How To Prepare for Automation? Or, Why We Need More “Artificial Intelligence Ecosystems” Now!

Artificial intelligence is transforming all aspects of everyday life. In a recent interview, American businessman and investor Mark Cuban — perhaps best known as one of the “shark” investors in ABC’s reality show Shark Tank — has gone as far as to say that if you don’t study artificial intelligence, you will be a “dinosaur” facing extinction within 3 years.

More and more corporate giants are following this advice. According to data analytics company, CB Insights, M&A activities involving AI companies is increasing rapidly.

M&A Activity Involving “AI” Companies — Source: CB Insights

The acquired companies are often “Silicon Valley-based” startups. But that doesn’t necessarily make it the “AI center of the world”. Even a cursory glance at the available data suggests that other regions are rapidly “catching up”. AI entrepreneurs are everywhere.

Location of Acquired “AI” Companies — Source: PitchBook

Yet, what is perhaps more surprising is that most “exits” happen in the United States, particularly Silicon Valley. Established Silicon Valley companies outperform their competitors in other regions of the world when it comes to AI-related M&A activity. The usual suspects, Google, Apple and Facebook, are leading the way.

Location of “AI” Acquirers — Source: PitchBook

What then needs to be done in order to shift the balance to other regions?

The answer is simple. We need to build more “AI ecosystems”.

If there are more ecosystems focused on AI, then everyone will be better prepared for what researchers, the business community, analysts and investors identify as the “next big thing”. Moreover, creating more ecosystems of this kind will help policymakers better address the challenges, opportunities and even the threat created by the implementation of artificial intelligence.

But, in pursuing this goal, we need to realize that we don’t need any sort of ecosystem, but the “right kind” of ecosystem. And understanding the “special sauce” that will result in the “right kind” of AI-focused ecosystem means re-visiting some of our basic ideas about what makes innovation ecosystems work.

The Question: How To Build the “Right Kind” of AI Ecosystem?

National and local governments see startup ecosystems as a necessity for preparing for the future. Putting in place the necessary infrastructure to stimulate the creation, growth and scaling of new and innovative business is now seen as an important and legitimate policy objective for all levels of government.

So, what can governments do to build an effective innovation ecosystem, AI-focused or otherwise?

Traditionally, the focus of policymakers looking to create an innovation ecosystem has been on making more risk capital available for startup and scale-up companies.

Take Europe as an example. We see a decline in venture capital activity at all the stages of a startup’s development. This tracks the worldwide trend (according to data provider and analyst PitchBook). What appears to be even more worrying is that right now the activity in first-time venture investments was the lowest in seven years.

VC Activity and First-time Venture Financing Activity in Europe — Source: PitchBook

In order to stimulate venture capital investments, governments have used several strategies.

First, there is long-standing evidence that government support has played a vital role in encouraging entrepreneurship and the launch of startup companies. For example, governments, in their efforts to establish a sustainable ecosystem, have become the main “post-financial crisis” investor in Europe’s startup scene.

Second, governments often introduce schemes that aim to activate private investments. A recent European example is the joint initiative by the European Commission and the European Investment Fund to set up Pan-European “VC Fund-of-Funds”. The investment of 25% of the total fund-size must encourage private investors, particularly, institutional investors, to invest in the next generation of innovative companies.

Third, regulatory measures can be implemented to make venture capital more accessible to investors. The proposed amendments to the European Venture Capital Fund (EuVECA) and the European Social Entrepreneurship Funds (EuSEF) are intended to give a boost to the venture capital industry in Europe. Also, investors are encouraged to make venture capital investments through fiscal incentives and tax breaks.

But even if these measures significantly increase the amount of venture capital available, entrepreneurs are not always better off.

As is the case with any industry that enjoys a boom, non-specialists will emerge looking to get a piece of the growing pie. There are a plethora of examples of “new” venture capital investors that have started to invest in innovative companies without doing their proper homework or understanding the rules of the game.

The fear of giving up equity and losing ownership and eventually control to less than stellar venture capital investors only feeds a growing skepticism among entrepreneurs about attracting venture capital or other sources of risk-capital.

What I’ve learned from informal discussions is that entrepreneurs prefer bootstrapping, perhaps supplemented with government grants or private loans from family, friends and fools (the “3 Fs”).

Although in some cases this can be an effective model, it undoubtedly exposes the founders to a much greater degree of financial risk and uncertainty. Grants can fill this gap, but drafting and submitting proposals can take a very long time. There is always a lot of competition and managing and administrating the grants can be cumbersome, costly and time-consuming.

The conclusion? Ironically, venture capital is not the missing ingredient for most innovation ecosystems. In general, we see that there are more businesses created, more money is available for innovation and innovative firms, and there is often an extensive infrastructure supporting entrepreneurs in starting a new business. Recall that the majority of the acquired AI companies came from outside Silicon Valley.

Yet, there is clearly something missing in most ecosystems. First, the exit opportunities appear to be centered in or around Silicon Valley. Second, startup companies often struggle to scale due to difficulties in identifying the most suitable sources for alternative funding. What is remarkable in this respect is that, according to CB Insights, “only” 46% of the acquired AI companies had attracted and received venture capital.

So, what is the “special sauce” that is missing from innovation ecosystems today? And what do we need to do in order to ensure the success of an AI-focused community?

The Answer? Established Corporations . . .

Large “established” corporations recognize that they have to engage with AI, robotics and automation. They have the motive and resources to play a crucial role.

And yet, all too often, existing corporate culture and governance structures mean that older, established firms struggle to adjust to new realities. 20th century companies rely too heavily on hierarchical, formal and closed organizations. As such, they are ill-prepared to make the bold and agile decisions necessary to succeed in a world of constant disruptive innovation.

To survive, it becomes imperative for established firms to re-invent their innovation strategies. In particular, this means learning valuable lessons from the Silicon Valley experience, particularly in terms of how to “organize for innovation”.

Crucially, younger firms in the innovation sector are typically organized around three governance principles that provide them with the energy and ideas to constantly innovate, namely a “flat” organization, “open communication” and a “best-idea-wins-culture”.

Since these governance principles are more likely to be found in the organization of startup companies, large corporations try to gain access to this — what Elon Musk termed — “Silicon Valley operating system” by cultivating open and inclusive partnerships with entrepreneurs, founders and startups in the innovation space.

When multiple established corporations build relationships of this kind the basis for the “right kind” of AI ecosystem can be put in place.

But to build a network, community or ecosystem around this new type of partnership, two strategies need to be understood and embraced. In this way, large corporations can become the crucial link in building an AI ecosystem.

Creating Genuine Opportunities for Serendipity

The first strategy is for established corporations to use “corporate incubator and accelerator programs” to offer an open architecture that affords startup-founder-employees with the opportunity to routinely mix with corporate employees. Such programs have become popular in recent years. In 2017, Amazon, Apple, Facebook, General Electric and Telefónica announced the opening of new accelerator programs in France, India, the United Kingdom and the United States.

Launch of Corporate Incubators/Accelerators — Source: Corporate Accelerator DB/TechCrunch

After all, the advantages seem obvious:

(1) For established corporations. Corporate incubator and accelerator programs allow large firms to work alongside startups and their founders. These collaborations give them a glimpse into the many potential “next big things” that already are or may become relevant in the near future. In fact, the date shows that many established corporations use incubator and accelerator programs to engage/partner with startups as a form of external R&D that is focused on understanding artificial intelligence.

(2) For startup companies. Corporate incubator and accelerator programs can be alluring on many fronts. They can provide startups with necessary capital, and deliver tremendous resources in the form of knowledge sharing, distribution channels to seemingly endless rolodexes.

Nevertheless, everything depends on the structure of the program. For instance, there are programs that are powered by incubator-accelerator service providers, such as TechStars and Plug and Play. Yet, most programs are directly set up by the corporate hosts themselves.

Structure of Corporate Incubators/Accelerators (2010–2016) — Source: Corporate Accelerator DB/TechCrunch

Also, we can distinguish between programs through which a large corporation acquires a minority stake in the startup company and “non-equity” programs.

And there are more structural differences. Most programs provide for temporary office space, mentors and general business skills training. Yet, the most lucrative incubators/accelerators also give access to the tremendous resources that can be of great benefit to any company that’s just starting out.

In particular, by putting actors together in this way, the boundaries between corporate and startup can be blurred and new opportunities for positive encounters and interaction created.

Support Offered By Corporate Incubators/Accelerators (2010–2016) — Source: Corporate Accelerator DB/TechCrunch

Big companies have established distribution lines, strategic partners, deep domain intelligence, not to mention an experienced marketing and sales force and, of course, a global presence. If a startup could access even a sliver of some of these resources, it can make all the difference in its development.

Incubator and accelerator programs are crucial in securing the long-term success of both established and large corporations and startups, but such success is contingent on corporations also recognizing the vital role they can play in the ecosystem.

Aligning Incentives & “Signaling”

The second set of strategies important in building an AI ecosystem around established corporate giants is to keep things simple and to avoid any misalignment between the interests of startups and the interests of the corporation.

The main focus of the corporation should be on assisting and accelerating startups by connecting them to potentially interesting networks, customers, etc. The possible strategic returns for the corporate partners are a secondary effect or by-product of collaborating with the startups. For instance, by working “cheek-by-jowl” with startup founders, the corporate employees are better placed to identify “out-of-the-box” solutions to particular business problems. The potential benefits are particularly high in the context of a “blue sky” field such as AI where the technology is complex and the “solutions” are non-obvious.

From this perspective, non-equity programs are preferred. Avoiding minority stakes help ensure continuity in relationships. It doesn’t need much explanation to see that financial interests will often complicate the arrangements between the startup and the corporation. For instance, making money (instead of accelerating startups) could easily become the first objective of the “corporate program”.

Also, managing minority stakes in portfolio companies is often daunting from a legal and accounting point of view. More importantly, many startups fear that accepting direct or indirect investments from a corporation will restrict their future funding opportunities and bring about the risk of “negative signaling” should the corporation decide not to support or continue the investment in the future,

Equity versus Non-Equity Based Corporate Incubators/Accelerators (2010–2016) — Source: Corporate Accelerator DB/TechCrunch

You may ask: why are the corporate giants the ones having to adjust their way of working, expectations, or goals?

The answer is rather simple. Mature corporations and startups are inherently different creatures. Whilst providing a broad range of strategic benefits to startups from industry partnerships, distribution opportunities and product development insights are all well intentioned, the corporation has to signal to the ecosystem that they are a real partner to startups and that they won’t sacrifice a founder for the sole corporate strategic of financial benefits. In this way, both sides are incentivized to remain committed to maintaining and deepening the relationship.

And There is More…

Interestingly, the most active corporations in the area of artificial intelligence already seem to understand their role very well. Even after they have acquired a startup (that they may or may not have “incubated” or “accelerated”), they seek to preserve that startup’s unique identity (often by retaining the founders on CEO positions) and do not seek to assimilate it (which has been the conventional M&A practice).

And it is precisely this kind of open and inclusive partnering that needs to be at the center of an AI ecosystem, if it is to be effective. Such relationships are the “special sauce”.

The “Right Kind” of AI Ecosystem

After all, a well-functioning AI ecosystem has multiple benefits for society as a whole. In particular, it can create the necessary links between complementary sources of risk finance. Also, it helps build the capacity of entrepreneurs to identify future partners that are best placed — not only to provide money — but to deliver a meaningful, long-term relationship that will fill expertise deficits (and by doing so add genuine value and give a young firm the best chance of scaling).

Finally, such an ecosystem can help policymakers develop more dynamic and responsive regulation. In the “right kind” of AI ecosystem environment, there is a clear opportunity for regulatory measures that are built on flexible and inclusive processes that involve startups and established companies, regulators, experts and the public. It is only to be expected that this trend will lead to better and smarter regulatory results that help everyone to prepare for automation (and artificial intelligence) before it is too late.

Please push the “heart button” or leave a comment. There is a new story every Wednesday. So if you follow me you won’t miss my latest ideas about how the exponential growth of disruptive technology is changing the way we live and work.