Bigger models require better alignment. So I would assume that the newer models will undergo more stringent alignment checks. For instance, I wouldn't put that much effort on GPT 4.
> So you are telling me Sol can't be used for hacking?
Who said anything about hacking? You can use it, but it is very restricted. But a much more capable model must have much higher restrictions on alignment. Obviously you wouldn't need to spend too much on alignment on a 3B gemma model. This must be obvious, I'm not sure why you bring this up?
> They didn't need to do that - they can just release when ready - nobody is expecting them to release a new model on a certain date.
Again a conspiracy theory? People expected a model during Dev Day. And they are just setting expectations. Whats with you guys and conspiracy theories?
I said if they were serious about 'safety' then given by their own graphs that Sol is less aligned than Astra, they should have pulled it.
As they haven't, the obvious conclusion is that safety isn't the main driver here.
I'm just following the logic - you are the one making assertions about motives that contradict the evidence.
Now I have some sympathy for the idea that these companies need to be responsible, but in the end - because these are universal machines - the regulation is going to be needed at the machine real world interface, not at the model itself.
This is, by the way, why OpenAI is responsible for the hacking that it's AI driven agents ( not models per se ), did.
It's not a problem to have an LLM that thinks it can drive, the problem is putting one it an actual car and putting it on the road.
Why would you expect something else?