmented using Machine Learning — for example, running various scenarios for investments or risk. Leading people, resolving conflict, applying empathy, choosing right from wrong; these are non-automatable tasks. Businesses that adopt AI quickly will be the winners of tomorrow, simply because they will be able to augment human potential before their competitors. So the problem is not too much Artificial Intelligence, but too little.
Despite all the hype and the media, too few companies are capable of adopting AI. Why? Because there are still too many obstacles and high costs that prevent the “democratisation” of AI. The obstacles are both in the supply side of AI, as well as the demand. On the demand, too many companies find it hard to identify a problem that can be efficiently solved by AI. Problem definition is far from trivial, especially when one needs to evaluate the availability and quality of data to train models; and most companies lack the human and technical capabilities to do this. On the supply side too much power — read models, platforms, frameworks, computing power and data — is currently concentrated in the GAFA (Google, Amazon, Facebook, Apple) oligopoly. This is a major bottleneck for innovating in AI. Thankfully, the invention of blockchain, seems to confer a new way of enabling AI innovation and adoption: decentralized AI marketplaces.
Decentralised AI Marketplaces
A decentralised AI marketplace is an exchange — indeed a “platform” — powered by Distributed Ledger Technology for AI tools, data and applications where network effects and open market economics accelerate innovation and lower costs. To effect such a marketplace one needs three principal building blocks. The first block is a Data marketplace, where businesses and individuals provide their data securely in an open data exchange. Ocean Protocol is one of the most prominent innovators in this space. The second block is cloud computing so that AI innovators can train their models in the most time-cost efficient way, and for best results. DADI are building exactly that, with a vision to provide a viable and preferable alternative to the current cloud computing oligopoly. And the third building block is a marketplace for AI models, applications and tools; the result of the AI innovation; but also the development and marketing framework where innovators can create, assemble and market their ideas. Agorai, is fast becoming a key player in this space. Indeed, Agorai, Ocean Protocol and DADI have partnered in order to create and deliver a comprehensive and decentralised ecosystem for AI innovation.
A decentralised AI marketplace ecosystem is a hub for fast innovation. Here’s how it works: AI engineers and data scientists solve for real business problems, and deliver Innovative AI applications to Business Leaders that can have a significant and positive impact in the growth of their companies. They do so by working closely with them to define the problem that needs solving. They then deliver the solution by accessing, combining and evolving more elementary AI tools and models, which are supplied by the wider AI developer community. Since data are the fuel for successful AI applications, the AI innovators have access to unique data sets that are supplied by other Business Leaders. Rewards flow across the ecosystem and each participant on the supply side is rewarded according to their contribution. This is a powerful incentive for all participants. AI innovators can monetise their work, and Businesses can turn their business data into profitable assets. Everybody wins.
A decentralised AI marketplace has many advantages for digital transformation leaders. Firstly, it can provide more, more varied, and better data compared to the GAFA oligopoly and other centralised entities. Data is the fuel of the AI engine that powers the 4th Industrial Revolution. So think GAFA as the OPEC, and decentralised AI marketplaces as solar panels. The second major advantage is the quality of the AI applications; more and better AI models and applications are possible because of the incentivised participation of wide developer and data science communities. As token economics incentivise supply side participation across AI models, data and computing, costs of use are much lower — and there is no GAFA lock in.
Five tips for Digital Transformation Leaders
Decentralized marketplaces provide the means for faster and better AI adoption by every business and organization in the world, irrespective of size. They have the potential to become the indispensable tool in the Digital Transformation toolset. So here’s a list of five tips for Digital Transformation leaders:
Forge a mixed leadership team that combines IT and HR professionals. For business to take full advantage of the AI opportunity it is vital that people and systems leaders collaborate closely. If there is something profoundly different in AI is the unprecedented rate and the continuously advancing scope of automation. This means that information systems must move in sync with people, rather than systems coming first and people following.
Define the business problem. Look at the core of your business model and ask how you could benefit, or indeed become transformed, using AI. Ask: what is the business outcome that you are trying to drive? And how will you make a real difference to your customers?
Identify tasks that will be automated. Look into your business processes and identify the routine, repetitive and well-defined tasks that can and should be automated.
Dive into data. As discussed, a decentralised AI marketplace can deliver one of the key building blocks for AI adoption currently missing, namely access to more data. Training models thus becomes easier. Nevertheless, businesses need to understand what real time and what stored data will be used in running their AI systems, to ensure success in deployment.
Augment the humans. Look in ways that humans and machines can collaborate to achieve more; in what world chess champion Gary Kasparov calls “centaurs”. AIs can perform heavy duty cognitive tasks freeing the humans to deploy their innate social skills. By strategising systems and work around human augmentation the “rise of the machines” will lead to the “rise of the humans”.