I just thought about Naval Ravikant's perspective on AI, which is somewhat different from the current viral narrative.

So here's the thing—everyone is panicking about AI taking over all jobs. The CEO of OpenAI once said AI will take 95% of speaker jobs, then the CEO of Anthropic said software engineers will be finished in 6-12 months. It’s like the world is ending. But Naval Ravikant, who is the co-founder of AngelList and an early investor in Uber and Twitter, has a different take.

According to Naval Ravikant, we all overestimate the productivity gains from AI. He argues that AI will always make mistakes, no matter how advanced it becomes. So software engineers? They are still irreplaceable.

This is what’s interesting from Naval Ravikant’s perspective—software engineers have two major advantages. First, they think in code. They understand the underlying mechanics, and all the weaknesses of abstraction. When AI or Claude Code generates a program for you, it will always have bugs, imperfect architecture, things that aren’t quite right. People who truly understand the logic can fix these holes immediately.

Second, there are still many problems in software engineering that AI cannot solve. Why? Because those problems are outside the scope of their data. AI can handle binary sorting or reverse linked lists—dudes look at thousands of examples, so they’re experts. But if you start doing something truly novel, high-performance code, new architecture, or solving completely new problems? You still need hands-on engineers writing actual code.

This situation will continue until there are enough data points to train new models, or until AI becomes capable of higher-level abstraction and independent problem solving.

Now here’s the thing Naval Ravikant emphasizes—markets don’t want mediocrity. If there’s a superior app in any segment, nobody wants the average one. The winner takes almost 100% of the market share. So the bad news: second or third place is pointless. The good news: there are unlimited niches where you can become the best.

Naval has this principle: keep redefining what you do until you become the best at something. And Naval Ravikant still believes this principle applies in the AI era. You don’t need to worry about being replaced if you’re truly an expert in your field, no matter how niche. The opportunities are endless if you’re willing to find your own lane.