Demis Hassabis: Why AI Job Cuts Are a Mistake
▼ Summary
– Demis Hassabis says Gemini 3.5 Flash can perform complex coding tasks like translating code bases and writing operating systems, but he dismisses claims that AI will eliminate software developer jobs.
– Hassabis believes engineers will become more productive with AI, allowing companies to take on more projects, and he suggests that predictions of job displacement may be motivated by fundraising.
– Google unveiled Gemini 3.5 Flash and an upcoming Gemini 3.5 Pro, along with a safer agentic assistant called Spark, as it seeks to catch up to rivals Anthropic and OpenAI in AI coding.
– Hassabis doubts that AI rewriting its own code will immediately lead to superhuman-level AI, and he notes that AI has not yet produced a blockbuster app or game without human help.
– Hassabis warns that companies replacing developers with AI lack imagination, while Alphabet is positioned to benefit from AI-driven software productivity gains.
Demis Hassabis, the CEO of Google DeepMind, is eager to highlight the capabilities of his company’s latest model, Gemini 3.5 Flash. This system has been trained to handle complex agentic coding tasks, such as translating entire code bases between languages, rooting out deeply embedded bugs, and even writing complete operating systems from scratch. Yet Hassabis firmly rejects the notion that this spells the end for software developers.
“I have no idea why people are going around talking with certainty about that,” Hassabis tells WIRED ahead of the model’s debut at today’s Google I/O event. He suggests that some of the dire predictions may have ulterior motives. “Perhaps there is an ulterior motive for putting those messages out; raising money or whatever,” he says. “From my point of view, from DeepMind and Google’s point of view, if engineers are becoming three or four times more productive, then we just [want to] do three or four times more stuff.”
The impressive coding prowess of modern AI has fueled widespread anxiety that programming roles and other white-collar jobs could soon vanish. Some AI company executives have forecasted mass displacement, and major tech firms like Amazon, Salesforce, and Block have cited AI as a reason for recent layoffs. Hassabis, however, believes that Alphabet,Google’s parent company,could be uniquely positioned to capitalize on a software productivity revolution. “I have a million ideas, from lab drug discovery to game design,” he says. “I’d love to have some free engineers to go and do those kinds of things.”
Hassabis warns that companies trying to replace developers with AI may be making a critical error. “I think it’s a lack of imagination,and a lack of understanding of what’s really going to happen,” he says.
At its annual developer event, Google unveiled a wave of new AI features. Its new coding tool, Antigravity, powers Gemini 3.5 Flash with frontier coding and reasoning capabilities while being faster and cheaper than rival offerings, according to the company. A more powerful version, Gemini 3.5 Pro, is slated for release next month. Google needs to close the gap in AI coding, which has become a vital and profitable area for modern AI models. According to a 2025 Stack Overflow survey, Anthropic and OpenAI currently lead in developer adoption with their tools Claude and Codex.
Google also demonstrated an agentic assistant named Spark, which operates within Google Cloud and has access to the company’s apps. The design is intended to be safer than tools like OpenClaw because it has limited access to personal data, Google says. Other agentic demos included an Android version with a built-in AI agent and a refreshed Google Search that uses agentic coding to generate a site or app on the fly in response to a search query.
AI coding has captured the imagination of the tech world in recent months, sparking hopes that models could one day rewrite their own code in a self-improvement loop. Hassabis says this is possible but doubts it will immediately lead to superhuman-level AI. He notes that progress in other scientific fields may require AI to develop a deeper understanding of the physical world and even the ability to conduct experiments within it.
Even in the seemingly solved domain of coding, Hassabis points out that AI has yet to produce a blockbuster app or video game without human assistance. “I think there’s something missing,” he says.
(Source: Wired)
