Dario Amodei, the chief executive of AI giant Anthropic, has said that artificial intelligence could handle 'most, maybe all' of what software engineers currently do within six to 12 months.
The warning landed like a grenade across the tech industry.
'I have engineers within Anthropic who say I don't write any code anymore. I just let the model write the code, I edit it,' Amodei said during an interview with The Economist at the World Economic Forum in Davos on 20 January 2026.
Large technology companies such as Google, Amazon and Microsoft are also increasingly utilising automated systems to build portions of codebases; rather than using entire teams to build codebases, AI can complete this with a single model and a prompt. This is not an example of augmentation; it is an example of substitution.
For thousands of developers worldwide, it felt less like a forecast and more like a redundancy notice.
What Amodei Actually Said at Davos
Amodei did not frame his remarks as speculation. He pointed to what is already happening inside his own company.
Engineers at Anthropic, he claimed, have shifted from writing code to reviewing and refining output generated by AI models. The role of the developer, he suggested, is moving from creator to editor.
'We basically have a Moore's Law for intelligence where the model is getting more and more cognitively capable every few months,' he told the Davos audience.
He went further. Amodei reiterated his earlier prediction that, by 2026 or 2027, advanced AI systems could conduct research and innovation at a level comparable to that of Nobel Prize winners.
The comments drew sharp attention. Not because the idea is new, but because of who said it.
Why the Industry Is Taking This Seriously
Anthropic builds some of the most capable AI coding tools on the market. When its chief executive says engineers are no longer writing code from scratch, it carries weight that a think-tank white paper does not.
Zoho founder Sridhar Vembu responded on X on 6 February 2026. His message was blunt.
'We better pay attention to him because he has the best coding tool in the world,' Vembu wrote.
Vembu later went further, advising software developers to begin considering alternative livelihoods. 'I don't say this in panic, but with calm acceptance and embrace,' he added in a follow-up post.
Ryan Dahl, the creator of Node.js, was even more direct. 'The era of humans writing code is over,' he posted on X.
Google, Amazon and Microsoft are already using AI to generate portions of their codebases. The shift is not theoretical. It is operational.
Not Everyone Agrees
Some industry leaders at Davos pushed back. Fortune reported that few C-suite executives outside AI firms shared Amodei's timeline.
Amodei had previously projected that up to 90 per cent of code would be AI-written by the end of 2025. That proved accurate for Anthropic itself — but not for the wider industry, where estimates put AI-written code at between 25 and 40 per cent across other software companies.
His sense of pace, some analysts suggested, may be skewed by operating at the frontier.
Still, the direction of travel is not in dispute. Entry-level coding roles face the sharpest risk, multiple observers noted. If AI can generate clean, functional code in seconds, the business case for hiring and training junior developers weakens fast.
What Happens Next for Engineers
The role will not vanish overnight. But it will change beyond recognition, industry figures claimed.
Developers may shift toward system architecture, product strategy, and oversight of AI-generated output. Fewer engineers could be needed to deliver the same volume of work.
'Developers won't disappear, but their job descriptions will change fundamentally,' one senior technology consultant told The420.in.
Amodei's broader warning extends well past software teams. He argued that AI improvement is accelerating across knowledge-intensive professions. Software engineering, he said, is simply the clearest test case — structured, digital and measurable.
What happens here may preview what follows elsewhere. The debate is no longer about whether AI will reshape coding. It is about how quickly — and who gets left behind.
Originally published on IBTimes UK
© Copyright IBTimes 2025. All rights reserved.
