For those operating at the cutting edge of AI, the new technology is leading to performance gains which would’ve seemed straight out of science fiction even a year ago.
Eric Schmidt, the former CEO of Google who ran the company through its most transformative years, has been sounding off about what AI is doing to the world of programming. His central argument: while AI will hollow out the middle of the programmer job market, the very best will not only survive — they’ll become dramatically more powerful.
“The very top programmers were worth ten times more than the ones right below,” Schmidt said. “There’s something special about the mathematical reasoning skills of programmers. Those people will become more valuable, not less valuable, because these systems need to be controlled by humans at the moment. Those people will be capable of grasping the parallelization and the activities of this.”
To illustrate, Schmidt described a conversation with a programmer at a startup he’s involved with:
“He said, ‘Here’s what I do.’ He’s working on UIs of various kinds, and he said, ‘I write the spec of what I want, and then I write a test function, an evaluation function, and then I turn it on.’ I said, ‘What time?’ And he goes, ‘Seven o’clock in the evening.'”
What happened next is what makes this remarkable. Schmidt continues:
“‘What do you then do?’ Well, he has dinner with his wife and he goes to sleep. I said, ‘Do you wake up?’ He said, ‘No, I sleep very well.’ ‘When does it finish?’ ‘Four in the morning.’ And then he gets up, has breakfast, does whatever he does, and then he sees what’s been invented.”
Schmidt was barely able to contain his astonishment:
“It’s mind-boggling. And this example I used with this young man — this is what the power of these systems are. If you can define the evaluation function and you can let it run, and if you have enough hardware, you’re inventing worlds. I mean, this stuff would’ve taken me six months and ten programmers at Google to do the same thing. This poor guy’s sleeping.”
What Schmidt is describing is now becoming a recognisable pattern, not an outlier. A quarter of Y Combinator startups already have more than 95% of their code generated by AI — produced by small, highly technical founding teams doing in days what once required entire engineering departments. YC has even updated its application process to ask founders to submit a coding agent session they’re proud of, a direct acknowledgment that directing AI effectively is now a core technical skill.
The story Schmidt tells about the programmer’s overnight workflow is essentially the agentic coding loop in practice: write a spec, define an eval, fire it off, and let AI generate and test solutions while you’re away. The human’s job has shifted from writing code to judging it. Schmidt himself has said elsewhere that programming as he knew it is over — every aspect of the design work he did in his career has now been automated — while also predicting that AI will replace the vast majority of programmers within a year. Anthropic CEO Dario Amodei puts the timeline for AI writing 90% of code at three to six months.
The bifurcation Schmidt describes is stark: the programmer who can frame the right problem, define a rigorous evaluation function, and interpret results at 4am has essentially become a force multiplier — one person doing what once required a team. Everyone else is being squeezed. The overnight software factory is already open. The question is who gets to run it.