November 1, 2025
Who needs a brain when you have a second monitor?
AI Broke Interviews
AI Broke Interviews: Script-reading candidates, whiteboard comebacks, and hiring chaos
TLDR: An industry insider says AI is wrecking coding interviews, from scripted answers to duplicate resumes, prompting calls for in-person testing. Commenters split: some want whiteboards and oral exams, others say you can spot cheaters online—either way, everyone agrees hiring needs a reset fast.
Software hiring just got messier. In a viral post, Yusuf Aytas says AI has “detonated” technical interviews—candidates read out flawless code and cookie‑cutter answers, sometimes freezing when asked a simple “what makes it interesting?” He even spotted four different CVs from the same person, and cites Google swinging back to in‑person interviews. The comments? Spicy. One camp says AI isn’t magic: ferrouswheel calls a lot of LLM output “slop,” pushing for sessions where candidates critique machine‑generated code live. Old‑schoolers cheer, with ForHackernews basically yelling “whiteboard is back!” Teachers like sega_sai report switching to oral exams that work—but burn everyone out. And yes, tools like InterviewCoder get name‑checked.
Another camp says we don’t need to drag everyone into offices: kace91 claims you can sniff out cheaters on Zoom—just ask them to lean back and talk through it. Meanwhile, greybeards like neilv reminisce about pre‑Google days when interviews were simple chats about real work. The meme energy is strong: picture three Spider‑Men—interviewers, AI cheat tools, and candidates—pointing at each other. The only agreement? The current game is broken. The real debate is what comes next: critique‑the‑bot, whiteboards, oral defenses, or back‑to‑basics storytelling. Grab popcorn; hiring season just turned into a reality show.
Key Points
- •The article argues that AI tools give candidates access to polished code, explanations, and behavioral answers during interviews.
- •Traditional software interview formats (LeetCode-style coding, behavioral, system design) are described as imperfect but previously functional filters.
- •Anecdotes include candidates reading AI-generated solutions without understanding and issues like subtle model-introduced code errors.
- •Screening challenges are reported, such as multiple distinct CVs submitted by the same individual, identified via a shared email.
- •The post states Google has returned to in-person interviews to counter widespread AI use and calls for rebuilding processes to measure genuine problem-solving.