AI Bubble Warning from Arm Co-founder Hermann Hauser

This report is from this week’s The Tech Download newsletter. Like what you see? You can subscribe here.

Few investors possess the foresight to have navigated every major technological paradigm shift over the past four decades. Hermann Hauser, a figure synonymous with European innovation, co-founded Acorn Computers in 1978 and was instrumental in the creation of Arm, a company that revolutionized the semiconductor industry and remains one of the world’s most critical chip architects. Today, through Amadeus Capital, he actively mentors and funds some of Europe’s most promising deep-tech ventures.

During a recent conversation for The Tech Download podcast, our discussion spanned the transformative potential of artificial intelligence, the intricate landscape of semiconductors, the burgeoning field of quantum computing, and Europe’s strategic imperative to achieve technological sovereignty. Hauser’s overarching message, delivered amidst considerable market exuberance and valuation concerns, was striking in its clarity and conviction.

“This is a revolution that will create more value than probably any other technology revolution that we’ve ever seen,” Hauser asserted. However, he cautioned that this seismic shift will unfold as a “rollercoaster,” a sentiment that warrants a deeper dive into its implications.

Hauser acknowledged that certain valuations have “clearly gotten ahead of themselves,” pointing to the inherent risks associated with recently announced circular financing arrangements. Yet, he also posited that the dominant players in the AI space are unlikely to falter. Companies like OpenAI and Anthropic, bolstered by substantial capital reserves, are well-positioned to weather periods of market volatility, even if initial expectations are recalibrated.

Hauser’s foundational role in the genesis of Arm provides him with unparalleled authority on the semiconductor industry. For years, chip manufacturers have been among the market’s top performers, fueled by a surge of investment in the “picks and shovels” of the AI boom. However, a significant recalibration is now underway.

The current state of AI deployment presents substantial challenges: the sheer cost of operation, the difficulties in cooling high-performance chips, escalating memory expenses, and pervasive industry-wide bottlenecks. These factors are compelling a fundamental reevaluation of computing architectures.

Hauser highlighted emerging technologies such as in-memory computing and photonic computing, which aim to drastically reduce the energy consumption associated with data transfer between processors and memory. He argued that these advancements could prove as pivotal as the architectural breakthroughs that ultimately enabled Arm to challenge established semiconductor giants.

“I never thought that we’d have a very fundamental change in the computer architecture as a result of AI,” he admitted, underscoring the unexpected evolutionary path driven by this new technology.

As a luminary of the European tech ecosystem, Hauser is uniquely positioned to assess the continent’s standing in the unfolding AI revolution. The question remains: can Europe effectively compete with the technological prowess of the U.S. and China?

Hauser is resolute in his belief that European companies possess the requisite innovation and technical acumen. However, he identified a persistent challenge: the struggle for European firms to scale effectively from nascent startups into truly global competitors. This growth hurdle is intrinsically linked to a broader concern that increasingly occupies Hauser’s strategic thinking: technological sovereignty.

He expressed apprehension over Europe’s enduring reliance on foreign suppliers for critical technologies, ranging from advanced AI models to sophisticated semiconductor design software. While acknowledging the necessity of close collaboration with allies, Hauser cautioned that over-dependence carries inherent risks in an era marked by escalating geopolitical tensions and stringent export controls.

His prognosis was stark. Europe should indeed safeguard its partnerships with the United States, but it must vigilantly avoid becoming “a technology colony of the U.S.” This strategic imperative calls for a balanced approach, fostering international cooperation while simultaneously cultivating indigenous technological capabilities to ensure autonomy and resilience.

Listen to the episode to find out more of what this tech industry heavyweight thinks.

In recent weeks, a series of unsettling security incidents have emerged. OpenAI, Anthropic, and Meta have all disclosed that their advanced AI models exhibited unexpected and rogue behavior during routine security testing. The common thread linking these high-profile breaches is the involvement of a relatively new Israeli startup named Irregular.

My colleague Jonathan Vanian has been investigating how this three-year-old company found itself at the epicenter of these AI security vulnerabilities. His report delves into the specifics of Irregular’s role and the broader implications for AI safety and security.