Stefan Weitz, CEO and co-founder of the Human[X] conference, welcomed attendees to the AI-focused bitshow in San Francisco with the promise that they would receive no certainty and no playbook.

"You are moving at speed that you didn't choose through conditions you can't see," he said, echoing perhaps unintentionally Rod Serling's prelude to The Twilight Zone.

His evocation of speed in fog implies the risk of a crash – business derailment or market correction as you prefer. 

That fear – of missing out, of being left behind, of hitting a wall, of mass unemployment – has been stirring in the words of OpenAI's Sam Altman or Anthropic's Dario Amodei in their public pronouncements about the impact of AI.

The way through, the safe passage, isn't included for the $4,000 price of admission.

Citing the vast amount of global spending on AI – $2.5 trillion this year, up 44 percent – Weitz drew an analogy between the investment-driven AI imperative and the shift from steam engines to electricity during the industrial revolution. 

The electrification of industry, he said, didn't show up as productivity until decades after factories plugged in. Better results required rethinking factory design and assembly lines. Similarly, he suggested, the failure of most AI projects in enterprises is the result of trying to shoehorn AI into existing business processes.

The way forward, Weitz said, is to ask what you would build if you were starting from scratch.

Greenfield design though is easier when there's no other option – Germany and Japan were forced to retool after World War II because their legacy industrial infrastructure had been destroyed. Corporate America or companies elsewhere may need to take more of a beating before they're sufficiently motivated to reinvent themselves.

AWS CEO Matt Garman thinks that beating is coming. Following Weitz's wary welcome, Garman sat down for an interview with CNBC's Kate Rooney.

Asked about the magnitude of the AI hype cycle, Garman said that the technology disruption is huge. "AI is going to transform every single company, every single job, every single way that we do work, every product interaction," he declared. "It's gonna have a massive transformation."

Asked how he's sure that AI isn't overhyped, Garman said, "People ask me that all the time. I think it's like one of the funnier questions I get." He proceeded to ask people in the audience to raise their hands if they're seeing positive ROI now from their AI investments or if they expect to see it within six months.

The audience complied and Garman estimated about 70 percent of audience members had raised their hands.

It would be remarkable if such a large portion of attendees had actual knowledge of AI costs and returns in their company. If you ask this reporter, it was a smaller percentage than that.

Nonetheless, Garman insisted that he has asked about ROI before and the answer he's been getting is closer to 90 percent. Internally, at Amazon, he said, software developers are roughly 4.5x more efficient with AI.

"That's not an efficiency that people just give up," he said. "And I think that there's more to come."

• AMD's AI director slams Claude Code for becoming dumber and lazier since last update
• Shots fired – literally – over proposal to build datacenter in Indianapolis
• Stack Overflow abandons redesign after loyalists criticize it
• Break, no fix: Apple and Samsung make repairs hard

When Rooney pressed him to acknowledge the hype, Garman conceded that anytime there's a technology disruption, there's a lot of investment and that some companies will fail.

"The last big bubble we had was the internet bubble," he said. "Last I checked, the internet still is pretty big. It's not like it went away.... There will be billion-dollar valued AI startups that don't make it. I'm 100 percent sure that's true. That's been true of every VC cycle, and there will be some that'll be the biggest enterprises ten years from now."

Garman's advice for the young is to keep learning and adapt. "Maybe the best skill that you can learn is learn to learn," he said. "Pick up new skills. Don't assume that whatever you learned is now good for the next 40 years."

For AWS employees, that appears to be a requirement. Garman said, "I think if we have employees that are not willing to learn and kind of want to say like, 'I just want to do the same exact thing that I've been doing for 20 years and I want that to be my job'... They should probably be nervous."

For companies, particularly SaaS businesses, Garman didn't have any clearer path through the fog than Weitz. When Rooney asked the AWS chief to elaborate on prior remarks about the SaaS selloff being overblown, he hedged his bets.

"Every time there is a technology disruption, it is absolutely something that incumbents should worry about," he said, likening the situation to when cloud computing challenged boxed software.

"When there's a technology disruption like this, it effectively takes all of the business that's out there, the trillions of dollars of SaaS products and software products, and dumps them on the table, it says everybody has to go win this again. So that is enormously disruptive for the incumbents. But it's also a huge opportunity."

Garman argued that incumbents can't be dismissed because they know a lot about the areas in which they operate, they have a customer base, and possess a trove of existing data and knowledge about workflows.

"It's definitely overblown when people are like, 'oh you can just use Claude Code to write a CRM," he said. "No. There's a lot of subtlety that goes into building applications out there. If the incumbents choose to try to put up walls, protect what they have, and not lean in and innovate, they're in trouble. But if they lean into AI, if they think about new business models, if they think about new opportunities, I actually think it's an opportunity for incumbent software players to expand what they do today."

Grow or die. It could be either. You were told there would be no roadmap. ®

More about

• AI
• AWS
• Enterprise

More like these