4 ms·
I wonder if this is the reason that ChatGPT is so heavily censored, and that Google is so nervous to make LLMs publicly available. These LLMs are clearly capabl
by Taek 4y ago
I wonder if this is the reason that ChatGPT is so heavily censored, and that Google is so nervous to make LLMs publicly available. These LLMs are clearly capable of instructing users how to make bombs, how to hack websites, they are clearly capable of producing copyright infringing images and bodies and probably bodies of text as well.
As far as I can tell, it's not obvious at all that output from an LLM would be protected by ChatGPT, and Congress is clearly resistant to setting up something similar for a new technology. People already feel that Section 230 is the enemy of justice and safety. I think it's rather likely that case law would intentionally side against LLMs if there is any grey area at all, and there seems to be plenty of grey area.
I wonder if hosted LLMs are going to become legally non-viable in the near future. Everyone will need to run their own at home simply because of legal liability.
- cma 4y agoWouldn't the same concerns apply to translations and AI generated snippits? They already ship those.
- inglor_cz 4y agoBuilding a bomb according to ChatGPT spec would be the ultimate Darwin Prize. The system spews so much nicely sounding bullshit that anyone who trusts it to produce a safe bombmaking recipe is a fool.
- jadbox 4y agoI wonder if the US legal system may consider that GPT AI could be viewed as a person exercising their free speech, similar to how we have limited corporate personhood?
- Animats 4y agoIt's not like it's hard to find out how to make bombs any more. The classic "Anarchist's Cookbook", from the 1960s, is on line as a Google scan. "Prepper" literature is full of that stuff. There are even old WWII training videos on the Internet Archive. ChatGPT isn't contributing to that problem.