In the first televised presidential debate in September 1960, presidential candidate John F. Kennedy challenged his opponent, Vice President Richard Nixon: “The Soviet Union now has in the field more missiles than we do … and unless we are willing to reappraise our whole defence policy, we are in danger of losing our position in the world.”

The “missile gap” became a central theme of Kennedy’s campaign and contributed to Nixon’s narrow defeat. But it was a myth: the U.S. had always maintained a clear lead in intercontinental ballistic missiles.

Yet, the narrative of the “missile gap” had already become a widely accepted fallacy in public discourse. Unable to walk back from his campaign rhetoric, Kennedy doubled down and justified the increase in defence spending with the need to maintain U.S. superiority in defence capabilities.

The “missile gap” is a striking example of how a flawed narrative can set the course for real-world decisions. Today, a similar dynamic is playing out in AI policy, where speculations about “superintelligence” shape the policy agenda, despite disagreement about what it is or would mean for society.

Towards the AI frontier

Shortly before his promotion to UK’s Trade Minister this month, Technology Minister Peter Kyle expressed confidence that we would achieve “superintelligence” in the very near future: “If my dream of winning a second term happens, I think it’s an inevitability”, he told TIME magazine.

Among industry representatives, this enthusiasm is met with scepticism: POLITICO quotes a lobbyist saying that Kyle’s statement feels “disconnected” to the corporate debate where many companies struggle to get a return on investment from their AI investments.

AI critic Gary Marcus dismissed the “superintelligence” narrative after what he viewed as a disappointing release of ChatGPT-5. According to Marcus, the technology behind current transformer models is not capable of reaching “superintelligence”.

What does the fox say?

Policymakers should approach both extremes – the superintelligence evangelists and the AI skeptics – with caution. Both narratives are more like strongly held beliefs rather than empirically tested theses. While there is some evidence that the pace of AI progress is slowing down, AI labs are also investing in new technologies that may continue to deliver breakthrough results.

In his book “Expert Political Judgment”1, Philip Tetlock finds that while people with one big idea (“hedgehogs”) often get a lot of attention because of their storytelling skills, they are not very good forecasters. By contrast, experts who build their knowledge on multiple, diverging perspectives and draw conclusions based on probabilities, not world views (“foxes”), are more accurate forecasters. Their predictions, however, are less headline-grabbing.

Subscribe now

When it comes to “superintelligence”, a “fox” might advise policymakers that it does not really matter whether we're on the path to “superintelligence” or not. Even if AI progress came to a halt today, there would still be an enormous opportunity to integrate this technology into our education systems and our economies. And if AI progress continued at scale, the diffusion of this technology would still take time which would allow for adjusting our societies and our economy.

Towards a “mission economy”

The benefit of the race to “superintelligence” is not necessarily the end goal of reaching a final frontier in AI research. It may be more useful to view “superintelligence” through the lens of Mariana Mazzucato’s “Mission Economy”, an approach where governments define bold public policy goals and then mobilize resources to achieve them.

Without any doubt, “superintelligence” is a powerful, inspiring mission, just like the Apollo program was in the 1960s.

But importantly, the U.S. did not strive to bring a man to the moon because the moon was such an attractive destination. The U.S. launched the moon mission because of all the technology that needed to be invented to eventually reach the moon.

Some of this technology will be highly specialised, but some will appeal to the broad masses. Business professor Ethan Mollick recently coined the term “mass intelligence” or powerful AI that is “as accessible as a Google search”. While “superintelligence” will surely require policy responses, “mass intelligence” also needs the attention of policymakers, both with regard to managing its risks and accelerating its adoption..

So instead of asking whether we're close to “superintelligence” or not, perhaps policymakers should think about how artificial intelligence can help improve our societies today. If and when “superintelligence” arrives, this will will give us a head start – and if it does not arrive in the end, we would still have made the best of this technology.

If you want to read more about AI and its impact on how we work and live, subscribe to Work/Code, a new newsletter I am launching with Markus Albers in October. In interviews with technologists, academics, creatives, workers, and managers, we look behind the headlines and uncover the quiet revolutions unfolding inside companies, economies, and countries.

IN politics, plans should not be overestimated – but they do reveal intend. That the United Kingdom, the EU and the United States have published “AI action plans” in recent months is a sign that the AI policy debate has shifted and that artificial intelligence is increasingly seen as a growth engine by policy makers.

But is there any other common ground between these plans? 2023 and 2024 have been characterised by a surprisingly large amount of international cooperation on the G7 and OECD level and even in the halls of the Vatican. Does the new focus on economic impact mean that cooperation on AI safety is replaced by competition?

Each of the three plans – the UK’s “AI Opportunities Action Plan”, the EU’s “AI Continent Action Plan” and “America’s AI Action Plan” have different intentions: The U.S. plan is an executive strategy. It is the basis for subsequent Executive Orders directing agencies, procurement rules, and standards guidance. The EU, on the other hand, can mobilise programme funding and shared infrastructure – and the EU AI Act already sets a common rulebook. The UK’s plan is a mix of government as a “first customer” and proposed regulation.

This means that both the tonality and the content of each of these plans differ slightly, and yet, they all touch on a number of key themes: AI infrastructure, data, AI adoption, skills, industrial policy, and regulation.

Let's look at each of these topics in more detail.

AI Infrastructure

Infrastructure is a key component of all three plans. The UK has the most internationalist approach and differentiates between “sovereign compute”, “domestic compute” and “international compute”. The U.S. and the EU on the other hand focus on expanding domestic compute through public investments (in the EU) and streamlining permitting (in the US). Both the UK and the EU intend to use the new compute power for supporting domestic companies to grow without having to worry about access to high-end compute power.

WHO’S AHEAD? The United Kingdom has the least protectionist and most pragmatic approach to building new data centres: The idea of getting access to compute power in other countries is an interesting and pro-trade approach and the Brits’ plan to use the national compute to advance the Labour government’s “missions” and projects of national importance promises a better bang for the buck compared to the EU which focuses a lot on providing access to SMEs. The UK also proposes “AI Growth Zones” with lighter permitting but still considers environmental concerns.

Data

Data features prominently in the EU’s “AI Continent Action Plan”. The U.S. merely commits to developing standards for scientific data and incentivise researchers to make more data available while the UK equally only proposes more open government data (which certainly is very helpful) and a “British media asset training data set” which I hope will contain “Yes, Minister”. The EU on the other hand has already adopted legislation to incentivise data sharing across industries (albeit with limited success) and is currently in the process of simplifying and modernising these rules.

WHO’S AHEAD? Not only did the EU show foresight by including the text-and-data-mining exemption into the EU Copyright Directive (something the Brits want to copy), but the Continent’s thinking about creating data marketplaces is also more advanced than in the UK and the U.S. However, there is a big gap between the EU’s intentions with data marketplaces and their implementation. It remains unclear what should incentivise companies to share data with others and what added benefits marketplaces would bring to that.

AI Adoption

Harnessing AI to kick-start growth is the key theme in all plans, starting with the U.S. that sees itself in a geopolitical race with China. The U.S. therefore proposes a “try first” culture across all industries and advanced adoption in government through interagency collaboration, talent exchange and sharing best practices. In addition, the U.S. wants to quickly adopt domain-specific standards to drive the adoption of AI.

The UK's “AI Action Plan” sees the UK government as the “first customer” and urges public administrations to use their purchasing power to support the British AI ecosystem by “moving fast and learning things”. The EU focuses their adoption efforts on SMEs and wants to transform the “Digital Innovation Hubs” that were created in the past five years into “Experience Centres for AI”. That is probably not going to go fast and will be hard to scale. The EU’s announcement to host “structured dialogues” to foster adoption in key sectors sounds more promising, even though not very innovative.

WHO’S AHEAD? The U.S. government gives officials and companies the license to integrate AI deeply into their operations. And while this is not without risks (contrary to popular belief, AI is not unregulated in the U.S.), this will surely lead to faster adoption.

Skills

When it comes to skills, there is one big difference between the U.S. plan and the visions of the UK and the EU: Immigration. While the UK and the EU want to attract international talent (the former even with a dedicated headhunting unit), the U.S. plan does not even mention the term once. But with such a density of talent (and large pay checks), there is probably also no need for a dedicated AI immigration push in the U.S.

The need to up-skill workers for the age of AI is a no-brainer and again, the UK has the most pragmatic approach by suggesting publicly funded skills programmes for AI like in Singapore, South Korea – or France.

WHO’S AHEAD? The UK has the most pragmatic approach. When it comes to education, public money is almost always well spent, and that is also true for AI skills development.

Industrial Policy

Each country starts from a very different position with regard to their industrial base, so naturally, the focus is quite different. To the surprise of many, the U.S. plan praised the role of open source AI and is also most advanced when it comes to accelerating scientific development through AI. The UK equally wants to appoint sectoral “AI champions” in sectors where the UK already has a strong presence and – ultimately – support the creation of “national champions”. The EU has the most comprehensive understanding of industrial policy – from building infrastructure to supporting European generative AI models.

WHO’S AHEAD? tbd; it is hard to directly compare the different approaches, but clearly each jurisdiction plans to push adoption in key industries.

Regulation

The perspective on regulation shows the greatest level of differences between each jurisdiction. After two and half years of close cooperation on AI safety, governments are starting to explore different paths.

In terms of regulation, the EU has already cast their dice with the EU AI Act which has already started to come into force. But in an effort to speed up AI adoption, Brussels is planning to follow-up with the “Apply AI” strategy and a simplify compliance with the AI Act.

The U.S. plan promises to revise or repeal regulations and rules that could hinder AI development and suggests to tie funding for federal states to their stance on AI regulation. In other areas, the Trump Administration adds compliance burden, such as with the obligation for federal agencies to only procure AI that is “free of ideological bias”.

The UK seeks a middle way by maintaining the AI Safety Institute’s mandate to do research on model evaluation, foundational safety and societal resilience. They also want to oblige regulators to regularly report how they “supported growth” by enabling safe AI adoption (I can already imagine Sir Humphrey writing this report).

WHO’S AHEAD? The transatlantic partnership. It was maybe a bit surprising that “America’s AI Action Plan” featured a whole section on international diplomacy (even though under the premise of maintaining U.S. leadership in AI). And the EU, too, committed itself to working with likeminded partners.

The first era of the internet was characterised by liberal policies such as Section 230 and the European E-Commerce Directive, followed by the Brussels Effect when European rules like the General Data Protection Regulation started to have a global impact to counterbalance the Wild West years of the world wide web.

The AI Safety Summits of 2023 and 2024 showed that there might be a third way: While every country is racing to reap the benefits of AI adoption, there is still a lot of interest in international cooperation on safety and standards. This can only lift all the boats.

Publisher’s note: In a previous post, I have compared the UK’s “AI Opportunities Action Plan” and the EU's approach towards AI.

Did you get the memo yet? In recent weeks, numerous CEOs have sent AI memos to their employees, musing about the impact of AI on the labor market. The most disturbing one came from Fiverr CEO Micha Kaufman who wrote that “AI is coming for your jobs” and that “your value will decrease before you know what hit you”.

Similarly, Shopify CEO Tobias Lütke bluntly stated: “In a company growing 20-40% year over year, you must improve by at least that every year just to re-qualify”. And: “Before asking for more Headcount and resources, teams must demonstrate why they cannot get what they want done using AI.”

While the purpose of these memos surely is to create a sense of urgency among employees, they also spark anxiety and cynicism, judging from employees’ comments in internet forums.

This is a bad starting point for a transformation as big as the emerging AI economy.

Leaders need to acknowledge workers’ anxiety, not fuel it

“When people are anxious about job security, layoffs, and the constant push for productivity, they’re not in a mindset to learn. We first need to address those anxieties directly – only then can people feel curious and open to exploring AI,” Helen Kupp, co-founder of Women Defining AI recently said at the Work Forward Forum.

When describing the magnitude of AI’s impact on the economy, it is fair to compare today to the industrial revolution (even though Sam Altman suggested over lunch with the FT that he prefers to compare the AI age to the renaissance – we will come back to this later).

The key point is that AI is a transformational general purpose technology, just like steam or electricity, that will reconfigure our economy. The question is how this can be done quickly and without disrupting society. This requires us to adapt how we run businesses, regulate markets and build institutions.

Running an AI business

Firms react differently to emerging technologies: Incumbent firms tend to favor quick productivity gains by using AI to automate processes. This may however lead to accelerated path dependency and a decline in quality (as Klarna found out – or not).

Firms at the “technological frontier”1 however are likely to use AI to reimagine work processes from scratch and become AI “superstar firms”. In these firms, AI agents will increasingly emerge as a “digital workforce” that works alongside humans.

A memo however is not sufficient to get there. AI is not yet another tech rollout, it requires more fundamental changes.

Luckily, there is a precedence for how companies can do this: The shift to remote work during the pandemic. In her excellent book “Redesigning Work”,2 Lynda Gratton describes how the pandemic created a window of opportunity for firms to rethink knowledge work around the concepts of energy, focus, coordination and collaboration.

Similarly, with AI, firms should assess roles and tasks of their employees and identify areas where AI can accelerate processes, create shortcuts or maybe even cut redundancies.

“Instead of thinking about how vibecoding accelerates engineers, focus on its transformative impact for product managers and designers”, Brian Elliott writes in his Work Forward newsletter. “Suddenly they can jump straight to rapid prototyping (…) before developing formal requirements docs. This fundamentally changes workflows, not just accelerates them.”

This is the opportunity for innovation and growth, not mere efficiency gains.

Distributing AI benefits fairly

Governments, on the other hand, must create frameworks for equal access and fair distribution of AI benefits.

In “Power and Progress”,3 Nobel-Prize winner Daron Acemoğlu and Simon Johnson stress the importance of institutions and social power for the adoption and social acceptance of new technologies: “Progress has a way of leaving many people behind unless its direction is chartered in a more inclusive way. Because this direction governs who wins and who loses, there is often a struggle over it, and social power determines whose favorite direction prevails.”

It is not surprising that disruptive technologies that threaten to upend the livelihood of workers create resistance. This was the case in 19th century Britain when the Luddites destroyed manufacturing equipment in protest of the de-humanizing and labor-replacing properties of theses new machines.

Britain’s initial reaction to the Luddites was harsh: In 1812, Parliament passed the Frame-Breaking Act which made the destruction of machinery a criminal offense that was charged with the death penalty.

But even though the Luddite movement itself was crushed with force, it sparked a wider debate about the need for worker representation in political life and a fairer distribution of the economic gains of industrialization.

This debate would transform Britain into an economic powerhouse: By the mid-nineteenth century, Britain had become a “nation of upstarts”: The dissolution of traditional social classes allowed citizens, often with no or only little formal education, to advance their social status as innovators and entrepreneurs. These self-made men drove the adoption of new technologies.

In Germany, the economic boom during the industrialization was accompanied by the creation of thousands of new and innovative companies is known as the “Gründerzeit” (or “founder’s period”). Some of these firms, such as Siemens, Bayer or BASF still exist today.

That is why the comparison of the age of AI to the “Gründerzeit” (if you want to avoid the term “industrial revolution” with its connotation of poverty and suffering) is still more fitting than a comparison to the renaissance (as Sam Altman suggested). While the renaissance saw scientific discoveries and the emergence of a class of super-rich merchants, not much changed for most of the people living at that time.

The industrial revolution however challenged not only the predominant economic model but also the structure of society just like AI changes both how we work and our self-image as humans and social beings.

Avoiding the pitfalls of the AI revolution

Luckily, we can learn from historic experiences in the industrial era to make the AI revolution a less painful transformation than industrialization:

GIVE WORKERS A WORKPLACE GUARANTEE To address short-term fears of being made redundant, CEOs could give a workplace guarantee for every employee for a certain period of time instead of writing threatening AI memos. During that time, each employee can safely experiment with AI and re-imagine old routines and work processes without instantly being judged against arbitrary productivity metrics (I used AI in various ways during the research and writing of this article and AI is by no means a simple shortcut to creative expression).

Companies should follow the advice of Wharton School professord Ethan Mollick: “The future of work isn’t just about individuals adapting to AI, it’s about organizations reimagining the fundamental nature of teamwork and management structures themselves. And that’s a challenge that will require not just technological solutions, but new organizational thinking.”

This task cannot be delegated to frontline workers – CEOs, management teams and boards must lead this effort.

Workplace guarantees are not uncommon: In transition phases, large German corporations often give workplace guarantees to their employees, and during COVID, some companies promised their employees to keep them on the payroll even though offices were closed to reduce anxiety among their workforce.

UPGRADE SOCIAL SECURITY NETS Voluntary workplace guarantees could be supplemented by a social security net styled after Denmark’s principle of “flexicurity”. In Denmark, employees have less protection from being made redundant but are supported with generous unemployment benefits as well as training to quickly transition into a new job.

INVEST IN UP-SKILLING AND RE-SKILLING Investments in skills and training are another essential element for a transition strategy towards an AI economy. Singapore for example has launched the SkillsFuture fund which supports citizens aged 40 or older to up-skill and re-skill with up to 3,000 USD per month. Similarly, France grants training credits of up to 5,000 EUR to every employee and job seeker to support the acquisition of new skills.

Towards a new “Gründerzeit”

AI is not a force of nature, it is a technology. And while it is obvious that AI will massively change the economy and the labour market, we as a society ultimately decide how AI will be deployed.

Governments need to make their social security nets “AI proof” and invest in up-skilling and re-skilling of workers to make the transition into AI jobs easier.

There may be parts of the population that are especially vulnerable to AI, such as graduates and workers in the second half of their working life. The latter could be supported with skills training, the first group with apprenticeship-style entry level roles that may be partially subsidized by the state and allow graduates to get into the workforce.

Governments should also incentivize the founding of new companies to pave the way to a “new Gründerzeit”. Just like the internet, AI massively reduces the barriers to start a new business. And even if many of these firms will fail, the experience of founding and running a company are important transferable skills.

In their essay “AI as Normal Technology”, Arvind Narayanan and Sayash Kapoor look at AI from a refreshingly sober perspective, without falling for AI doomsday scenarios or unrealistic promises of endless growth. Just like I have written previously, Narayanan and Kapoor argue that the discovery of a new technology is something very different than the diffusion of this technology into our everyday life, which often takes decades, not years.

For policy makers and CEOs, the reassuring message is that this is the time for an AI strategy, not an AI memo. Narayan and Kapoor believe that building resilience is the best approach to navigate through this period as we still discover how AI will shape our work and our lives.

This is sound advice that policy makers and CEOs should adhere to.

Update (6 March 2025): Added a link to Bertelsmann Foundation’s report on the EuroStack.

THE recent “AI Action Summit” in Paris and the speech by US Vice President JD Vance at the Munich Security Conference were important inflection points for Europe.

The US attempt to reach a peace agreement in Ukraine through bilateral negotiations with Russia proves that Europe no longer has geopolitical influence. Global trade relations, meanwhile, are becoming increasingly transactional. This is particular harmful for Europe, whose “soft power” relies on trade. And lastly, in tech policy, relations with the US are equally adversarial.

The EU reacted by launching an initiative to cut red tape and increase the region’s competitiveness. At the same time, the Commission announced increased spending on tech infrastructure, while French President Emmanual Macron presented an ambitious plan to invest EUR 109 billion in France alone – a sum that according to Macron matches the American Stargate project.

Tech infrastructure: Now or never?

JD Vance’s speech at the Munich Security Conference (video) left European diplomats wondering whether the US can still be considered an ally. If there ever was a time to build a home-grown alternative to US infrastructure, it is now.

Earlier European initiatives such as Gaia-X were overly politicized from the beginning. But slowly, European industry is catching up. One of the challenger firms is Stackit, a cloud platform that was built for internal use by the German retail company Schwarz Group before it was made publicly available in 2022.

Similarly, the recently announced “EU AI Champions Initiative”, led by VC firm General Catalyst, is committed to invest EUR 150 billion into the European AI ecosystem in the next five years. A report published by the initiative at the Paris AI Action Summit reads like Europe’s response to the UK’s “AI Opportunities Action Plan”.

This is a refreshing alternative to Europe’s traditional “me too” approach to industrial policy that focused on copying US business models. Instead, European firms now take a page from the playbook of French AI firm Mistral: The company was created in 2022 as a result of a sense of “panic” that Europe was behind in generative AI. But because Mistral lacked the capital and compute to compete head-on with the large US labs, the company had to innovate and developed a competitive model at a fraction of the cost of the leading foundation models.

Scaling AI in Europe however remains a challenge: According to the think tank Group d’Études Géopolitiques, Europe would need to invest EUR 500-700 billion into energy infrastructure to achieve 16% of global AI compute power, which is proportionate to the EU’s weight in the global economy. A far cry from what is currently in the budget.

A tech trade war will threaten Europe’s competitiveness

Europe’s dependency on US tech is a “critical weakness”, writes Marietje Schaake, a former member of the European Parliament. And the Financial Times’ Martin Sandbu believes that the EU must refrain from the natural desire to prevent a tech trade war with the US.

This view, however, is mistaken. The digital services that Europe buys from the US cannot be replaced easily with competing offers. In some instances (social media, for example) this may be annoying for consumers but not business critical. But in other cases like cloud infrastructure or AI, it is important to retain access to the best products in the world. A world-class large-language model cannot be replaced as easily as Russian gas.

Both Schaake and Sandbu point to the Eurostack, a new initiative that aims to reduce Europe’s dependency on Big Tech by creating a full-blown IT stack to firmly position the EU across the entire technological value chain, from physical infrastructure to software and, ultimately, digital applications and services.

So far, however, the Eurostack is not more than a “pitch”. And while its authors recognize that traditional approaches to industrial policy in the tech sector such as Gaia-X have failed, it is far from sure that this approach can be more successful.

The Eurostack does not change geopolitics

The Eurostack report itself recognizes that “complete self-sufficiency is neither feasible nor desirable” and that international partnerships are still required. Computer chips and energy production all require critical input in terms of raw materials, intangible assets and technical components that that are subject to the same geopolitical tensions that the “sovereignists” want to overcome.

There are also economic questions around the Eurostack that remain unsanswered so far. Ansgar Baums, co-author of “Tech Cold War”1 believes that developing a Eurostack is economically not viable and that the EU should instead double-down on co-dependency with the US by creating more firms such as ASML that are indispensable for the Western IT stack.

Becoming a tech leader takes time. But there is a shortcut, and this shortcut is rapid adoption. Europe’s AI opportunity is the wide adoption of AI models across all sectors and industries. This does not mean using chat bots to cut costs in customer support, but using AI in in research, development and robotic process automation where it can truly transform industries and support the creation of AI-first companies that will hopefully become “superstar firms” one day.

The proposed “28th Regime”, a new, pan-European legal framework for startups and scale-ups might be the right path towards this (provided member states share the Commission’s ambitions). But there is more work to be done on strengthening the Single Market, providing access to cheap energy and capital and attracting international talents.

The EU tech policy flywheel

The US did not become dominant in tech by creating national champions but by designing a market. And the best way to do this is to create a “flywheel”, where every piece of the flywheel reinforces the whole.

Speaking to Bloomberg, French tech mogul Xavier Niel lays out in a compelling way (“almost as a general would”) how such a flywheel could look like:

The priority (...) is talent and how to keep it. That means funding startups and entrepreneurs and offering them the computing power that resource-hungry AI demands. Then comes the need for actual users – to “democratize access to AI.” (...) This will be a critical path for long-term productivity gains that Europe badly needs.

Only then comes infrastructure (understood in the broader sense of the Eurostack) which in turn deepens the maturity of the talent market and the diffusion of new technologies.

Another way to look at the flywheel is from an impact perspective: Where can Europe get the biggest impact with the smallest input? Building data centers, cloud infrastructure and networks has lots of external dependencies, such as access to chips, raw materials and insecure supply chains. Increasing the quick adoption of new tech merely requires access to these technologies while attracting talent requires next to no input from third parties.

Europe should start with this first step, and then go faster and further.

📚 Read on

• In their recently published report, the EU AI Champions initiative urges policy makers to streamline tech regulation, accelerate AI adoption, invest in tech and energy infrastructure and support AI skills development.

• In this podcast, my former colleague Jack Keevill gives and excellent overview about the Commission’s simplifaction agenda and its implications for tech.

• The “pitch” for the Eurostack is worth reading and apparently, another report by Bertelsmann Stiftung and Francesca Bria.

• Chamber of Progress researcher Kay Jebelli has published in in-depth thread on the opportunities and obstacles for the Eurostack.

THE worst case scenario for Germany was narrowly avoided last night: For a long time, it looked like the BSW, a new, populist movement led by Sarah Wagenknecht, would snatch enough votes to enter Parliament. In that case, Friedrich Merz, the likely new Chancellor and leader of the conservative CDU, would have been forced to form a threeway coalition with the Social Democrats and the Greens if he wanted to stay true to his promise to not form a government with the biggest winner of the election, the righ-wing AfD, which came in second in the polls.

In the end, BSW did not get into Parliament and a "grand coalition" between the CDU and the Social Democrats is now the most likely outcome of the coalition negotiations that will start in the coming days.

Many in Europe (including me) hoped that Friedrich Merz would make Germany a more decisive player in the Council and end the "German vote" where Germany stays silent in policy debates or abstains in important decisions in Brussels because the coalition cannot agree on a joint position. Merz himself at least has frequently (and rightly so) criticized Chancellor Olaf Scholz' policy on Europe and promised to do better.

However, as the dust settles, it is not entirely clear how he can deliver on this promise: On many of the most controversial topics in Europe like migration, budget and the green transition, the position of the Conservatives and the Social Democrats are far apart. In addition, the Social Democrats in the German government might be tempted to use German support as a bargaining chip to stop the EPP from forming voting blocks with the ECR at the expense of the progressive parties.

There is another reason why Friedrich Merz ist not the power broker many think he might be: His party's performance was in fact not as good as it seems at first sight: looking only at Eastern Germany, the AfD is the strongest party with 32 per cent of the vote (CDU: 18.7 per cent!). In Western Germany, the only state where the Conservatives outperformed the national average considerably was Bavaria.

However, Friedrich Merz was not on the ballot paper in Bavaria: The CDU does not have their own candidates in Bavaria. Instead, voters vote for their Bavarian sister party CSU, led by Markus Söder. Markus Söder however, just like the party's chiefs in Eastern Germany, are considerably more right-wing than Friedrich Merz is. They might argue that in order to win back votes, the CDU needs to become more like the AfD.

In last nights talk shows, everyone agreed that time is of the essence when entering coalition talks now. But this is especially true for Friedrich Merz. Alice Weidel, the head of the AfD, proclaimed that her party is open to a coalition with the Conservatives and emphasized the huge overlap in their party's respective election manifestos in areas like migration, energy and economic policy. The fundamental shift in policy that the Conservatives said was needed for the country will not be possible with the ailing Social Democrats, Weidel added.

Merz categorically ruled out talks with the AfD, but the longer talks with the Social Democrats last (or if they were to fail in the end), the more likely it is that the mood in the Conservative Party will shift against Merz and his politics of isolating the AfD. It was Merz, after all, who said just a few weeks before the election that if the policy is right, he doesn't care who supports it.

These words may still haunt Merz who, two decades after Angela Merkel ousted him from politics, is close to fulfilling his lifelong dream of becoming Germany's chancellor. And yet, he is not quite there yet.

THIS Monday, the UK government published its “AI Opportunities Action Plan” (“AI Plan”), a set of 50 policy proposals drafted by British entrepreneur Matt Clifford. As the UK is trying to develop a “third way” for AI policy, it is worthwhile to have closer look at how the British plan compares to the EU.

The AI opportunity

The AI Plan follows a vision that is very similar to the EU’s: The UK aims to be an “AI maker, not an AI taker”. In other words, the UK wants to create national champions along the whole AI value chain. To achieve this, the UK pledges to invest in AI foundations (particularly computing and data infrastructure) and push the adoption of AI. With regard to adoption, the plan’s focus is on the public sector with the hope that this will subsequently encourage private sector investments in AI.

WHAT THIS MEANS FOR THE EU: While the vision is similar, the sequence of priorities is different in the EU. EU digital policy is driven by the implicit tension between creating European champions and driving adoption and sometimes, the EU creates champions without customers. The UK tries to mitigate this risk by leveraging the purchasing power of the UK government, which is more difficult (although not impossible) to achieve for the European Commission as national governments will be more hesitant to coordinate their procurement decisions.

Building AI infrastructure

In the context of the AI Plan, things like compute power, training data, skills and talent, but also regulation and AI safety are summarised under the broad umbrella of AI infrastructure.

According to Clifford, “access to compute will be a key pillar to economic security” which makes “ownership of critical strategic assets” essential. To achieve this, the AI Plan suggests a tiered approach that distinguishes between “sovereign AI compute”, “domestic compute” and access to “international compute” through agreements and partnerships.

In relation to the EU’s own ambitions in AI, a few policy proposals in the AI plan stand out:

• Clifford proposes to expand the capacity of the AI Research Resource (ARR) by 20x until 2030. While he emphasises that this does not equal an increase in budget by 20x (the ARR's budget is currently around one billion Euro), it is still possible that the UK’s investments in AI compute will surpass the EU’s current financial commitments.

• Dedicated ARR programme directors with significant autonomy shall decide on the allocation of the UK’s sovereign AI compute, based on the five “missions” (or priorities) that Labour has set.

• “AI Growth Zones” shall be established to speed up the building of new data centers; and

• international compute agreements with likeminded partners (including the EU) should be negotiated.

WHAT THIS MEANS FOR THE EU: The tiered approach to compute infrastructure could be a model for the EU, given the current deadlock in the sovereignty debate. It admits that some level of sovereignty is useful, but does not attempt to make sovereignty a de-facto standard and thus a protectionist measure.

The proposal to designate “compute czars” that help would-be champions to get access to AI compute in strategic areas such as health, manufacturing or green technologies is an interesting variation of the EU’s “AI Factories” that support initiatives which promise a high return on investment.

Access to data

The AI Plan acknowledges that data is a key ingredient of AI-driven growth and proposes the creation of a National Data Library that not only publishes data sets, but also shapes which data is being collected for the future training of AI models. Access to proprietary data sets could be linked to the allocation of compute power and incentivise research and development. The UK government is also encouraged to license a copyright-cleared data set with media assets – a controversial proposal which has already attracted some criticism.

WHAT THIS MEANS FOR THE EU: With regard to data access, the EU is probably ahead of the UK with initiatives such as the Common European Data Spaces and laws like the General Data Protection Regulation (GDPR) and the Data Act that regulate data access. But while the effectiveness of the Data Act still needs to be tried and tested, the GDPR has its own flaws, including a high bar for data anonymisation (which is a key element of the UK’s plans to make health data accessible to research).

Teaching skills and attracting talent

The amount of AI talent that the UK needs to train according to the AI Plan is enormous: Within five years, Britain is supposed to “train tens of thousands of additional AI professionals”. Given that a Bachelor degree already takes three years, these professionals essentially needed to start studying yesterday. That is why there is also an urgent need to broaden educational pathways into AI through professional education and up-skilling.

An interesting approach to convincing talent to relocate to the UK is the proposal to establish an “internal headhunting capability on par with top AI firms to bring a small number of elite individuals to the UK”.

WHAT THIS MEANS FOR THE EU: Given that the EU is not a nation state but a union of states, it will be politically difficult to mirror the UK’s efforts in AI headhunting at the EU level. But I can certainly imagine leaders from France and Germany, Poland or Portugal jet-setting the West Coast, poaching founders to build their business in Europe ...

Regulating AI (or not)

The chapter on regulation is full of ambiguity from a European point of view: While the EU gets a shoutout for the innovation-friendly text and data mining exemption in the Copyright Directive and for the idea of regulatory sandboxes for innovative AI solutions, the AI Plan generally proposes a more conservative approach to regulation.

The work of the AI Safety Institute (which among other things conducts pre-deployment evaluations of new foundation models) gets praise by Clifford, too, while the AI Plan still suggests a formal regulation of foundation models (a consultation on this topic is expected soon).

In general, however, the UK is likely to follow a sectoral approach to AI regulation. In addition, the UK government pledges to support the “AI assurance ecosystem”, which essentially means a self-regulatory approach to most AI use cases.

WHAT THIS MEANS FOR THE EU: The EU has of course already implemented AI regulation. With regard to foundation models and low risk AI, the UK seems to be following an approach similar to the EU AI Act, while the regulation of riskier AI models will follow a sectoral approach. The EU on the other hand has both horizontal laws (the AI Act) and sectoral rules e.g. for AI in medicine technology.

Driving the adoption of AI

The adoption of AI is seen as a critical element of the AI Plan. Central to this goal is the rapid adoption of AI in the public sector, following a “scan, pilot and scale” approach where the public sector identifies an AI use case, creates a pilot and scales adoption quickly when a pilot succeeds.

WHAT THIS MEANS FOR THE EU: Given that the EU has only very limited say in public procurement, this approach cannot be easily used as a blueprint for the EU, even though member states could adopt this model for themselves.

Driving adoption in the EU will in reality be slower than in the UK, because existing laws make adoption more complex. The AI Plan for example states that teachers were able to cut down the 15+ hours per week that they spend on lesson planning and marking. Under the EU AI Act, however, the use of AI to grade an essay could be a “high risk” use case. Guidelines on how AI can be safely used in education and other sectors should therefore be a priority to accelerate the adoption of AI in Europe.

Comparing the UK’s and EU’s approach to AI

Of course, the AI Plan cannot be a blueprint for the EU: the EU has only limited powers and a lot of the proposals made by Clifford would have to be adopted by EU member states rather than the European Union.

And still, it is valuable to compare the approaches and acknowledge both the similarities (e.g. on the importance of developing national champions along the AI value chain) and differences (a sectoral rather than horizontal approach to regulation). Even though the EU might not be able to scour the world for AI talent, it can create standards for data collection and use, for safe anonymisation of personal data for AI training and for the use of AI in schools.

Finally, what I found striking is the inspiring language of the AI Opportunities Plan which written in the spirit of Mariana Mazzucato’s “mission economy”. Unfortunately, this language does not seem to resonate well with decision makers in the European Union as the Financial Times reporter Martin Sandbu wrote recently. Sandbu diagnoses a “committment phobia” in the EU: while it often surpasses itself in times of crisis, it is less good in setting long-term goals and then vigorously pursuing them.

This seems to be a great moment in time to learn this skill.

📚 Read on

In this noteworthy essay, Carl-Benedikt Frey and others sketch out how policy choices influence the adoption of AI and productivity growth.

This slightly older (September 2023) but still relevant focus group study by Milltown Partners) sheds some light on the public’s expectations to achieve responsible AI. The interviewees were based in the UK, the US and Germany.

In this video, Google’s Nicklas Lundblad speaks about the delicate balancing act of regulating AI for safety and economic benefits, drawing heavily on many examples from (contemporary) history of tech regulation.

OVER the holidays, an excerpt from an interview with Microsoft CEO Satya Nadella went viral in which Nadella allegedly predicted the “death of software-as-a-service” (SaaS) because of the emergence of ever smarter artificial intelligence (AI) agents.

But Nadella did not at all speak about the “death of SaaS apps”. He repeatedly referred to business apps as “canvases” for writing, calculation or other business functions.

AI agents will change, but not replace SaaS

Still, it is clear that the way how (business) users interact with apps will change dramatically as AI agents become the primary interface to SaaS apps for many users.

An AI agent might for example proactively identify customers for a specific seasonal upsale based on information in your CRM, draft and send a sales email using your mail service, register a purchase on your website and trigger the shipment of the product in your e-commerce platform, monitor complaints through your customer support platform and feed back all this information into your main CRM.

Looking at the complexity of these workflows, it seems unlikely that a single AI system will be able to integrate all of these activities in one single piece of software in the near future (and at this point, the agent will essentially have become a SaaS product itself).

In the interview, Nadella also seems to suggest that certain elements of a software (such as a spreadsheet) could be created “on the fly” by a large language model. But it is doubtful whether this would be economical. While it is nice to be able to create a minimum viable product in an AI chat bot, it would not make sense to re-invent the wheel every time someone makes a certain standard query.

Legacy software will slow down adoption of agents

The current pace of technological progress is extremely fast, but as these technologies meet legacy technologies, there will be all kinds of unforeseen challenges to unlocking the full potential of AI. Organisations and societies naturally change at a slower pace than start-ups do.

In his essay “Machines of Loving Grace”, Anthropic co-founder Dario Amodei introduces the concept of a “compressed 21st century”1 because advanced AI is likely to allow us to make discoveries in health and biology in a time span of five to ten years that would require 50 to 100 years of research without AI.

In the same essay, Amodei however also acknowledges that existing regulations may significantly slow down the impact that new technologies can have. It is therefore not surprising that it may take decades until we see the economic impact of AI (I actually think it's more one decade than two because of the emergence companies built on AI but in any case, it will be more than two to three years).

AI may trigger a surge of new custom-built software

There is also an alternative scenario to the one outlined by Satya Nadella where AI accelerates the release of new software. This scenario was described most eloquently by Scott Belsky in his newsletter “Implications”:

DIY software will revolutionize apps for consumers AND the enterprise. There has been much discussion of AI code reviews, GitHub co-pilot, and no-code application builders for the enterprise, but what are the implications of agent-assisted software development for consumers? Quick apps for your home or family were too hard to build until now. I think we’ll see some pretty remarkable and super niche software applications emerge in 2025, by and for consumers. And in the enterprise, the cost calculation of building your own internal tools will start to merit AI-made homegrown solutions to workflows and enterprise functions (and increasingly agents will replace these functions, per the last forecast on the list!) as opposed to the usual “find a SaaS product to solve every need.”

In this scenario, AI agents trigger a surge of custom-made apps that are built with just one customer in mind.

While it is hard to predict how the market for software will develop, it is easy to see that the change will be significant: If AI agents become the primary point of interaction with users, developers will have to reimagine the architecture of their software and tailor them specifically to AI agents. Expenses for example may be filed easily just by taking a photo of a bill without the user having to open an app.

This however means that certain parts of a software business will become less relevant. As Azeem Azar puts it in a recent episode of "Exponential View":

I think that you’ve got a classic innovator’s dilemma if you are a traditional SaaS company, because the way in which you build, the way in which your products work, the way in which your teams are organized doesn’t necessarily sit nicely with an AI model.

On the other hand, larger SaaS firms will double down on their investments in well-designed user interfaces (let’s face it: the chat interface it not the end of UX design for AI). And many SaaS firms have already done a pretty good job in integrating AI and will continue doing so.

Complex regulation increases uncertainty among businesses

But what does this mean for the Commission's political agenda?

AI agents will play a large role in closing the productivity gap between the US and the EU. At a time where talent is scarce, AI bots will allow companies to build a “limitless workforce”, according to Salesforce CEO Marc Benioff.

But in order to do so, they will need to understand how AI agents fit into the existing digital rulebook of the EU.

In his interview, Nadella already pointed out some of the regulatory issues related to AI agents: Maintaining cybersecurity is naturally a key concern, shortly followed by questions around how access to these various apps by an AI agent should be governed. Should a company be allowed to restrict access for agents that for example are not build by them? If they do, could this be considered anti-competitive behaviour?

On a related note, it remains to be seen whether the emergence of AI agents shifts where value is created in the software market and how this shapes market dynamics in the software industry.

The more complex questions will resolve around how the deployment of AI agents will interact with existing laws such as the GDPR, cybersecurity regulations or the Product Liability Directive. Without a doubt, the software value chain will become more complex and it will become more difficult to assign responsibilities to individual economic actors.

Finally, while the Data Act mostly does not cover SaaS, its main idea of making data accessible is highly relevant for agentic AI. The law, however, only comes into force later this year and it remains to be seen whether it is even fit for purpose for the Internet of Things for which the law was originally created for ...

Reducing regulatory complexity will be crucial for the Commission’s goal to increase the uptake of advanced technologies such as AI agents. Next week, the Commission plans to announce its “Competitiveness Compass”. This will be indication of how serious the new Commission is about streamlining the proliferation of digital rules.

📚 Read on

How Satya Nadella remade Microsoft and manoeuvred the tech giant into a pole position for the AI race.

Radical deregulation in the US risks sucking business investment away from Europe much as subsidies did under Joe Biden’s Inflation Reduction Act.

The EDPB’s opinion on processing of personal data by AI models maintains some “strategic ambiguity” but this leaves companies at the mercy of how regulators interpret this opinion.

Why the Product Liability Directive’s categorisation of software as a product rather than a service may be reduce the competitiveness of European firms.

AI reshapes what computing means – interesting podcast on AI and SaaS with Ben Thompson and Andrew Sharp.

As the year comes to an end, I am looking forward to a few quieter days in front of the fireplace with a book in my hands. If you’re the same and you need some inspiration about what to read, here are a couple of books that I heartily recommend.

“Atomic Habits” by James Clear

I am admittedly late to the party, having read “Atomic Habits” (originally published in 2018) only this year. It is one of these books that makes you think “right, I get the idea” when reading the back cover; however, there is a lot more to discover.

According to James Clear, personal and professional success is built on the tiny habits that we develop over time; it is not ambition that makes us successful (naturally, every team has the ambition to win a basketball game for example), but how we go about getting there. Or in Clear’s words: “You do not rise to the level of your goals. You fall to the level of your systems.”

To me, starting to imagine who I want to be and then thinking about which habits this person would deploy was a significant change in how I think about personal development.

“Co-Intelligence” by Ethan Mollick

“Co-Intelligence” is hands-down the most important book on AI that was published in 2024. Fueled with practical insights from Ethan Mollick’s own experience with AI, the book proposes four pragmatic rules for adjusting to a world that will inevitably be infused with AI in the coming years. Far from the exhausting antagonism of “boomers” and “doomers”, these rules include experimenting with AI whenever we can, always being mindful of the shortcomings of current AI systems and assuming that the AI we interact with today will be the worst AI that we’ll ever see.

In the second part of the book, Ethan gives practical examples of how AI can be used in our lives, for example for creation, task automation, or as a coach and tutor (a vastly underestimated and yet already quite developed area for AI application). Its practical relevance and accessibility makes “Co-Intelligence” a compelling and interesting read for beginners and even advanced users of AI and sets it apart from many of the other books on AI on the market. A must read.

Ethan also publishes a Substack called “One Useful Thing” that I also recommend to read!

“EU Superlobby” by Milos Labovic

“EU Superlobby - Winning in Brussels” is a concise and practical guide for everyone who has an interest in understanding and influencing policy making in the European Union. Having worked in Brussels for years, Labovic not only knows how things should work but also how they actually work. He also discusses how important stakeholders can be reached most effectively, whether through direct engagement, media or events. Essential reading for everyone in the “Brussels bubble” and for everyone who has to deal with it at least sometimes.

Milos also writes the “Superlobby Community” newsletter which is a great resource for all politicos – you can subscribe here for free!

“Power and Progress” by Daron Acemoğlu and Simon Johnson

Already published in 2023, I still want to recommend “Power and Progress” as this year, Daron Acemoğlu and Simon Johnson have been awarded with the Nobel Price for Economics. In this groundbreaking study, Acemoğlu and Johnson argue that technological progress is closely linked to social power and that this power balance decides how technology is used and how its benefits are distributed. To solve this dilemma, the authors explore ways how technology can be used to increase worker productivity rather than to replace humans and to strengthen the voice of those that are impacted by technology in democratic debate. A topic that could not be more timely.

“The Strongmen” by Hans Kribbe

“The Strongmen” is an intriguing and extremely relevant analysis for everyone interested in foreign affairs. As authoritarian leaders such as Donald Trump, Vladimir Putin, Xi Jingping and others come to power across the globe, the European Union faces a fundamental challenge: The EU is built on norms of international cooperation, consensus and multilateralism – values that the strongmen explicitly do not follow. This puts the EU at a significant disadvantage when dealing with these leaders.

In the first part of the book, Hans defines what he means by “strongman” and what makes them so different to the European heads of state. In the second part, he proposes strategies how the EU can effectively deal with strongmen leaders, based on experiences during past encounters with them. With the second Trump administration coming in in 2025, this book provides European leaders with a useful, but equally sobering, playbook for the next years.

To understand what the new Trump administration means for EU tech policy, I interviewed Hans earlier in November. You can read the interview here.

Some more books!

Two other books I read, Amitav Ghosh’s “The Nutmeg’s Curse” and Jenna Odell’s “Saving Time” both explore in different ways how humankind has exploited the planet in the pursue of profits and growth at least for four hundred centuries now. Both authors pledge that we need to re-connect to nature and learn from indigenous people that have never lost touch with earth as we have. Two books that really made me think.

And while I don't read a lot of fiction, “All Fours” by Miranda July was an entertaining, intriguing and funny read, a “coming of age” story of a middle-aged women who suddenly founds herself questioning her entire previous life and embarks on a journey to find herself again.

What have you read this year? Please do share your recommendations in the comments!

Elon Musk featured prominently in the parliamentary hearing with Henna Virkkunen, the designated Commissioner for Tech Sovereignty. The EU and the controversial tech billionaire have a long history of conflict: Former Commissioner for Digital Thierry Breton clashed repeatedly with Elon Musk over the lack of content moderation on X, for example.

On several occasions, Members of the European Parliament indicated that the incoming Trump II administration required the EU to develop a European tech stack and decrease dependency on the US.

To really understand how the political dynamics related to the EU’s tech policy may play out in the new Trump administration and in particular with Elon Musk having a set at the President’s table, I spoke to someone who wrote a whole book about how “strongmen” like Donald Trump and Elon Musk think: Hans Kribbe, the founder and director of the Brussels Institute for Geopolitics and co-founder of Shearwater Global.1

Hans spoke eloquently about how security policy and economic policy become ever more intertwined in a “strongmen world” and that in order to compete, Europe needs to ensure that it creates the economic incentives for companies to thrive and survive. Without these basics in place, he says, Europe is unlikely to achieve technological sovereignty any time soon.

Read the full interview with Hans below; the interview has been edited for length and clarity.

In your book, you describe how the EU increasingly has to deal with leaders that you call “strongmen” such as Donald Trump, Vladimir Putin or Xi Jinping. Can you briefly explain what you mean by “strongmen”? How does the worldview of a strongman differs from the worldview of a typical Eurocrat?

In modern states, government is highly institutionalised and also bureaucratised: We have government agencies which are put in charge of specific tasks, we have our legal system, we have presidents and prime ministers and parliaments. And they’re all connected to rules and laws that delimit and codify these powers.  

The strongmen essentially do not wish to govern through this institutional framework, preferring to act through informal, parallel systems or a “shadow state” based on personal loyalties of friends, oligarchs and family. And this network is of course invisible and outside the purview of public scrutiny. 

How do strongmen conduct foreign policy? While the European Union is built on the idea of multilateralism, it seems that strongmen are realists when it comes to foreign affairs, is that correct?

Yes, I do think they are realists. We in Europe love multilateralism, international organisations and international law because in this way our interests are protected on the global stage. The strongmen’s approach to diplomacy is very different. They reject international bureaucracy and wish to conduct diplomacy at a much more personal level – leader-to-leader, man-to-man, as Trump put it once. 

You’ve written your book during the first Trump presidency. What have we learned about how he governs?

We know that Trump is deeply transactional. This will not change. He goes into high-level international meetings to make deals that benefit the United States. America first, that is his only principle. Transatlantic ties and so-called “special relationships” have no intrinsic value to him. 

He generally feels that the United States has been getting a very bad deal with NATO – “the steal of the century”, he calls it. But also in all kinds of other international fora. The WTO for example. This is why he wishes to renegotiate America’s relationships with the main players in the world, such as China and Europe, for example on trade.

In addition to the deal, Trump also craves respect. He needs to be told how great and smart he is. Europeans looked down on Trump as somebody who was not sophisticated or literate enough to understand how trade really works. They felt they needed to educate and enlighten him. But we learned that this petulant approach is not going to pay dividends.

What does work is giving him lots of respect, rolling out red carpets, slapping him on the back in front of the cameras. In his first term, the Queen invited Trump for a state visit to Buckingham Palace. He loved it.  

If Trump believes that America got a bad deal and he wants to change this deal, this analogy works well for international trade agreements. But during the European Parliament’s hearing with designated Commissioner Henna Virkkunen, a lot of questions evolved around the newly acquired power of Elon Musk and whether he may use it to dodge some of the EU rules such as the Digital Services Act (DSA). Would you share this fear?

Yes, this is a serious danger. When Trump talks about redefining trade relationships he is not just talking about tariffs, but also about how we regulate US players, including Musk’s X.

It is unlikely that Trump himself has clear ideas about the DSA and how we regulate tech in Europe. But we do know that Elon Musk has Trump’s ear. We also have the comments of J. D. Vance, the vice president-elect, who explicitly said that if Europeans want to “censor” and fine X under the DSA, it stops making sense for the US to continue to guarantee Europe's security in NATO.  

I am not sure this possibility has fully sunk in in Brussels and other European capitals. But if this is the position the Trump administration takes, little remains of our freedom to set our own regulatory policies. With the war going on, European leaders would be tempted to let X off the hook, water down the Digital Services Act or find creative solutions for a negotiated settlement that Musk can live with it. He now holds the upper hand, not the European Commission. 

So you would expect that the incoming US government will be much more proactive than the Biden administration in protecting the interests of American companies abroad and that this engagement will be tied to completely unrelated policy areas such as defense?

I think we will have to wait and see. I don’t think tech regulation in Europe is very high on Trump’s agenda. He seems to care more about other sectors such as cars, energy and steel. And I also think that some tech companies would be hesitant to lobby the White House to play the security card in this way, which on the long term could clearly also backfire. But we obviously need to be aware this risk exists.

On the second point, economic and security policy are of course already linked. Just look at China and European supply chain dependencies in areas such as the green economy. Or look at Russian gas and oil. Why should Trump not use the security leverage he has over Europe, including by threatening to abandon Article 5 of NATO? 

Clearly, the fear of dependence on the US is top of mind for EU policy makers at the moment. This is also reflected in Henna Virkkunen’s mission letter. When it comes to tech, she has two big tasks ahead of her: One is to accelerate the adoption of new technologies so that we can close the productivity gap between the US and the EU. And the other is to build a European tech stack. The question for me is what is the right order here? Building a European tech stack takes a lot of time and money. And in the end, the result is not necessarily better than what is already on the market. Do you think that increasing Europe’s tech sovereignty is really the only option that we have? Or is there common ground that we can find with the Trump administration? 

We will continue to rely on US technology in the foreseeable future because we have no European alternatives. And I don’t see the biggest threat to European sovereignty arising from the US threatening to withhold certain digital technologies and services from the European market. For companies such as Google, Apple or Microsoft, Europe remains a big market, so such a move would come at a significant cost to the US.

I think the bigger risk lies in the United States using our security dependency to influence and shape digital and other economic policies in Europe, for example in how we trade with China.

I do agree Europe needs to invest in its technological sovereignty. But we have to be realistic here. There is a huge technology gap between the EU and the United States and also China. It takes massive investment to close this gap and the economic conditions for attracting such investment just don’t exist at the moment.  

Without a sound economic basis, achieving greater industrial autonomy or resilience is going to be hard in any sector. Take the Swedish battery factory Northvolt, meant to offer a European homegrown alternative to Chinese-made batteries. Geopolitically, it was the right move. But today the company faces bankruptcy. Calling for greater tech sovereignty and European champions makes sense politically. But you also need to be able to make the business case, and for now that is not always easy.

For the time being I believe Europe has little choice but to rely on the US tech industry. Yes, we do need things such as Europe-based data centres and cloud infrastructure, but we’re not going to build them without collaboration and joint ventures with US tech companies.

📚 Read on

In this article (in German), Ansgar Baums and Nicholas Butts argue that export controls for technology are often more expensive than assumed and may even strengthen your enemies.

In this essay, Sujit Raman seeks to find common ground in approaches to digital sovereignty, privacy and cybersecurity between the EU and the US.

EU regulations make it harder for firms to do “radically new things”, Pieter Garicano writes and this makes is extremely difficult to build “superfirms” in Europe.

WHEN Henna Virkkunen, the designated European Commissioner for “tech sovereignty” will be quizzed by the European Parliament tonight, she has little to fear; Members will not reject her nomination. The challenge for Virkkunen will only come afterwards, as her mission letter reveals a strategic dilemma.

In her political guidelines, Commission President Ursula von der Leyen wrote that the “insufficient diffusion of digital technologies” is one reason for the lack of competitiveness of the European economy (a finding that is widely echoed by economists such as Isabel Schnabel and Mario Draghi). But the EU also wants to increase its “strategic autonomy” by building European infrastructure.

What’s first: diffusion or development of digital technologies?

The order of things is important: Do we want to spend the next decade building tech infrastructure or accelerate growth by increasing the uptake of cloud and AI across the economy?

Both are important obviously, but Europe’s opportunity lies with the latter.

Today more than ever, speed is essential and growth does not necessarily come from the invention of the best AI model but from a smart adoption of this technology in research and development and in business. Just like retailers who ignored e-commerce in the early 2000s are today challenged by direct-to-consumer brands, companies that do not embrace AI now will be replaced by AI in the future. The new superstar firms will be built on AI.

In her response to the European Parliament, Virkkunen wrote:

Our goal is to develop technologies that empower and enhance human capabilities and deliver on European values. This can be a competitive advantage for “Made in EU” digital products and services.

No one can possibly be against “safe, trustworthy and human-centric technologies”, but it is hard to believe that digital products and services “Made in EU” have a competitive advantage, especially because it is unclear what “European values” actually mean in the context of emerging technologies.

The excruciating debate about EUCS, the EU’s cybersecurity certification for cloud services is a case in point: the ongoing uncertainty poses a risk especially to smaller firms who may not have the expertise to accurately assess the cybersecurity posture of their vendors.

Will the EU merely “survive digitally” or embrace the opportunity?

The United Kingdom has a more pragmatic approach: The government of UK Prime Minister Keir Starmer believes that the biggest AI opportunity for Britain is to focus on the application layer with regulation that encourages the uptake of AI across the economy. British entrepreneur Matt Clifford, who led the review of AI opportunities for the British government, is due to present his “AI Opportunities Action Plan” later this month.

This approach is what the European People’s Party’s Axel Voss seems to favour as well: Speaking to POLITICO, he said: “Tech sovereignty doesn’t mean digital revolution”. After a mandate full of new tech regulations, Europe now has to think about how it can “survive digitally” in a fast-moving world, he added.

Re-defining Europe’s digital policy mission

Instead of building regional tech barons such as a European search engine or a European large language model, Europe should have the aspiration to build global empires based on the application of emerging technologies.

Mariana Mazzucato shows a way to think about how this could be achieved. Visionary, energetic and unconventional, the Italian is one of Europe’s most admired economists who embodies what Europe aspires to be.

In her book “Mission Economy: A Moonshot Guide to Changing Capitalism”1, Mazzucato writes that government should not just be about fixing problems (with regulation), but about creating public value by launching “missions” that can be very specific (like sending a man to the moon) or centre around shifting attitudes.

In terms of digital policy, Europe's mission should be to create frameworks to enable companies to scale successfully. While the EU is ‘good’ at reigning in on Big Tech and increasingly also on supporting small start-ups, it is a bad place for companies to scale due to a lack of finance and high costs of compliance.

For this reason, it is encouraging that von der Leyen has promised to introduce a new category of “small midcap” companies and to assess where existing regulation for large companies creates disproportionate burden for smaller firms.

Secondly, rather than putting money into some companies or sectors, hoping that they’ll succeed, the EU should think about the digital transformation in much broader terms than just IT and hardware or cloud infrastructure. A pharmaceutical company that develops a new drug using a large language model or a construction firm that discovers a lighter construction material with AI is as much a digital company as a cloud service provider.

Thirdly, there is evidence that the General Data Protection (and probably not so much the actual law but subsequent guidelines and court rulings) have a chilling effect on innovation and venture capital flowing into the EU. In the future, we will probably see similar effects from other laws such as the EU Data Act. Complexity and legal uncertainty reduce appetite for innovation, especially in the light of high financial risks. As the new Commission is aiming for a faster and simpler Union, it needs to review laws that need to be simplified, streamlined or abolished.

Without adoption, there is no innovation. To get to the technological frontier, every firm in Europe needs to become a digital company.

📚 Read on

Europe lacks competitiveness compared to other regions because of a lack of adoption of new technologies and overregulation, according to ECB board member Isabel Schnabel.

Angela Strange, General Partner at VC firm Andreessen Horowitz believes that AI agents can significantly increase a company's productivity and create competitive advantages for early adopters.

Andrew Bennet of the UK VC form Form Ventures has some insightful thoughts on AI policy for the application layer.

Mario Draghi correctly identifies the origins of Europe's slowdown in productivity, but draws the wrong conclusions, finds Bruegel's Bertin Martens.

Former MEP Luis Garicano writes that since GDPR's entry into force, web traffic went down, firms became more concentrated and EU start-ups closes fewer venture capital deals.

In her bid to get confirmed as President of the European Commission, Ursula von der Leyen claimed that the “insufficient diffusion of digital technologies” is one reason for the lack of competitiveness of the European economy. Yet, the announcement of both Apple and Meta to delay the launch of their AI features in Europe due to regulatory uncertainty signal that the diffusion of the next generation of digital technologies will not happen more swiftly.

Quickly, two different interpretations emerged: The first stated that the EU’s regulatory zeal is a risk to the continent’s competitiveness and the second that the rules of the game in the EU are clear and that companies simply do not want to comply.

Europe's decade of regulation

Not surprisingly, EU regulators and lawmakers primarily belong to the latter camp. This is former Vice President of the European Commission Viviane Reding:

Importantly, it seems that Meta’s comments about regulatory uncertainty did not refer to the newly enacted EU Artificial Intelligence Act (AI Act), but the block’s privacy rules. The company recently complied with a request to stop training its models on data of European users, but according to Meta’s VP of Policy Rob Sherman, the company would not “be able to serve European consumers properly” without the ability to train on European data.

The uncertainty of EU tech rules

Even people involved in the negotiation of the AI Act like Kai Zenner have criticised that during the final 36-hours negotiation marathon, “Time pressure led to an outcome where many things remain open. Regulators couldn’t agree on them and it was easier to compromise. It was a shot in the dark”.

This includes guidance on the use of personal data for AI training. In recent months, Data Protection Authorities across Europe have published position papers and guidances that range from liberal (like this position paper) to extremely restrictive. Given the fines attached to violations of the EU’s General Data Protection Regulation (GDPR), the financial risk is big, especially for large firms.

Benedict Evans explained the dilemma nicely in his newsletter (No. 550, 23 July 2024):

There is a narrow story here about highly specific legal issues, but the broader issue is that the EU has written some very broad and complex laws that have a lot of scope for interpretation and ‘creep’ over time, so that companies might think they’re complying with the law, but then get a billion dollar fine they weren’t expecting, and a demand to redesign their proposition. I don’t remember anyone reading GDPR and saying that this would oblige Meta to provide its services to the EU without any interest-based advertising, for example, yet that’s where the EU has ended up. One of the tensions of regulatory practice is between setting general principles (so you don’t need thousands of individual laws) and predictability and standardisation, and the EU might have swung too far.

Is the “Brussels effect” losing its grip?

This has two potentially dramatic consequences: If companies continue to delay (or even cancel) the launch of their most powerful AI models in Europe, European firms will have to work with lower quality models and won’t be able to reap the benefits of AI to the same extent as other countries can. This will further increase the GDP gap between the US and Europe.

And European companies may not be able to fill the gap: Elisabeth L’Orange, co-founder of AI firm Oxolo recently announced that her company pivoting because of regulatory challenges (Oxolo’s AI generated videos would be labelled as deepfakes under the AI Act) and a slow adoption of new technologies in European companies.

Secondly, it seems that recent years have shown that a tougher macroeconomic environment and high compliance costs might have clipped the wings of the “Brussels effect”, i.e. the assumption that stricter European regulatory norms will be adopted globally.

It is often assumed (especially by EU officials), that the global adoption of European norms is a more or less automatic process, but that was of course never the case.

The Brussels effect is tied to Europe's economy power

First of all, the rules need to be clear: The EU often claims that they are “open for business” for every company that wants to be present in Europe and adheres to local laws. But when the law is unclear, there can by definition be no “Brussels effect”.

Companies may also think they are forced to “fork” their software and produce a “localised” version solely for European users. This however fundamentally changes the economics of tech businesses.

In his newsletter Stratechery, tech analyst Ben Thompson has repeatedly pointed out that firms like Meta earn significantly less money per user in the EU compared to the US and yet, the most important European tech laws calculate fines not on the basis of European, but global turnover:

This is the point, though: the E.U.’s worldwide regulatory power is ultimately derived from the structure of technology, and structures can change. This is where that ten-percent-of-worldwide-revenue figure looms large: it fundamentally changes the calculus in terms of costs. Fines from a regional regulator are not the equivalent of engineering and server costs that you’re already paying, so you might as well capture pennies of extra revenue from said region; they are directly related to revenue from that region. In other words, they are more like marginal costs: the marginal cost of serving the E.U. is the expected value of the chance you will get fined more than you earned in any given year, and for big tech that price is going up.

If you add the costs of forking (and maintaining) an EU version for your product, coupled with a smaller return per user and the potential of fines based on your global turnover, it is not difficult to imagine that companies may think twice before launching a high-end product in Europe in the future.

Europe needs to do better: A recent analysis by the Centre for European Research (CER) finds that the Brussels effect might still prevail as long as EU regulations “focus on benefiting consumers, keeping Europe’s markets open and not forcing firms into EU-specific business practices”. Europe needs to find better ways to regulate and it cannot source out its competitiveness to regulators — it needs to become more innovative on its own.

On this day nineteen months ago, OpenAI published a blog post: "Introducing ChatGPT".

The article doesn't read like a launch post; it reads more like an announcement at the blackboard of the Department of Computer Science. But ChatGPT went trough the roof and acquired new users at a speed like no other software product before. The rest is history.

Attention is not all you need

Still, not only tech pundits like Ben Evans wonder what the use case of AI is. As it turns out, attention is not all you need.

But that's ok. There also wasn't an obvious use case when Gutenberg invented the printing press at a time when almost no one could read. And Thomas Watson, President of IBM, famously said that he expected the world market for computers to be “five". Still, today I own at least four computers and have access to vast amounts of computing power provided through the cloud.

Welcome to AI's "floppy disk" era

In his insightful book "Co-Intelligence"1, Ethan Mollick asserts that today's AI will likely be the least advanced we'll ever see. We are, in essence, in the "floppy disk" era of AI

We may soon be using more than one model or AI system, each of which will be designed for specific tasks or grounded in your own data such as in Anthropic’s “Projects” feature. When we don't have to rely on one system but can pick and choose the system best suited for the task at hand, the usefulness of AI will increase significantly.

Understanding AI’s “Engel’s Pause”

Innovation rarely occurs in a single "Eureka moment." It unfolds when technological advancements align with changes in our economy and in society. The true value of the printing press is not the machine itself but that it enables authors to write, publishers to publish and book stores to sell. Institutions like universities couldn't exist if the printing press was never invented.

Carl Benedikt Frey, in "The Technology Trap”2, notes that it took a generation for the steam engine's impact to reflect in growth statistics. This lag, where a new technology exists but its economic value isn't immediately realized, is termed "Engel’s Pause," named after Karl Marx' collaborator Friedrich Engels.

That does not mean we have to wait for another thirty years until we see our economy growing through AI: In a recent edition of his newsletter, Ethan Mollick shared that researchers asked the public to solve problems with the help of large language models that even the most advanced models could not solve yet (or so they thought). Still, smart users were able to use the models in ways that allowed them to solve the challenge within a day.

In other words: In the case of AI, the Engel's Pause may be shorter than ever before in the history of mankind.

Hidden use cases of AI

We tend to recognize only use cases that are immediately apparent: It is obvious that a smart phone is superior compared to a Windows phone or a Blackberry. However, when the iPhone was launched, nobody predicted that a GPS connected handheld device would lead to the employment of hundreds of thousands of people driving cars for Uber. That is the type of innovation that general purpose technologies enable.

AI is likely to follow a similar trajectory. Today, almost every CEO touts their company's "AI strategy." But in six to ten years, AI might be so embedded in daily life that we take it for granted, much like hailing an Uber has become second nature. Life doesn't need a use case.

European news portal Euractiv writes that regulators in Europe become increasingly concerned about Twitter’s ability to comply (or not) with the Digital Services Act, the EU’s flagship regulation for online safety. Yet it seems they draw the wrong conclusions from it:

If Twitter, by corporate decision or negligence, ends up pulling out of the EU market, it would be an important first time. (…) Still, tech companies seem increasingly making cost-benefit analyses on whether Europe’s regulated market is worth the trouble, increasing the importance of establishing EU rules as the international standard via the ‘Brussels effect’.

The last sentence hints at a severe misunderstanding of how the “Brussels effect” works.

What is the “Brussels effect” – and what it is not?

The Brussels effect is described in Anu Bradford’s book (2020) of the same title. In essence, Bradford argues that in many cases, European rules and regulations become de-facto global standards because of the importance of the European market. It is often simply easier to roll out European standards globally, especially if they are higher than other national standards.

But the Brussels effect is also an example of a concept that seems simple and easy at first sight and is therefore often misunderstood because the reality is more nuanced.

It' it just like the butterfly effect: just because the flap of a butterfly’s wings in Brazil may set off a Tornado in Texas does not mean that it will do so. Likewise, not all European laws become a global standard.

In fact, Bradford herself highlights a few preconditions that must be met so that the Brussels effect kicks in, namely a large enough size of the European market for a specific product, the ability to enforce rules and impose sanctions, the political will to regulate, a lack of ways to “circumvent” regulation (e.g. by re-directing capital flows), and an inability to customize products or services to different regulatory standards.

In reality, therefore, the Brussels effect requires a complex set of conditions to come into effect. This explains why the Brussels effect is clearly visible in the privacy world: The EU’s General Data Protection Rregulation is a horizontal regulation that applies across industries (i.e. the market is huge). On the other hand, the Brussels effect is less likely to emerge automatically in online safety bills such as the Digital Services Act where cultural norms (i.e. the “political will to regulate”) play a bigger role.

Why regulation does not always translate into a competitive advantage

Which brings us back to the original claim of the Euractiv article: If a company’s cost-benefit analysis concludes that compliance with European regulation does not lead to a return on investment, the Brussels effect becomes irrelevant. Only when Europe is an attractive market and companies want to sell (or produce!) their goods and services here, companies have an inventive to comply with European regulation. If compliance costs become too high, companies will simply leave for more attractive markets.

That’s the message that a representative of German chemicals giant BASF recently shared with the Financial Times: “The regulatory burden that’s building up might be manageable for global players but I don’t know how a midsized company of 100-200 people is supposed to digest it“.

Despite what some European lawmakers may think: the Brussels effect is no regulatory silver bullet. As beautiful as the European butterfly may be, the flap of its wings does not always create a regulatory storm.

Thanks for reading 🎪 Techleash! Subscribe for free to receive new posts and support my work.

Was heißt „beautiful business“ im Zeitalter der Automatisierung und der Künstlichen Intelligenz, in der „Superstar“-Firmen mit purpose werben aber Shareholder-Value meinen?

Nach vier Tagen Diskussion und dem Singen von Kampfliedern im House of Beautiful Business in Lissabon kristallisieren sich zwei sehr unterschiedliche Antworten auf diese Frage heraus.

When you get to start your day by singing fight songs in the most beautiful library at the @_houseofbb pic.twitter.com/G6lVcik5By

— Samira Zippel (@SamiraZippel) November 5, 2019

Die erste Denkschule ist überzeugt, dass unser Wirtschaftssystem kaputt ist.

„Wenn ich Mark Zuckerberg wäre, würde ich Facebook in ein öffentliches Infrastrukturunternehmen umwandeln und die Shareholder müssten sich dann einfach damit abfinden“, meint etwa der Autor und Tech-Kritiker Douglas Rushkoff. (Digital-) Unternehmen müssten zu Genossenschaften werden, weil börsennotierte Unternehmen nicht in der Lage seien, den Shareholder-Value-Kapitalismus zu überwinden (und ja, es gibt natürlich gute Beispiele dafür, dass Kooperativen beautiful bss sind).

Douglas Rushkoff tells @_houseofbb #House19 how every business should become a coop! @rushkoff #postitoftheday pic.twitter.com/iUQUi4TxlY

— John Monk

(@johnmonks) November 5, 2019

Die zweite Denkschule wird durch „leidenschaftliche Pragmatiker“ wie den Ökonomen Sir Paul Collier repräsentiert. Collier zufolge ist das Ringen um eine neue Form des Kapitalismus der größte Kampf unserer Zeit. Neue Konfliktlinien zwischen boomenden Metropolen und verödeten ländlichen Regionen sowie zwischen gut ausgebildeten Fachleuten und wenig gebildeten Hilfsarbeitern beschreiben die zentralen Herausforderungen vor denen wir stehen.

„Ich glaube fest an den Kapitalismus“, meint Sir Paul. „Aber wenn man ihn auf Autopilot laufen lässt, gerät er aus der Bahn.“ Niemand wache morgens auf und denke sich: „Heute möchte ich den Shareholder Value erhöhen!“

Dabei helfe bloße Wut nicht bei der Lösung der Probleme, so Sir Paul, vielleicht mit Verweis auf den Vortrag von Anand Giridharadas am Vorabend. Giridharadas hatte kritisiert, dass der Diskurs über die Transformation von Wirtschaft und Gesellschaft ausgerechnet von denjenigen Eliten geführt würde, die am meisten zu verlieren hätten. „Das ist, als ob man den Schutz der Hühner an den Fuchs delegiert“.

Demgegenüber betonte Sir Paul: „Von Ideologie getriebene Revolutionen führen nicht nach Utopia. Wir wissen, wohin sie führen.“ Der Autor des Buches „The Future of Capitalism“ (Amazon-Link) propagierte stattdessen eine Reform des Kapitalismus.

Rushkoff, Sir Paul und viele andere Gäste des House of Beautiful Business waren sich einig in ihrer Kritik an den großen Technologiekonzernen, ihrer Dominanz im öffentlichen Raum und ihrer Kontrolle über unser Leben.

Doch trotz der vielfach berechtigten Kritik sollten wir uns vor dem Zynismus schützen, den Giridharadas versprühte („Ich glaube nicht, dass jemand, der eine App zum Bewerten von Frauen an der Uni erfunden hat, soviel Kontrolle über unseren Diskurs haben sollte“).

Dass Unternehmen eine erhebliche Kraft zur Veränderung haben zeigen Beispiele wie die dänische Energiefirma Ørsted, die ihren CO2-Ausstoß bis 2023 um 96 Prozent reduzieren wird. Ein purposeful capitalism ist also möglich und Veränderungen können auch durch und in Unternehmen geschehen und nicht nur auf der System-Ebene.

Auf dem Weg zu beautiful business ist es nicht nur der alleinige Fokus auf Shareholder Value, der uns bremst, sondern auch eine Form des Zynismus, der nicht daran glaubt, dass auch große Organisationen sich wandeln können – vielleicht nicht alle und vielleicht nicht sofort, aber hoffentlich genügend und schnell genug.

„Wir brauchen die Menschen in Organisationen und den Druck von Außen“, fasste die Journalistin Georgia Frances King die Diskussion in einer der letzten Sessions mit dem Titel „Falling in Love with Tech Again“ zusammen.

Und Sir Paul beendete seinen Auftritt im House of Beautiful Business mit dem Satz: „Bitte, seien Sie Botschafter des Wandels. Ich bin verängstigt über das, was ich im Internet sehe, aber wir brauchen das Netz!“

Best of House19

Das Haus of Beautiful Business hatte eine Menge spannender Residents zu bieten, hier nur eine kleine Auswahl von Vorträgen, die mich besonders beeindruckt haben:

1. Die Designerin Leyla Acaroglu hat darüber gesprochen, wie man mit besserem Design eine nachaltigere Welt schaffen kann.

2. Sir Paul Collier, „steward of capitalism“.

3. Die Ingeneurin Dara Dotz, die cutting edge-Technologie zu den Menschen bringt, die damit am Meisten anfangen können (Foto).

4. Der Extinction-Rebellion-Aktivist, Slow Traveller und ehemalige Berater Ed Gillespie mit seinem flammenden Plädoyer für den Baum (Foto).

5. Ebru Köksal, die eine sehr persönliche Geschichte mit uns teilte (Foto).

6. Jess Kutch, Gründerin von Coworker.org mit spannenden Einblicken in die Gewerkschaftsbewegung in den Vereinigten Staaten.

7. Mathieu Lefèvre, einer der Gründer der NGO More In Common, die kürzlich auch in Deutschland gestartet sind.

8. Mariana Lin, die uns daran erinnert hat, dass Künstliche Intelligent genauso merkwürdig sein sollte wie wir Menschen es sind.

9. Jennifer Petriglieri, Autorin von „Couples That Work“ (Amazon-Link).

10. Carolin Ramade, die für mehr Diversität in der Tech-Branche streitet.

Im Jahr 2013 veröffentlichten die Oxford-Wissenschaftler Carl Benedikt Frey und Michael A. Osborne eine der am meisten zitierten Studien der letzten Jahre zur Zukunft der Arbeit: „The Future of Employment: How Susceptible are Jobs to Computerisation?“ (PDF). Die Studie, die – vereinfacht gesagt – prognostozierte, welche Jobs am ehesten durch Automatisierung bedroht werden, führte zu einem weltweiten Echo und passte perfekt zur notorischen Zukunftsangst der Deutschen.

Nun hat Carl Benedikt Frey mit „The Technology Trap“ ein neues und sehr lesenswertes Buch herausgebracht, das die Angst der Deutschen vor Automatisierung durch Technologie in einen historischen Kontext stellt.

Zeitalter der Maschinenstürmer

Die erste Industrielle Revolution bracht erheblichen sozialen Unfrieden in Europa hervor. Doch während Maschinenstürmer und Ludditen in Frankreich und Deutschland erfolgreich die Einführung neuer Technologien verhinderten, stellte sich die britische Regierung auf die Seite der Industriellen. Die Briten wollten ihre Wettbewerbsfähigkeit nicht durch das Verbot neuer Technologien beschränken.

Hinzu kam, dass die Ständegesellschaft in Großbritannien wesentlich weniger ausgeprägt war als in Kontinentaleuropa und dass Meinungsfreiheit, Forschungsfreiheit und die freie Berufswahl dazu führten, dass sich neue Ideen leichter durchsetzen konnten.

Warum Karl Marx’ Prognose nicht Realität wurde

Trotzdem: Von der Industrialisierung profitieren zunächst ausschließlich die Industriellen. Während die Produktivität eines Arbeiters in der Zeit von 1780 bis 1840 um 46 Prozent stieg, wuchsen die Löhne nur um 12 Prozent. Dieser Zeitraum – in der neue Technologien Arbeit ersetzen, ohne das adequate neue Arbeitsplätze entstehen, bezeichnen Ökonomen als „Engels-Pause“, benannt nach Friedrich Engels, einem engen Wegbegleiter von Karl Marx.

Marx‘ Prognose einer Arbeiter-Revolution aufgrund der unmenschlichen Arbeitsbedingungen der Industriellen Revolution trat jedoch nicht ein. Im Gegenteil: Die Industrielle Revolution – das wissen wir heute – brachte einen erheblichen Wohlstandsgewinn für die breite Mehrheit der Bevölkerung, wenn auch erst mit einem Zeitverzug von einigen Jahrzehnten. Unsere heutige Konsumgesellschaft wäre ohne die Industrielle Revolution nicht möglich gewesen.

Die Industrielle Revolution brachte Wohlstand für Viele

Spätestens im 20. Jahrundert sorgte die Automatisierung dafür, dass die Gesellschaft gleicher wurde. Die Zeit zwischen 1900 und 1970 gilt in den USA als „die größte Nivellierung aller Zeiten“. Und trotz zweier Weltkriege und dem Kalten Krieg lässt sich dies auch in Europa für die Zeit nach dem Ende des Zweiten Weltkrieges sagen.

Diese zweite Phase der Industrialisierung – charakterisiert durch die Einführung des Fließbandes durch Henry Ford und durch die Elektrifizierung der Produktion – brachte im Vergleich zur ersten Industriellen Revolution weniger soziale Verwerfungen mit sich. Frey führt das darauf zurück, dass die entscheidenden Technologien dieser Zeit Arbeit nicht (wie der Webstuhl oder die Dampfmaschine) ersetzten, sondern unterstützen.

Elektrifizierung, Reorganisation und modernes Management waren alle Teil desselben Prozesses.

Carl Benedikt Frey

Aber auch die Gewerkschaften spielten in der zweiten Industriellen Revolution eine erheblich konstruktivere Rolle als die Gilden des 18. und 19. Jahrhunderts. Anstatt den technologischen Fortschritt zu blockieren, legte die Arbeiterbewegung des 20. Jahrhunderts ihren Fokus darauf, den technologischen Wandel zu begleiten – und dabei sicherzustellen, das die Interessen ihrer Anhänger – mehr Gehalt, weniger Arbeit und eine frühere Rente – ausreichend berücksichtigt wurden.

Zugleich wandelten sich nicht nur Produktionsprozesse, sondern auch die Natur des Unternehmens an sich: „Elektrifizierung, Reorganisation und modernes Management waren alle Teil desselben Prozesses.“ Und obwohl auch die zweite Industrielle Revolution mit massiven Arbeitsplatzverlusten verbunden war – etwa durch den graduellen Ersatz der Straßenbahn durch das private Automobil – ergaben sich zugleich neue berufliche Chancen für die arbeitslos gewordenen Straßenbahnfahrer – etwa als Trucker oder Arbeiter in einer der zahlreichen Automobilfabriken.

Kommt der „Techlash“?

Apokalyptische Szenarien zur Zukunft der Arbeit sind nicht neu. Nicht wenige glaubten zur Zeit der ersten Industriellen Revolution fest daran, dass die Produktion bald komplett automatisiert werde und dass die Arbeitsbedigungen der neuen Arbeitswelt die moralische und intellektuelle Entwicklung der Arbeiter beeinträchtige.

Frey glaubt jedoch nicht, dass technologischer Fortschritt verlangsamt werden sollte. Der Computer sei die ultimative Universaltechnologie, die in der Lage sei, jeden Aspekt unseres wirtschaftlichen Handelns grundlegend zu verändern. Und schließlich brachte auch die Industrielle Revolution langfristig gesehen einen noch nie dagewesenen Fortschritt mit sich.

Aber er mahnt dazu, die sozialen Konsequenzen der gegenwärtigen digitalen Transformation nicht aus dem Blick zu verlieren und wirtschaftliche Einbußen abzufedern. Sonst könne es – vergleichbar mit den Maschinenstürmern des 18. und 19. Jahrhunderts – zu einer öffentlichen Gegenreaktion kommen, mit der Folge, dass andere Nationen schneller voranschreiten und damit auch die ökonomische Wertschöpfung woanders stattfindet.

Die von Marx vorhergesehene Revolution ist nicht deswegen ausgeblieben, weil die Einführung neuer Technologien verhindert wurde, sondern weil Technologie begann, im Interesse der Arbeiter zu wirken und die Arbeiter sie zunehmend als ein zentrales Instrument für die Erhöhung ihres eigenen Wohlstands sahen.

Frey warnt jedoch vor einer zunehmenden Spaltung der Gesellschaft. „In einer Welt, in der Technologie wenige Arbeitsplätze und enormen Wohlstand erzeugt hat, ist die entscheidende Herausforderung die Verteilungsfrage“.

Die Herausforderung betrifft nicht nur Unternehmen

Einige Besprechungen des Buchs von Frey haben kritisiert, dass der Ökonom wenig Antworten auf die von ihm geschilderten Herausforderungen bietet. Das finde ich nicht. Allerdings sieht Frey nicht alleine Unternehmen in der Verantwortung, sondern schlägt unter anderem auch Lohnauffüllungen, die Liberalisierung von Flächennutzungsplänen oder die Abschaffung von Zugangsbarrieren durch Zertifikate in manchen Jobs vor, die einen ähnlichen Effekt auf den Arbeitsmarkt haben wie Gilden im Mittelalter.

All das klingt weit weniger sexy als eine Robotersteuer oder die Zerschlagung von Internetkonzernen – zeigt allerdings, dass die Herausforderung auch wesentlich komplexer ist als so mancher talkshowerprobte Philosoph uns glauben machen will – ein absolut lesenswertes Buch!

Foto: Robert Friedrich Stieler (1847–1908): BASF-Werk Ludwigshafen 1881, Gemälde im BASF-Archiv

Die Digitaldebatte in Deutschland ist in der Regel binär, wie auch der Rückzug von Robert Habeck aus den sozialen Netzwerken heute gezeigt hat: Dafür oder dagegen – drin oder draußen. Die ideologisch motivierte Polarisierung spiegelt sich auch in den Büchern wieder, die hierzulande zum Thema veröffentlicht werden.

Das Buch von Benedikt Herles ist anders. Es lässt sich nicht in ein Lager einordnen. Vor allem aber gelingt es dem Investor und Buchautor, dem man eigentlich unterstellen könnte, einen überwiegend wirtschaftlichen Blickwinkel auf das Thema einzunehmen, auch die politische und gesellschaftliche Dimension der digitalen Transformation in seiner gesamten Tragweite zu vermitteln.

Wie kann sich der Sozialstaat in Zukunft finanzieren?

Eindrucksvoll schildert Herles, wie neue gentechnische Methoden es ermöglichen, das menschliche Erbgut kostengünstig zu verändern, wie Unternehmen im Silicon Valley an der Unsterblichkeit arbeiten oder wie der Einsatz von Robotern und Algorithmen in der Wirtschaft zu einer zunehmenden Spaltung zwischen den Abgehängten und den Superreichen führt.

Man muss dem Autor dabei nicht einmal in jeder Hinsicht folgen; selbst wenn nur ein Drittel seiner Vorhersagen eintritt, bedeutet das einen fundamentalen Wandel in unserem gesellschaftlichen Zusammenleben.

Weil unser Sozialsystem etwa auf der Besteuerung von Lohn basiert, wirft der zunehmende Einsatz von Robotern und Künstlicher Intelligenz in Unternehmen die grundlegende Frage auf, wie unser Sozialstaat in Zukunft finanziert werden soll.

Die Entstehung des heutigen Sozialstaats ist eng mit der industriellen Revolution des 19. Jahrhunderts verbunden. Aber zu dieser Zeit vollog sich der technologische Fortschritt linear. Es dauerte Jahrzehnte, bis Automobile das Pferd als das dominante Fortbewegungsmittel ablöste und noch einmal so lange bis moderne Fabriken das Handwerk ablösten.

In der Politik betrifft das innovator’s dilemma uns alle

Heute leben wir jedoch in einer Zeit des exponentiellen Wachstums, frei nach dem Mooreschen Gesetz, wonach sich die Rechenpower von Computerchips alle zwölf bis 24 Monate verdoppelt. Die Politik, so Herles, ist dieser Dynamik schlicht nicht gewachsen und immer noch in einer Industriegesellschafts-Logik verhaftet.

Hier beginnt der wirklich spannende Teil von “Zukunftsblind”, denn Herles sieht die Politik in einer Art innovator’s dilemma gefangen: das, was erfolgreiche Startups ausmacht – Agilität, Flexibilität, Diversität – ist aus Sicht der Politik ein Risikofaktor. Stattdessen denken Parteien wie Konzerne: sie versuchen, Risiken zu minimieren, verpassen dabei jedoch viele Chancen. Kurz: ihr Erfolg in der Vergangenheit versperrt ihnen den Erfolg in der Zukunft.

Was ist der Wirtschaft aber lediglich ein Problem für Aktionäre und Mitarbeiter wäre, betrifft in der Welt der Politik das gesamte Gemeinwesen – also uns alle.

Eine neue Sichtweise auf Politik

“Zukunftsblind” ist das erste Buch, das ich in diesem Jahr gelesen habe – und zugleich meine erste Buchempfehlung. Die Parlamentsbuchhandlung sollte einen extra großen Schwung bestellen.