The Age of the Platform

Introduction

Good evening everyone,It is a real pleasure to be here today at the Digital Leaders Annual Lecture. First, I would like to extend my heartfelt thanks to all of you for taking the time to join us. I know many of you are leading critical efforts in digital transformation and innovation, and your presence here tonight is a reflection of the shared commitment we all have towards building a better, more connected future through technology.

I also want to thank the event organizers and our hosts for this wonderful opportunity to speak about something that I am deeply passionate about—artificial intelligence and its future. As we gather here in London, I believe the collaboration between France and the UK, particularly in the field of AI, holds immense potential.

It's interesting to come and talk about AI in one of the most advanced countries in the field. This is the very country of Alan Turing, Donald Michie, Geoffrey Hinton, Andrew Blake, Demis Hassabis, David Silver, John McCarthy, Wendy Hall, Luciano Floridi, Zoubin Ghahramani, Margaret Boden, and Michael Wooldridge, just to name a few—some of whom have recently been awarded a Nobel Prize.

The AI landscape.

In this competition, there is one clear leader: the USA, and one clear challenger: China. Together, they account for 40% of overall AI publications and nearly 45% of venture investment in AI.

And yes, we are lagging behind—each of us individually and collectively as Europeans. Even the UK, the most advanced European country in AI, can only claim 4.7% of AI publications and a respectable 7.25% of AI venture investments. However, we must admit that this is not enough to compete with the two AI superpowers.

Among the top 10 AI companies, 7 are based in the US, and 3 are in China.

Defending who we are

Here comes the first question: should we compete? Or should we collaborate with the US, and even with China? After all, over the last 70 years, Europe, including the UK, France, and other Western countries, has experienced a very fruitful collaboration with the US.

Indeed, we are already collaborating. We share research, and there are many research centers from the "Seven Magnificent" in the UK, Ireland, and other European countries. However, we also need to defend our differences—our culture, our political stance, and our economic and geopolitical autonomy. If there’s one thing I’ve learned from my journey into the public policies of France and the EU, it’s that we have much more in common than things on which we disagree.

Nevertheless, we are lagging behing and I believe it is a crucial aspect to understanding why we’ve been outpaced by these US and Chinese companies. I would be tempted to point to at least three main factors:

1° The Market Size Effect: The US has a unified market of 320 million people. Compared to any European market, it is at least four times easier to reach critical mass for a startup. In the US, you need a certain percentage of consumers to reach that critical mass; in Germany, France, or the UK, you need four to six times more consumers to reach the same point. The same is probably true for China, where, despite the larger population, the high-end part of the market is estimated to be between 250 and 350 million people.

2° Regulatory Fragmentation: This is also true for transactional services. While it’s easy to reach global readers as a free media, when offering more sophisticated services, you have to comply with complex and diverse national regulations.

3° Access to Capital: Although pension funds represent only a small share of primary venture capital financing today, they played a structural role during the early days of the industry, from the 1960s to the 1990s. If Europe had a more integrated financial system, it could unlock an investment capability of 10 trillion euros.

4° and 5° Additionally, one could consider the lower cost of energy, which is becoming critical in the GPU data center era, and the ecosystem effect of Silicon Valley, where expertise is readily available in an almost “one-stop shopping” manner.

There’s no question that there is a temptation to mimic the US model for those who recognize the importance of the AI revolution (if we can call it a strategy). But one can easily question the success of those who have attempted this strategy. In France, there have been attempts to compete with LinkedIn, Facebook, WhatsApp, Amazon, and Microsoft. There was even a publicly funded venture created to challenge Google, which ended in failure.

This raises the question: is there only one way to approach information technologies as a nation ? Additionally, shall we we follow the US?

At the very least, one could agree that many countries exhibit clear differences. Looking at the list of unicorns (startups valued at more than $1 billion), the UK has a dominance of finance companies, Germany focuses more on industrial startups, and Israel has numerous cybersecurity players. There are also clear national strategies, like Estonia’s leadership in e-government. India, too, is interesting, having replaced commercial systems in messaging and payment systems with open-source alternatives and a lot of other countries have interesting specifications.

Productivity

For what regards AI, this might just be the beginning—the real game is only starting. Why do I say this? Because, as Nobel laureate Robert Solow famously observed, "You can see computers everywhere except in the productivity statistics." In other words, computers have not significantly improved workplace efficiency. In fact, the efficiency with which we produce goods and services is actually declining.

However, productivity experts agree that AI is a game-changer. Although it’s difficult to discern a clear trend in these early years of widespread AI adoption, it’s likely that, over time—maybe a decade—we will see an increase in productivity greater than anything we’ve seen before, thanks to the cognitive nature of the technology, which can be applied practically anywhere.

If this turns out to be true, we are facing a civilization-level change, and we need to get organized. Indeed, for this to happen smoothly, the very nature of our systems must change if possible in a coordinated way.

I see four key areas that societies and government must address:

Culture of Innovation

There is no doubt that AI carries risks, possibly even colossal risks. I don’t know precisely how risky it could be, and I believe no one truly does. We will need to handle these risks appropriately. But nothing will happen if we are not audacious, innovative, and entrepreneurial. This is a key success factor in California. After the physical frontier of the West, after the digital one, we are now facing the cognitive frontier. California has managed to maintain a spirit of curiosity and entrepreneurship that some European countries have seen fade over time.However, this spirit has not disappeared entirely; some European countries have managed to reignite this culture of innovation.Usually, when you visit a country from afar, you avoid discussing politics. I know it can be controversial, but it seems that, despite its excesses and weaknesses, the Thatcher era undoubtedly revived the entrepreneurial spirit in the UK, which might partly explain its success in the digital and AI fields.We must encourage innovation and innovators, and for that, it is essential to foster a collective culture of progress, innovation, and entrepreneurship. I will return to this point in a few minutes.

State Platformization

The state itself must transform into a platform if it wants to remain relevant compared to the services we experience with private online actors. States must also be capable of regulating algorithms to keep pace with AI advancements. Should we accept AI for minor court rulings? Should we allow AI to advise the unemployed? Should we let AI determine social subsidies?These questions will arise sooner than we expect, especially as states struggle to find financial resources to continue functioning efficiently.This transformation poses a gigantic challenge for our democracies. It requires more democratic oversight to prevent bureaucracies from seizing the power of these tools.The advent of AI presents an unprecedented challenge for modern states. I cannot think of anything comparable in the past, not even the two World Wars. How can we combat disinformation at a democratic scale? How can we handle new forms of cyber threats enabled by AI?States must be able to enforce laws not only through legal channels but also through platforms, APIs, user experience (UX), and standards. I can foresee many platforms in the future—related to cybersecurity, anti-AI scams, carbon market management, social security allocations, and payments—that may require collective oversight, whether it comes from the state, open-source communities, or civil society.But state platformization also represents an opportunity—an opportunity to chart our own course, potentially one that differs from other regional powers.

Environmental Transition

Europe is probably the region most committed to environmental transition. According to the OECD, Europe uses 14 units of carbon energy for every dollar of value produced, compared to 22 in the US and 29 in China. In other words, Europe is providing the world with a free environmental service.This is where the Carbon Border Adjustment Mechanism (CBAM) comes into play, aiming to tax imports with a higher carbon footprint than European products. In theory, this seems simple, but in practice, it’s extremely complex. Defining the full scope of carbon emissions (Scope 3) is tricky. The challenge for the EU—and possibly for the UK, should it adopt something similar—is to avoid creating a massive bureaucracy and instead use this as an opportunity to build a platform. This, after all, was part of the rationale behind the UK's decision to leave the EU.I have no doubt that AI will be one of the most promising fields for reducing CO2 emissions. I’ve written a book on the subject, which provides examples from dozens of sectors. However, certain conditions must be met. One of them is the creation of a level playing field that allows us to value and monetize CO2 emissions, as well as biodiversity. This is essentially what British economist Arthur Cecil Pigou advocated a century ago in The Economics of Welfare.Both Europe and the UK, one of the most disciplined countries in environmental transition, can take advantage of AI to facilitate this transition and platformize it as mentioned earlier.

Education and Culture of AI

Earlier, I mentioned the culture of innovation. One thing that strikes me, not just in France but in many countries, is the deep skepticism around AI and, more generally, the concept of progress. People believe it doesn’t serve them. In France, I was struck by a survey that revealed 43% of French citizens explicitly did not want to be trained in AI. To better understand this, the National Digital Council, which I chair, engaged with rural communities and organized 120 events to discuss topics related to the digital revolution and AI. The results were so impressive that we went on to convince the President and the current government to launch "Café IA."[Café IA description here]

French-UK Cooperation

Before wrapping up, I want to make a comment about French-UK cooperation.Since Brexit, cooperation between our two countries has slowed, and I find that regrettable.

Final Remarks

For some the final judgment may come from God, but for all of us, it will most certainly come from the future generations. When they look back, I believe they will focus on this very moment—this decade and the next—as the time when we had to make the right decisions on the environment, geopolitics, and AI. It is therefore our most pressent duty to be able to be able to resume the discussion about long term question.