News for hackers. Stuff that matters.
Re:I played chess against ChatGPT-4 and lost by orbital-decay on Sunday March 19, 2023 @09:39AM (#35218794)
Simple storage is unable to apply it, it can only give the data to you verbatim. The model is able to make actual predictions based on it.

But yeah, the "knowledge" and "understanding" are hard to define formally, so this discussion can be endless. Common well-defined terms are required.

Re:I played chess against ChatGPT-4 and lost by 2-718-281-828 on Sunday March 19, 2023 @10:49AM (#35219493)
the model does not extract knowledge. an external algorithm trains the models parameters and then the model is fed a string that is also evaluated externally based on the models configuration.
Re:I played chess against ChatGPT-4 and lost by orbital-decay on Sunday March 19, 2023 @06:53PM (#35224582)
Semantics. You could say the same about the disk - the data doesn't get magically teleported from the magnetic plates to the RAM, it needs a lot of underlying hardware to read and transfer it.

Model is not just a set of weights, it's inseparable from the underlying architecture, the way to train and to apply them in practice.

Re:I played chess against ChatGPT-4 and lost by 2-718-281-828 on Sunday March 19, 2023 @08:11PM (#35225323)
> You could say the same about the disk

that's exactly my point

Re:I played chess against ChatGPT-4 and lost by orbital-decay on Sunday March 19, 2023 @11:22PM (#35226876)
Yeah, and that's arguing about semantics. You could do that in a loop endlessly, ignoring the fact they're fundamentally different.