How do you interview devs in a post-AI world?

Since the rise of AI coding assistants, about 80% of the dev candidates that I interview tell me that they aren't writing much code themselves anymore - they are directing agents instead. This makes me deeply uncomfortable (although maybe I'm just being old-fashioned). I still want to know that devs on my team can actually write code themselves, and have some idea of what they are doing when it comes to system design. This is increasingly challenging given how rapidly the AI tools are evolving.I'm curious what other folks are seeing in a post-AI hiring landscape and how you are approaching candidates that tell you they are fully agent-pilled when it comes to their development process. Should I keep doing the conventional leetcode and design interviews? Or just give up and expect everyone is going to use Claude Code no matter what?

Hacker NewsToolAISource
0Sign in to voteCopy link

FL score

42

out of 100

Verdict

SKIP

high confidence

Competition

No competitor data yet

Trend

No signal yet

A hiring manager is uncomfortable that AI-assisted developers no longer write code by hand, but no one has figured out how to assess real coding ability in a post-AI world.

The pain

Hiring managers cannot tell if candidates can actually code or just prompt-engineer. Candidates are increasingly relying on AI agents for all development work. Traditional leetcode and design interviews feel outdated but nothing has replaced them. The uncertainty creates hiring risk.

The gap

No tool or framework exists that reliably measures coding ability in an AI-assisted world. The post is a question, not a solution. Existing assessment platforms (HackerRank, Codility, Interviewing.io) have not adapted their approach. The gap is real but the solution is undefined.

Build angle

You could build an interview platform that forces candidates to code without AI assistance, or one that evaluates how well they use AI tools, or one that tests system design thinking instead of syntax. But each approach has different customers and different willingness to pay.

Strengths

  • The problem is growing and will only get worse as AI tools improve
  • Hiring managers are actively struggling with this right now, not in theory
  • The pain is concentrated in tech companies with real hiring budgets
  • Early movers could establish credibility in a new category

Risks

  • No consensus on what good assessment looks like in an AI world, so product-market fit is unclear
  • Hiring managers may just change their interview process themselves rather than buy a tool
  • Candidates will find ways to use AI during any assessment, making enforcement hard
  • Existing assessment platforms have distribution, brand, and customer relationships
  • The market is small, limited to tech hiring at companies that care about this specific issue
  • The problem may self-correct as companies accept that AI-assisted development is the new normal

Questions about this idea?

FlyBot reads the scoring and gives you a second opinion on “How do you interview devs in a post-AI world?”.

Open FlyBot