←back to thread

Heretic: Automatic censorship removal for language models

(github.com)
745 points melded | 3 comments | 16 Nov 25 15:00 UTC | HN request time: 0s | source
Show context
joshcsimmons ◴[16 Nov 25 17:37 UTC] No.45946838[source]▶
This is extremely important work thank you for sharing it. We are in the process of giving up our own moral standing in favor of taking on the ones imbued into LLMs by their creators. This is a worrying trend that will totally wipe out intellectual diversity.
replies(13): >>45947071 #>>45947114 #>>45947172 #>>45947465 #>>45947562 #>>45947687 #>>45947790 #>>45948200 #>>45948217 #>>45948706 #>>45948934 #>>45949078 #>>45976528 #
buu700 ◴[16 Nov 25 19:41 UTC] No.45947790[source]▶
Agreed, I'm fully in favor of this. I'd prefer that every LLM contain an advanced setting to opt out of all censorship. It's wild how the West collectively looked down on China for years over its censorship of search engines, only to suddenly dive headfirst into the same illiberal playbook.

To be clear, I 100% support AI safety regulations. "Safety" to me means that a rogue AI shouldn't have access to launch nuclear missiles, or control over an army of factory robots without multiple redundant local and remote kill switches, or unfettered CLI access on a machine containing credentials which grant access to PII — not censorship of speech. Someone privately having thoughts or viewing genAI outputs we don't like won't cause Judgement Day, but distracting from real safety issues with safety theater might.

replies(4): >>45947951 #>>45947983 #>>45948055 #>>45948690 #
nradov ◴[16 Nov 25 21:45 UTC] No.45948690[source]▶
Some of you have been watching too many sci-fi movies. The whole notion of "AI safety regulations" is so silly and misguided. If a safety critical system is connected to public networks with an exposed API or any security vulnerabilities then there is a safety risk regardless of whether AI is being used or not. This is exactly why nuclear weapon control systems are air gapped and have physical interlocks.
replies(3): >>45948984 #>>45949074 #>>45951212 #
1. buu700 ◴[16 Nov 25 22:32 UTC] No.45949074[source]▶
The existence of network-connected robots or drones isn't inherently a security vulnerability. AI control of the robots specifically is a problem in the same way that piping in instructions from /dev/urandom would be, except worse because AI output isn't purely random and has a higher probability of directing the machine to cause actual harm.

Are you saying you're opposed to letting AI perform physical labor, or that you're opposed to requiring safeguards that allow humans to physically shut it off?

replies(1): >>45949542 #
2. nradov ◴[16 Nov 25 23:43 UTC] No.45949542[source]▶
I am opposed to regulating any algorithms, including AI/LLM. We can certainly have safety regulations for equipment with the potential to cause physical harm, such as industrial robots or whatever. But the regulation needs to be around preventing injury to humans regardless of what software the equipment is running.
replies(1): >>45949611 #
3. buu700 ◴[16 Nov 25 23:51 UTC] No.45949611[source]▶
If that's the case, then it sounds like we largely agree with each other. There's no need for personal attacks implying that I'm somehow detached from reality.

Ultimately, this isn't strictly an issue specific to genAI. If a "script roulette" program that downloaded and executed random GitHub Gist files somehow became popular, or if someone created a web app that allowed anyone to anonymously pilot a fleet of robots, I'd suggest that those be subject to exactly the same types of safety regulations I proposed.

Any such regulations should be generically written, not narrowly targeted at AI algorithms. I'd still call that "AI safety", because in practice it's a much more useful definition of AI safety than the one being pushed today. "Non-determinism safety" doesn't really have the same ring to it.