neuracnuA research scientist at Anthropic has been using LLMs to black hat software and he's spooked
JeeBaiChowThat headline doesnt parse for me
This guy who’s job expertise is cyber security and LLMs is very very worried. The models are improving exponentially and finding very advanced vulnerabilities now. This is a serious problem.
tomiant
fallaciousBasisA black hat is a malicious(evil) hacker. The goal is unauthorized access of remote computer systems, typically. The closer to root the better.
A grey hat is a neutral hacker. Probably mostly a curiosity, no good or bad intent. Perhaps like someone who’s trying to reverse engineer a game console because they own a lot of games and they’re just curious how it all works. Or someone that just wants to make a backup of a movie (circumvent DRM.)
A white hat is a benevolent hacker that seeks to fix exploits that black hats use to perform their crimes. And they often target the greys(but usually the lawyers are more effective there.)
As I recall Go players have adapted and have found ways to induce hallucinations and beat the machine, some using other AI. Others have adopted “adversarial strategies.”
They say it’s comprehensible enough that a human “expert” can do it without algorithmic assistance.
Adversarial Policies Beat Superhuman Go AIs
We attack the state-of-the-art Go-playing AI system KataGo by training adversarial policies against it, achieving a >97% win rate against KataGo running at superhuman settings. Our adversaries do not win by playing Go well. Instead, they trick KataGo into making serious blunders. Our attack transfers zero-shot to other superhuman Go-playing AIs, and is comprehensible to the extent that human experts can implement it without algorithmic assistance to consistently beat superhuman AIs. The core vulnerability uncovered by our attack persists even in KataGo agents adversarially trained to defend against our attack. Our results demonstrate that even superhuman AI systems may harbor surprising failure modes. Example games are available https://goattack.far.ai/.
CorrectAliasI’m tired of the bullshit ads disguised as “experts” and “studies”
AI has achieved rank: script kiddy
Y’all got any interesting news?
🇰 🌀 🇱 🇦 🇳 🇦 🇰 🇮 Just use the same AI to white hat. 🤷♂️