• Last week, OpenAI launched the newest version of its flagship software — and told users they'd have to give up the ones they already used.
• Redditors and other people on the internet complained — and OpenAI CEO Sam Altman changed his mind.
• Does that mean Sam Altman is a weak CEO? Maybe. But for now, he just seems like a flexible one.

All hail the new ChatGPT, which is much better than the old ChatGPT, which we're getting rid of.

That was the messaging from OpenAI CEO Sam Altman and his team last week.

A day later, Altman changed his mind. He told the world the older version of ChatGPT was going to stick around, after all — in addition to the new version that was meant to replace it.

I'm not getting into the weeds here about the change and change back, which is confusing for people who use ChatGPT, and impenetrable to non-users. (If you want to, I suggest you head to Business Insider's coverage, or this post from analyst Ben Thompson, for details.)

I'm most interested in Altman's incredibly quick pivot.

Because I'm having a hard time thinking of a CEO hyping a new product launch, and almost immediately changing course afterward, supposedly because his customers didn't like it.

Can you think of one? The most obvious one I can recall is New Coke, which you have to be pretty old to have tried. It only lasted for a few months in 1985, because lots of Old Coke drinkers hated it, and it's now synonymous with Corporate Mega Flops. But it still lasted for a few months — not a single day.

And this is different than product flops like the Samsung Galaxy Note 7, which was pulled off the market after a couple of months because some of them exploded.

And to be clear — Altman isn't recalling his newest, very high-profile AI engine. It still exists; he's just reversing his call to get rid of the older one.

(Business Insider owner Axel Springer has a commercial agreement with OpenAI. And our CEO thinks we should all use AI in our day-to-day work.)

It's possible that there are other, yet-to-surface explanations for Altman's change of heart. But so far, the only one he's offered is that he heard from people on Reddit and presumably other places who were upset to lose the versions of the service they already had.

If you are an Altman fan, you can paint the episode as a story that shows you how nimble and responsive a Big Tech CEO can be.

If you are less generous, you might argue that this was something Altman and his company should have seen coming, and acted accordingly. Either by not budging, and explaining to users that they were wrong, and would learn to love the new tech. Or by not making the move in the first place.

Thompson, in his Stratechery newsletter, worries that Altman's quick flip is a sign of a bigger problem — that he's too willing to tell people what they want to hear:

The real question for OpenAI is if they are in fact ... just a bit too obsequious and sycophantic. The paradox of successful consumer companies from Apple to Facebook is that they give customers what they want, but they don't ask them; they make decisions and then seek out revealed preference through data, not stated preference on social media. Hopefully OpenAI did that in this case; my concern is that the more realistic explanation is that this is a company that, in the end, can't say "no" to anyone.

Maybe! But I think this is probably a pretty small chapter in the OpenAI story — a visible, but ultimately not-that-consequential misstep. Maybe OpenAI really did misjudge its customers. But it was pretty easy to make those customers happy, simply by … not taking something away from them.

It also helps that this was a do-over Altman and crew could do with a couple key strokes. There were no devices (yet) to recall, no refunds to issue.

In that sense, this reminds me of something closer to a branding or marketing snafu, like a new Gap logo that lasted for 10 days in 2010, or that Kendall Jenner Pepsi ad from 2017 that disappeared after people called it stupid and tone-deaf. Embarrassing screw-ups, but not the first thing you think about when you think about those companies.

I myself had forgotten those stories until ChatGPT reminded me of them, when I asked for comparable flip-flops. (I don't use ChatGPT to write my stories, but I definitely find myself using it as a superior version of Google more and more these days.)

And yes, if we see more waffling from Altman in the months and years to come, we'll be able to point back to this botched rollout as the start of a pattern.

But for now, this one seems like an odd and interesting footnote, and not much else.

• This post originally appeared in the BI Tech Memo newsletter.
• Sign up for the weekly BI Tech Memo newsletter here.

There's a fear in investing when a sector swells rapidly. Booming stock prices and aggressive spending feel great, until things inevitably cool off. Then comes the reckoning: Who overdid it in irreversible ways?

Big Tech is in an AI arms race, each company trying to outspend the others on data centers, GPUs, networking gear, and talent. Engineers can be let go. But the infrastructure? That's permanent. If the AGI dream fades, you're stuck with massive, costly assets.

So when Google announced it would hike capex by $10 billion to $85 billion in 2025 eyebrows went up. Most of it is for things you can't walk back: chips, data centers, and networking.

Google is "jumping aboard the AI crazy train," Bernstein Research analyst Mark Shmulik wrote, referencing a song by the late bat biter Ozzy Osbourne.

Meta's Mark Zuckerberg brags about Manhattan-sized data centers. And Elon Musk keeps hoarding GPUs. While Sam Altman is building mega-data centers with partners. JPMorgan dubbed this "vibe spending," warning OpenAI might burn $46 billion in four years.

It's no shock when Elon, Zuck, and Sam flex on capex. But Google? That's surprising. "Google doesn't do this," Shmulik said. The company has been viewed as measured in recent years, prioritizing investment intensity with care. Not anymore.

Now investors want to know: Will these swelling bets pay off?

There are promising signs. Since May, Google's monthly token processing (the currency of generative AI) has doubled from 480 trillion to nearly a quadrillion. Search grew 12% in Q2, beating forecasts. Cloud sales surged 32%. CEO Sundar Pichai said Google is ramping up capex to support all this growth.

But it's still a huge gamble. "Does the current return on invested capital seen in both Search and Cloud hold up at higher [capex] intensity levels," Shmulik asked, "or is the spend a very expensive piece of gum trying to plug an AI-sized hole?" He leans optimistic.

Still, Google shares rose just 1% after these results. Not exactly a resounding endorsement.

Sign up for BI's Tech Memo newsletter here. Reach out to me via email at [email protected].

• A think tank boss believes AI will replace most jobs by 2045, leaving billions without work.
• Sex work, coaching, and politics may survive, but not at the scale society needs, Adam Dorr said.
• He said the future could bring mass inequality or "super-abundance," depending on our response now.

By 2045, robots and artificial intelligence could render most human jobs obsolete — and there's little time to prepare for the fallout, according to Adam Dorr, director of research at the RethinkX think tank.

In a Wednesday interview with The Guardian, Dorr warned that machines are advancing so rapidly that within a generation, they'll be able to perform virtually every job humans do, at a lower cost and with equal or superior quality.

Drawing from historical patterns of disruption, he compared today's workforce to horses in the age of cars, or traditional cameras in the age of digital photography.

"We're the horses, we're the film cameras," he said.

Dorr and his research team have documented more than 1,500 major technological transformations. In most cases, he said, once a technology gains even a few percentage points of market share, it quickly dominates — typically within 15 to 20 years.

"Machines that can think are here, and their capabilities are expanding day by day with no end in sight," he said. "We don't have that long to get ready for this."

Still, he said, not every job is destined for extinction. Dorr believes a narrow set of roles may survive the AI takeover, especially those grounded in human connection, trust, and ethical complexity.

He pointed to sex workers, sports coaches, politicians, and ethicists as examples of jobs that could remain relevant.

"There will remain a niche for human labor in some domains," he said. "The problem is that there are nowhere near enough of those occupations to employ 4 billion people."

Dorr argued that the looming upheaval could lead either to mass inequality or to what he called "super-abundance" — a society where human needs are met without traditional labor. But achieving the latter, he said, will require bold experiments in how we define work, value, and ownership.

"This could be one of the most amazing things to ever happen to humanity," he said — but only if we're ready.

The AI takeover debate is heating up

Several top AI researchers and tech leaders have shared Dorr's concerns, though views on which jobs will endure vary.

Geoffrey Hinton, often called the "Godfather of AI," warned that "mundane intellectual labor" is most at risk. On the Diary of a CEO podcast in June, he said he'd be "terrified" to work in a call center or as a paralegal.

Hinton believes hands-on roles like plumbing are safer, at least for now, saying it will be a long time before AI is "as good at physical manipulation" as people.

In May, Anthropic CEO Dario Amodei told Axios that he believes that half of all entry-level white-collar jobs, including roles in tech, finance, law, and consulting, could disappear within five years.

But Nvidia CEO Jensen Huang and Meta's Yann LeCun have pushed back, saying AI will transform jobs, not eliminate them entirely.

OpenAI CEO Sam Altman also said AI will displace many roles, but believes new ones will emerge, even if they look "sillier and sillier" over time. "We have always been really good at figuring out new things to do," he said.

MIT economist David Autor took a darker view: AI may not wipe out jobs, but it could make people's skills worthless, ushering in a "Mad Max" economy where many fight over a shrinking pool of valuable jobs.