For the longest time, ideas weren’t thought to be worth very much in the entrepreneurial world — conventional wisdom said that ideas were cheap, and it was execution that mattered. But AI could turn that idea on its head.
That, at least, is the view of Amjad Masad, CEO of Replit, the browser-based coding platform that has been at the forefront of making software development accessible to non-programmers. In a recent conversation, Masad argued that as AI continues to drive the cost of building toward zero, the bottleneck in entrepreneurship is shifting — away from execution and toward the quality and velocity of ideas themselves.
“I think (idea generation) is actually the core skill in the AI era,” Masad said in an interview. “If you want to work on a skill, it’s going to be about idea generation, because the cost of implementation of those ideas is going down rapidly. It’s going to go to zero at some point. So the bottleneck becomes: how fast can you generate ideas?”
For Masad, idea generation isn’t an abstract or mystical talent — it’s a habit of observation. Being culturally plugged in, attuned to emerging trends, and alive to what people are talking about is, in his view, a genuine competitive edge.
“One component is perception — just looking around you in the world and seeing what’s happening. What are the trends? Are you plugged in on social media? What are people talking about? What is the most interesting thing that’s happening, and is there a market for that?”
Masad offered a concrete example to illustrate the kind of culturally-aware ideation he has in mind — a ‘looks maxing’ app.
“We were discussing this idea of a looks maxing app. It’s something in the ether right now — people are discussing it, people are interested in it. Can you build an app that gives you feedback on your looks maxing progress?”
From there, Masad made what may be his most provocative point: that traits and habits long considered unproductive — chronic social media use, a short attention span, an obsession with novelty — may be revalued as genuine assets in an AI-accelerated world.
“Being plugged in is super important. A lot of the vices that older generations think are vices might actually become advantages. If you’re a terminally online person, that might be an advantage because you know what’s happening in the world,” he says. “If you’re also someone who’s ADHD, really interested in novelty, wanting to try a lot of different things — that’s actually an advantage, because AI really benefits people who can try a lot of things really quickly. Obviously you need to get things to completion; you need to have some grit at some point once you’ve got some validation. But trying a lot of ideas is key.”
Masad’s framing tracks with a broader shift underway across the technology industry. The rise of vibe coding — writing software by describing intent in natural language rather than writing code directly — has already begun to collapse the distance between an idea and a working product. Anthropic CEO Dario Amodei has predicted that AI could be writing 90% of all code within months — and essentially all of it within a year. NVIDIA’s Jensen Huang has gone so far as to say that his greatest joy would be if his engineers never had to write code at all — freed instead to focus on problem identification. Investor Naval Ravikant has argued that vibe coding is now doing the work once done by product managers. If these predictions bear out, Masad’s thesis follows almost inevitably: in a world where building is nearly free, the scarcest input is the idea worth building in the first place.