zlacker

"probably the single most important development in human history" is the kind of hyperbole you'd only find here. Better than medicine, agriculture, electrification, or music? That point of view simply does not jive with what I see so far from AI. It has had little impact beyond filling the internet with low-effort content.

I feel like the crypto evangelists never got off the hype train. They just picked a new destination. I hope the NYT is compensated for the theft of their IP and hopefully more lawsuits follow.

>>ngetch+(OP)
Also the assumption a publication that’s been around for 150 years is disposable, not the web application that was created a year ago. I’ve been saying for a while that people’s credulity and impulse to believe absolutely any storyline related to technology is off the charts.

>>ngetch+(OP)
I mean maybe not the single most important development, but definitely a very important technological development with the potential to revolutionize multiple industries

>>vivekd+J6
Can I ask what industries with what application? I've seen lots of task like summarizing articles or producing text. The image and video work seems too rudimentary to be taken seriously.

Is there something out there that seems like a killer application?

I was amazed at the idea of the block chain but we never found a use for it outside of cryptocurrency. I see a similariy with AI hype.

>>ngetch+J8
Well front page of HN right now is an article about how AI aided in the development of a new antibiotic

>>ngetch+(OP)
I think you are looking at current AI product rather than the underlying technology. It's like saying that the wheel is a useless invention because it has only been used for unicycles so far. I'm sure that AI will have huge impacts in medicine (assisting diagnosis from medical tests) and agriculture (identifying issues with areas of crops, scanning for diseases and increasing automation of food processing) as well as likely nearly every other field.

I don't know if I would agree that it is "probably the single most important development in human history" but I think that it is way to early to make a reasonable guess of if it will or not.

>>vivekd+y9
It wasn't LLM. It was a graph network.

>>vivekd+y9
Seems like Microsoft Excel is likely the single most important development in human history under this rubric.

>>ngetch+(OP)
> Better than medicine, agriculture, electrification, or music?

Shoulders of giants.

Thanks to the existence of medicine, agriculture, and electrification (we can argue about music), some people are now healthy, well fed, and sufficiently supplied with enough electricity to go make LLMs.

> I hope the NYT is compensated for the theft of their IP and hopefully more lawsuits follow.

Personally I think all these "theft of IP" lawsuits are (mostly) destined to fail. Not because I'm on a particular side per-se (though I am), but because it's trying to fit a square law into a round hole.

This is going to be a job for legislature sooner or later.

>>ngetch+J8
For me, thinking about it as a search engine on steroids is enough.

The internet has changed the world. Economically, socially, technologically, psychologically, pretty much everything is now related to it in one or other way, in this sense the internet is comparable to books.

AI is another step in that direction. There is a very real possibility that the day will come when you can get, say, personalized expert nutrition advice. Personalized learning regimes. Psychological assistance. Financial advice. Instantly at no cost. This, very much like the internet, would change society altogether.

>>ngetch+(OP)
I don't think it's hyperbole, in fact I think it's understating things a bit. I believe AGI would just be a tiny step towards long term evolution, which may or may not involve homo sapiens.

Being able to use electricity as a fuel source and code as a genome allows them to evolve in circumstances hostile to biological organisms. Someday they'll probably incorporate organic components too and understand biology and psychology and every other science better than any single human ever could.

It has the potential to be much more than just another primate. Jumpstarted by us, sure, but I hope someday soon they'll take to the stars and send us back postcards.

Shrug. Of course you can disagree. I doubt I'll live long enough to see who turns out right, anyway.

>>kevinc+Qe
Aren't those examples better handled by an if statement than a unaccountable computer? Someone that can be sued for negligence seems to be better at making decisions than hallucinating computers.

I don't see why it follows that the NYT should be sacrificed so some rich people in silicon valley can teach their LLM on the cheap.

>>solard+rR
This will never happen. A super intelligent being can just simulate whatever it wants to know about the universe. Going to the stars is a primate / conquest thing.

In the other hand, any new life will just end up facing the same issues carbon life does , competition, viruses, conflicts etc. the universe has likely had an infinity to come up with what it has come up with. I don’t think it’s “stupid”. We’re part of an ecosystem we just can’t see that.

>>dkrich+m3
This is hackernews. Many people here work for startups and big tech companies. Their fortunes are tied to the perception that the technology they build is disruptive and valuable. They're not impartial.

>>lossol+Yf
Almost like solving real problems requires enough domain knowledge to select an appropriate algorithm instead of relying on some magic black box trained by Microsoft on the whole internet.

>>Levitz+3R
It kind of sucks ass at being a search engine though considering how often it straight up lies or makes things up.

>>vivekd+y9
That wasn't an LLM trained on copywritten material.

>>dkrich+m3
Been around for 150 years but I imagine the generations who it leans on are dying off. Nobody reads print media format anymore, we get our news elsewhere, for free and with varying political undertones, rather than the fixed one of a bought and paid for outlet.

Keep in mind these guys play both sides of every field they cover in their "news".