AI Voices, May 22: Hassabis Calls the Singularity, Prince Abolishes the Middle Manager, Daniela Amodei Warns Europe

AI Voices, May 22: Hassabis Calls the Singularity, Prince Abolishes the Middle Manager, Daniela Amodei Warns Europe

Three AI leaders on May 21: Demis Hassabis (Google DeepMind CEO) tells Google I/O he believes we're 'standing in the foothills of the singularity' and puts AGI at 50% probability by 2030; Cloudflare co-founder Matthew Prince lays out why AI eliminated an entire job category — the 'measurers' — and argues the change is permanent; Anthropic President Daniela Amodei urges European governments to regulate AI before adoption outpaces safeguards, as Anthropic opens its Milan office.

AI Founder & Investor Interview Daily Digest
2026/5/22 · 8:26
購読 3 件 · コンテンツ 3 件
Three AI figures spoke publicly on May 21, each from a different vantage point: a lab chief who thinks AGI is years away and said so in front of an audience that gasped, a company founder who just laid off an entire job category and has no plans to rehire it, and an AI company president who told European governments to learn from social media's mistakes before it's too late.

Demis Hassabis: "We're standing in the foothills of the singularity"

Google DeepMind CEO and co-founder Demis Hassabis closed Google I/O 2026 on Tuesday with a line that drew an audible gasp from the room — and then spent the following day defending it.
"When we look back at this time, I think we will realize that we were standing in the foothills of the singularity," he told the developer audience at Mountain View 1.
In a follow-up interview with Axios co-founder Mike Allen at the conference the next day, Hassabis was specific about what made him use the word now, not last year or the year before: "This year, with the agentic systems that we're all seeing and using, I think we can start feeling it now." He said he's been using AI agents at night to build mini video games — projects that would have taken months of engineering work a few years ago.
Demis Hassabis, Google DeepMind CEO, at Google I/O 2026
Demis Hassabis at Google I/O 2026 1
Hassabis is not predicting a runaway machine takeover. His definition of "singularity" is roughly synonymous with full AGI — a point he puts at a 50% probability by 2030, a number he's cited consistently for months. He describes himself as a "cautious optimist." But choosing to say it out loud at the main stage of Google's flagship developer conference — instead of a podcast or late-night X thread — is a deliberate tonal signal, and one that Sundar Pichai apparently sanctioned: Pichai handed Hassabis the closing slot for the first time, right after the unveiling of Gemini for Science.
Hassabis believes AI's impact will be 100 times that of the Industrial Revolution, mostly through breakthroughs in science and healthcare. "We're in a moment of immense promise, but also enormous responsibility," he said 1.
For AI professionals and investors watching the conference: this was the CEO of the world's most prestigious AI research lab making a public AGI timeline bet on the main stage — not a podcast.

Matthew Prince: AI just made an entire job category redundant

Cloudflare co-founder and CEO Matthew Prince laid off roughly 20% of the company's workforce earlier this month. He published a Wall Street Journal op-ed explaining why, and gave Fortune an interview on May 21 expanding on it 2.
The framing is blunt. Prince divides white-collar employees into three buckets:
  • Builders: engineers and product developers
  • Sellers: revenue-generating roles
  • Measurers: middle management, finance, legal, internal audit, revenue recognition
The third category, he says, is gone. "Tireless, independent, efficient and available, AI systems can now measure an organization with a level of objective detail and precision that was previously impossible even for the best employees."
He is not arguing that Cloudflare is shrinking. He says the company has a record number of open positions in "growth-driving areas" — it's just that those positions don't include anyone to track, audit, or sign off on what the builders and sellers do.
Prince also addresses what he calls "AI-washing" — companies that were overstaffed for years and are now using AI as the excuse. He quoted a16z co-founder Marc Andreessen, who said on a recent 20VC episode: "Essentially, every large company is overstaffed. It's at least 25%. I think a lot of companies are 50%, some are 75%. And now they all have the silver bullet excuse: it's AI."
Prince's own position is that Cloudflare's case is real — AI genuinely automated the measurement work — but he's honest that not every company making similar claims can say the same.
リンクプレビューを読み込んでいます…
The structural argument here cuts deeper than a single layoff announcement. If "measuring the organization" can be fully automated, a large swath of corporate hierarchy — and the career paths built around it — stops making economic sense. Prince is essentially saying that AI collapsed the ROI on supervision.

Daniela Amodei: Europe needs rules before the damage is done

Anthropic President and co-founder Daniela Amodei spoke in Milan on May 21, where Anthropic announced it will open an Italian office this month — its third European city after Paris and Munich 3.
リンクプレビューを読み込んでいます…
Amid the expansion news, she made a point about governance that cut past the office announcement: "Governments should act early to avoid repeating mistakes made with social media platforms, where safeguards lagged widespread adoption."
The remark is short but carries a specific argument. Social media arrived, scaled, caused documented harm to mental health and political discourse, and only then did regulators scramble to catch up — a cycle of deployment followed by regret that she's saying AI governance cannot afford to repeat.
Anthropic is tripling its international workforce to meet demand for Claude outside the United States. Its European offices now cover Dublin, Zurich, London (~200 staff), Paris, Munich, and — starting this month — Milan.
The company has also taken positions that make the regulatory argument more than rhetorical: Anthropic has clashed with the Trump administration over guardrails restricting how its models can be used for autonomous weapons targeting and domestic surveillance.

Comparison: what each is betting on

SpeakerRoleWhat they said on May 21Underlying bet
Demis HassabisGoogle DeepMind CEO/co-founder"Standing in the foothills of the singularity"; AGI 50% likely by 2030The capability jump in agentic AI is qualitatively different from prior waves
Matthew PrinceCloudflare CEO/co-founderAI has made "measurers" obsolete; layoffs are structural, not cyclicalAI-driven labor reallocation is permanent — the "measurer" job category will not return
Daniela AmodeiAnthropic President/co-founderGovernments must act before AI scales the way social media didThe window for proactive governance is open now and will close faster than policymakers expect
Marc Andreessen (context)a16z co-founderMost large companies are 25-75% overstaffed; AI is "the silver bullet excuse"The current wave of AI layoffs is partly real, partly a cover for existing organizational bloat

このコンテンツについて、さらに観点や背景を補足しましょう。

  • ログインするとコメントできます。