wewbull

The standard library is where project go to die.

Computer programming, regardless of language, is hard. The computer does exactly what you tell it to.

...and people worry about the name of a git branch.

I don't view free-use models as open-source. Open-source means I can rebuild it from scratch and I can't because I don't know what the training data is, or have access to it.

The scale is the difference and who is harmed.

Billion dollar company losing $100. Who cares?!

Billion dollar company stealing from all artists in the world. We care.

Points 2 and 3. Basically make restrictions on normal user accounts which are fine for humans but that will make bots swear and curse.

Unless you mean "what should the registration process be" I think API keys via a user account would do.

...but if they don't know I expect them to say so. An LLM isn't trustworthy until it says "I don't know".

Exactly the reason I suggest it.

Is that an armadillo? Forgetting how my own code works is my forte.

An LLMs "intent" is always to give you a plausible response even if it doesn't have the "knowledge". The same behaviour in a human would be classed as lying IMHO.

1. Make bot accounts a separate type of account so legitimate bots don't appear as users. These can't vote, are filtered out of post counts and users can be presented with more filtering option for them. Bot accounts are clearly marked.

2. Heavily rate limit any API that enables posting to a normal user account.

3. Make having a bot on a human user account bannable offence and enforce it strongly.

Agreed, It wouldn't be a good thing. However it's their own failures and mismanagement that are causing it.