I don't see why'd you have to prove humans are anything at all, to validate the claim that GPT is a word predictor. ChatGPT doesn't really need defending, the proof is in it's massive success.. right?
It seems the news cycle has settled into two possible options for future code releases. It's either the second coming of Christ (hyperbolically speaking) or it's an overly reductive definition of GPT's core functionality.
I can't help but be reminded of the first time the iPod came out [0] and the Slashdot editor of the time, dismissed it out of hand completely.
My point isn’t that LLMs are anything more than pattern predictors; it’s that calling them such as some sort of dismissal doesn’t really strike me as the “gotcha” it initially seems.
We don’t know that humans themselves aren’t just prediction machines. Yet , humans are insanely capable! And thus the same might apply to an LLM.
It’s hard to strike a balance between having excited, rational, discussion and not coming across like a religious AI nut.
Oohh it always struck me as the comparison to humans was to marginalize humans.
I'm often on the reductive "gotcha" side and it is always a rummaging around to find and understand the essence of the thing in front of me causing all this news and hype. Peel open the black box and see it's basic shape, as much as one can from the outside..
Thank you for the insight, I genuinely have been annoyed at people who are "dragging down The Human Consciousness to the level of basic boolean logic", when they compare computing to humans and I never got the other side of it.
It seems the news cycle has settled into two possible options for future code releases. It's either the second coming of Christ (hyperbolically speaking) or it's an overly reductive definition of GPT's core functionality.
I can't help but be reminded of the first time the iPod came out [0] and the Slashdot editor of the time, dismissed it out of hand completely.
[0] https://slashdot.org/story/01/10/23/1816257/apple-releases-i...