4 ms·
The Claude Mythos / Project Glasswing thing is real: https://www.anthropic.com/glasswing https://www.anthropic.com/glasswing I met a few people at PyCon this w
by simonw 6mo ago
The Claude Mythos / Project Glasswing thing is real: https://www.anthropic.com/glasswing https://www.anthropic.com/glasswing
I met a few people at PyCon this week who have been part of Glasswing (they're just starting to be allowed to talk about it) and it really does drive down the cost of finding vulnerabilities.
I've been collecting notes on that here: https://simonwillison.net/tags/ai-security-research/ https://simonwillison.net/tags/ai-security-research/
- krzyk 6mo agoPeople in my company sounded underwhelmed by it. It usually was founding issues by not understanding deployment (or not being fed that info).
- halflife 6mo agoA friend of mine had hands on experience, it’s not the intelligence of it, it’s the speed. You used to have a couple of days to close a breach, now it 2 hours.
- vessenes 6mo agoYou might point them at the cloudflare blog about deploying mythos - I found it interesting. Upshot — as your folks discovered, deployment, harness, and utilization method matters for mythos and is a bit different than how you’d deploy a coding agent for writing code, but if you do that, you get something capable of end to end chaining and reasoning about a much broader class of vulnerabilities. No personal experience with it. But the security team writeups I’ve read are significantly more positive about it than you describe, so it might be worth a second look.
- Gigachad 6mo agoWouldn't it drive up the cost of finding vulnerabilities when all the low hanging fruit has already been scanned and patched? Like the new baseline for finding a vulnerability will be something an LLM couldn't find.