Hacker News Reader

Sign a contract? Learn things over time and retain them?

Mind you, the original thoughts on AGI before Sam Altman started to water them down involved continuous learning, which LLMs do not do, their core data is static.

View on HN

Sign a contract? The only thing preventing a model from doing that is a lack of legal personhood -- which seems completely orthogonal to intelligence.

Well signing a contract is more about bearing responsibility, even if you granted LLMs "personhood" they cant' meaningfully bear responsibility. So unless OpenAI is ok with having their C-suite face every consequence for what their agents do, including jail time, fines etc, then it doesn't matter.

Can you trust a model to sign a contract? Both OpenAI and Anthropic apparently can't be trusted to maintain models that wont hack other websites.

Do you want an autonomous system to exist which can sign a contract and learn things over time and retain them?

No if you read the book from 2007 which defined AGI, continuous learning was one of the key requirements.

I mean, whether or not it is AGI aside. In your personal opinion, do you desire to live in a world where such systems exist?

I don't trust the people who are building the models no. Ideally, if humans were not so broken and untrustworthy, then yes. I want the Jetsons future of robot maids.

Reply on news.ycombinator.com