The SaaS Model Of Amortizing Software Building Costs Across Millions Of Users Could Soon Be False: Anthropic CEO Dario Amodei

Anthropic CEO Dario Amodei had been largely on the money when he’d said last year that 90% of code would soon be written by AI, and he now has a disconcerting prediction for the SaaS industry.

In a recent interview at Davos, Amodei shared observations from within Anthropic that paint a striking picture of how AI is already transforming software development. His comments suggest that the fundamental economics of the software-as-a-service industry—built on spreading development costs across millions of users—may be approaching obsolescence. What makes his perspective particularly noteworthy is that it comes from someone whose company is actively building the tools driving this transformation, giving him a front-row seat to the changes underway.

Speaking about Anthropic’s latest model, Opus 4.5, Amodei revealed the dramatic shift happening inside his own company. “I have some engineering leads within Anthropic who have basically said to me, I don’t write any code anymore. I just let Opus do the work and I edit it,” he said. The proof of this productivity leap came in the form of Cowork, a new tool the company recently released. “This was a version of our tool, Claude Code for non-coding. This was built in a week and a half, almost entirely with Claude Opus.”

But Amodei was quick to acknowledge the transitional nature of this moment. “There’s still things for the software engineers to do, right? Even if the software engineers are only doing 10% of it, they still have a job to do or they can take a level up. That’s not going to last forever. The models are going to do more and more, and so there’s an incredible—this is a microcosm—you can see there’s an incredible amount of productivity here.”

Then came the most striking prediction. “Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users—that may start to be false,” Amodei said. He painted a picture of a future where custom software becomes disposable: “For this meeting, it might cost a few cents to just say, I don’t know, let’s make some apps so people can talk to each other. It just may be very flexible and recyclable.”

Yet Amodei didn’t shy away from the darker implications. “But at the same time, there are whole jobs, whole careers that we built for decades that may not be present. And I think we can deal with it. I think we can adjust to it, but I don’t think there’s an awareness at all of what is coming here and the magnitude of it.”

Amodei’s warnings align with a growing chorus of concerns about AI’s impact on the software industry, though not everyone shares his pessimism about SaaS platforms specifically. In a recent interview, he suggested we could be just 6-12 months away from AI handling everything software engineers do end-to-end. Microsoft CEO Satya Nadella has similarly predicted that SaaS applications will “collapse” in the AI agent era, arguing that agents will eliminate the need for traditional software interfaces. However, Nvidia CEO Jensen Huang has offered a contrasting view, suggesting that AI won’t disrupt SaaS platforms because they’re sitting on valuable data goldmines that will become even more crucial in the AI era.

The market appears to be taking these concerns seriously. Software stocks have experienced significant turbulence, with the Morgan Stanley SaaS index showing steep declines even as the broader Nasdaq 100 has maintained gains. The divergence suggests investors are reassessing the long-term value proposition of traditional software companies in an AI-driven future.

If Amodei’s vision materializes, the implications extend far beyond software engineering roles. The entire architecture of the tech industry—from venture capital models that bet on scaling software across massive user bases, to the careers of millions of developers, product managers, and related professionals—could face fundamental disruption. The question is no longer whether AI will transform software development, but how quickly, and whether the industry and society can adapt at the pace Amodei suggests is coming.

Posted in AI