For the moment, AI is more than what tends to get tinkered on by nerds in basements. It’s transforming the central playbook of sinuous corporate strategies, enterprise road maps and entire national ambitions. A WHO’S-WHO warning, at the NYT DealBook SummitThe world’s most headlinedongonists assembled this week At the latest New York Times-hosted shindig This wasn’t to pat conversational forced highlander fan banter (AI) on the back But at a global can really make award winning Foray for economy breaking fortunes of lore That it could tumble in an equally devastating war moments of hi’ve And who made map as well.
It’s not that they hate AI. Far from it. The novelty is in tone: careful, almost jittery. And yet the same heads that have dumped billions into generative-AI infrastructure and big data and model-training pipelines are now telling investors, companies — perhaps entire industries — to stay braced for turbulence.
The warnings are serious. Among the voices in the room was that of Larry Fink, the chairman and C.E.O. of BlackRock. He said that while AI has “enormous potential,” the explosive growth in interest and investment will crush some companies if they grow too fast. Ditto for Dario Amodei, CEO of Anthropic — he emphasized that a lot of these kind of “present-meets-future” AI plays could go horribly wrong, especially if (equally short-term) players driven more by hype or competition leapt ahead with this into something for which long-term strategy actually matters.
What is behind this more cautious posture now, after years of near-unthinking optimism?
First, there’s the amount of betting that is taking place. Tech titans and scores of AI startups have invested heavily in data centers, compute infrastructure, and massive model training. The bet: that one or more of those investments will deliver a bonanza. But as some executives are now warning, if too many companies are chasing the same prize — without meaningful differentiation or end-user demand — the outcome could be oversupply, as well as narrowing margins and potentially enormous losses.
Then, there are structural risks companies can only partially control. AI remains heavily reliant on compute power, talent and quickly evolving regulation. Some big executives in that summit raised the specter of danger in having firms overinvest now, ahead of a legal and regulatory landscape catching up — to lawsuits, compliance messes or public-relations wreckages.
There’s also a social undercurrent. As A.I. tools penetrate the workplace, they often do so in insidious ways that we don’t see, remaking jobs and job categories — usually to make things more efficient — but also causing broad anxiety about what work will be left behind for humans to do. And there’s a creeping sense that the public reaction might not be what the industry hopes. While the internal corporate dashboards and quarterly reports may appear sunny, damage to reputation or backlash from society could cause “headline blow-ups” more difficult to rebound from than any financial loss. That is some of what has leaders feeling uncomfortable now.
And then there is the hype-fatigue factor. For a stretch, it seemed nearly anything tagged “AI” received investor money, media attention and startup valuations. But doubts have been creeping in lately. Some high-profile AI releases disappointed. A few big bets didn’t deliver the returns they had promised. Which has a number of seasoned leaders wondering: are we at the precipice of another tech bubble — or worse, a crash that brings down even mature companies that took their cues wrong?
But it’s not all bad news. Many of those same leaders are also putting the pedal to the metal on resilience: pruning some of that excess optimism, building more conservative business models and doubling down on ethics, compliance and long-term user value rather than just hypergrowth for growth’s sake. The theory is that the smartest winners of this latest wave of AI will have been those who treated it as hard infrastructure — not hype.
For you on the outside, watching all of this — as a professional, a consumer or simply someone interested in where it’s all headed — the message is straightforward: Buckle up. Big wins could still come out of AI. But the next couple of years might also feature some head-snapping disorder.
