I fail to see how this particular "anime girl" and the potential for clients seeing it, could make you think that's a fair request. That seems extremely ridiculous to me.
Safety in the context of LLMs means “avoiding bad media coverage or reputation damage for the parent company”
It has only a tangential relationship with end user safety.
If some of these companies are successful the way they imagine, most of their end users will be unemployed. When they talk about safety, it’s the companies safety they’re referring to.
Investor safety. It's amazing that people in hn threads still think the end-user is the customer. No. The investor is the customer, and the problem being solved for that curtomer is always how to enrich them.
It feels hard to include enough context in the system prompt. Facebook’s content policy is huge and very complex. You’d need lots of examples, which lends itself well to SFT. A few sentences is not enough, either for a human or a language model.
I feel the same sort of ick with the puritanical/safety thing, but also I feel that ick when kids are taken advantage of:
I also don't get it. I mean if the training data is publicly available, why isn't that marked as dangerous? If the training data contains enough information to roleplay a killer or a hooker or build a bomb, why is the model censored?
If you don’t believe that you can be harmed verbally, then I understand your position. You might be able to empathise if the scenario was an LLM being used to control physical robotic systems that you are standing next to.
Some people can be harmed verbally, I’d argue everyone if the entity conversing with you knows you well, and so i don’t think the concept of safety itself is an infantilisation.
It seems what we have here is a debate over the efficacy of having access to disable safeguards that you deem infantilising and that get in the way of an objective, versus the burden of always having to train a model to avoid being abusive for example, or checking if someone is standing next to the sledgehammer they’re about to swing at 200rpm
The first one has been argued against quite nicely by Piketty, it's how you get plutocracy
The three other ones should not be treated as Rights since the concerned individual is no more, and they don't matter much anyway if coming against the rights of people (that means "living"). For instance collecting organs for the good of those who need, when evaluated, should trump any opposition on frivolous grounds.
I'm indead asking if the whole concept is not wrong and deeply harmful to societies
Unless someone is hurt you can believe what you want. Otherwise it's necessary to weight what's to be gained and lost by entertaining net negative stances on frivolous grounds; and why we should then chose to do so.
America is a common law jurisdiction. Juries determine the true law and you never know what that is until a jury decides. You can make so good guesses though. Custom is the rule so an autocar doing what people normally do is the best way to have it behave.
Not only in science.
Even (or especially) in impartial journalism or public debate, there's no valid reason why any unfounded (thus illegitimate) opinions should get as much consideration as sound and researched arguments.
Are you being purposefully controversial (to not say trollish)?
To the exact contrary to what you assert, one of the prominent argument against Gnome that I've been seing times and times again in DE debates, is the "dogmatic" opposition to SSD from the Gnome project.
