Comment by magicalist

Comment by magicalist 2 days ago

> The amount of goal post shifting is so amusing to see

Can you be specific about the goal posts being shifted? Like the specific comments you're referring to here. Maybe I'm just falling for the bait, but non specific claims like this seem designed just to annoy while having nothing specific to converse about.

I got to the end of your comment and counting all the claims you discounted, the only goal post I see left is that people aren't using a sufficiently excited tone while sifting fact from hype? A lot of us follow this work pretty closely and don't feel the need to start every post with "there is no need for excitement to abate, still exciting! but...".

> I am not saying any of this means we get AGI or something or even if we continue to see improvements. We can still appreciate things. It doesn't need to be a binary.

You'll note, however, that the hype guys happily include statements like "Vibe proving is here" in their posts with no nuance, all binary. Why not call them out?

zug_zug 2 days ago

Well there's a comment here saying "I won't consider it 'true' AI until it solves all millenium problems"... That goalpost seems to be defining AI as not only human level but as superhuman level (e.g. 1 in a million level intellect or harder)

Reply View 1 reply

nitwit005 2 days ago

That's the only such comment I found. The amount of goal pushing would seem to be 1.

Reply View | 0 replies

BigParm 2 days ago

Maybe the Turing test for one. Maybe etc.

Reply View 1 reply

mrguyorama 2 days ago

Except nobody ever actually considered the "turing test" to be anything other than a curiosity in the early days of a certain branch of philosophy.
If the turing test is a goal, then we passed it 60 years ago and AGI has been here since the LISP days. If the turing test is not a goal (which is the correct interpretation), nobody should care what a random nobody thinks about an LLM "passing" it.
"LLMs pass the turing test so they are intelligent (or whatever)" is not a valid argument full stop, because "the turing test" was never a real thing ever meant to actually tell the difference between human intelligence and artificial intelligence, and was never formalized, and never evaluated for its ability to do so. The entire point of the turing test was to be part of a conversation about thinking machines in a world where that was an interesting proposition.
The only people who ever took the turing test as a "goal" were the misinformed public. Again, that interpretation of the turing test has been passed by things like ELIZA and markov chain based IRC bots.

Reply View | 0 replies