The New Turing Test

When OpenAI announced its latest funding round — $40 billion at a post-money valuation of $300 billion — one line in the press release jumped off the page: “New funding to build towards AGI.” Not just AI, but AGI. That small distinction signals something big. OpenAI is no longer just developing systems that understand language or recognize images — it’s now explicitly chasing Artificial General Intelligence: machines capable of performing any human task as well as, or even better than, we can.

That’s no small feat. Until recently, AGI was the exclusive domain of nerds, visionaries, and science fiction writers. But now, the term suddenly appears in an official press release from the most influential AI company in the world. As if it’s the most normal thing in the world. As if they’re saying: hey, we’re almost there.

According to Bill Gates, within the next decade we may no longer need doctors or teachers — at least, not in the way we know them today. AI will diagnose illnesses, recommend treatments, respond to patient questions, design lesson plans, tailor them to individual needs, and continuously refine them. What’s left for humans is the warmth, empathy, and presence that machines can’t replicate, assuming we’re still willing to make time for that.

The one-person unicorn

Meanwhile, Silicon Valley is placing bets on a different kind of milestone: the world’s first one-person unicorn. That’s a billion-dollar startup, run by a single individual, powered by a swarm of AIs that code, sell, negotiate contracts, design visuals, and handle communication. No bloated teams of 100 people, no endless funding rounds, no Slack channels full of misunderstandings. Just one person with a vision — and a tireless army of digital collaborators.

The entire AI industry is shifting toward agents. After generative AI, which was able to “generate” text and images, comes its successor: AI that can do things: make payments, book travel, schedule appointments, or even manage other AIs. Agents are about action, not imitation.

In this light, the classic Turing Test suddenly feels outdated.

As early as 1950, Alan Turing asked a simple yet profound question: “Can machines think?” Since “thinking” is hard to define, he designed a practical test: if a machine could hold a conversation indistinguishable from that of a human, it should be considered intelligent. Fast forward to today, and we’re chatting with models that pass for human without breaking a sweat. They write job applications, flirt on dating apps, offer comfort after heartbreak, and more often than not, we don’t even notice they’re machines.

The Turing Test? Consider it passed.

AIs making billions for you

So what should be the new Turiung Test, then?

According to Mustafa Suleyman, co-founder of DeepMind and author of The Coming Wave, it’s simple. Give an AI $100,000 in seed money, a hefty stack of GPUs, an Amazon.com account, and instruct it to “Come back in a month with $1 million. Start trading.” No role-play, no illusion. Just: go do something in the real world. Act. Build. Grow.

The foundations of work and the economy are changing before our eyes — and faster than we ever imagined.

Ethan Mollick, a professor at Wharton and leading voice on AI and the future of work, sees the shape of what’s coming. He doesn’t describe AI as a tool, but as an intern — a tireless digital colleague that summarizes reports around the clock, conducts market research, builds pitch decks, and maps out strategies. As he often puts it: “Everyone works with a team now — even if you’re working alone.”

The line between solo entrepreneur and mini-multinational is quickly fading. What takes ten people today might soon be handled by one person and ten AIs. And not long after — perhaps without that one person at all.

While we’re busy generating Studio Ghibli-style portraits of our pets, the real questions are becoming harder to ignore. What happens when AI and capital converge, and human labor becomes increasingly optional? What if the most efficient “employees” don’t need sleep, coffee breaks, or parental leave? What if productivity becomes exponential, but only for those with access to the right models?

Because the true disruption isn’t that AI will replace us. It’s that it will supercharge a few and leave the rest behind and invisible. Those who own or can command this new intelligence will hold an advantage that’s not just significant, but exponential.

When my country, Belgium, recently came to a standstill amid a wave of strikes, I was struck by an uneasy feeling. The actions felt like a desperate attempt to reclaim a past that’s never coming back. Meanwhile, the real conversation — about the future of work and society — is barely happening. And that’s not going to cut it.

As Alan Turing warned long before AGI became a buzzword: “We can only see a short distance ahead, but we can see plenty there that needs to be done.”

If you want to learn more about emerging technologies and innovation, join me and 5 other top moderators on the Flanders Technology Tour 2025 in October!

This is a translation of a piece originally published in Dutch in the newspaper De Tijd.