Quick Hits: June 29, 2026

1. GLM 5.2 Takes the Lead in Cyber Benchmarks
What happened: GLM 5.2 outperformed Claude in our latest cyber security benchmarks, scoring a solid 716 and sparking 344 comments on Hacker News.
Why it matters: If you’re in the market for a new AI to bolster your security, GLM 5.2 might just be the “mythos at home” you’ve been waiting for. Claude, consider yourself warned.

2. Claude Code Diagnoses MRIs (Sort Of)
What happened: A brave soul used Claude Code to get a second opinion on their MRI, and the internet went wild with 527 comments.
Why it matters: While it’s tempting to let AI take a peek at your medical scans, remember: it’s still a bit like asking a calculator to perform surgery. Proceed with caution.

3. Professor Calls Out AI Exam Fraud at Brown
What happened: An academic at Brown University publicly denounced a wave of AI-assisted cheating on exams, sparking a debate on academic integrity with 487 comments.
Why it matters: As AI gets better at everything, from writing essays to acing exams, universities are scrambling to keep up. The future of education might just depend on who can outsmart the bots.