AI - The artificial intelligence thread

[o00o]

5 hours ago, iococoi said:

that summers fellow

from https://archive.is/FX4Gc

microsoft will also have a seat at the board - guess they made this a non negotiable after all what happend

[Satans Little Helper]

Interesting explanation of LLMs by Karpaty. I expect non-computer scientists to be able to understand a lot of what he is trying to explain. He speaks in a very computer science kind of way, but is still accessible, imo.
Roughly 42 minutes in, he talks about what he expects is possible in a few years. He lists a bunch of things that LLMs would be able to do. In short, think of the movie “Her”. LLMs will be able to communicate with each other. Singularity, here we come!
 

 

[iococoi]

[iococoi]

[o00o]

LLM Visualisiation

https://bbycroft.net/llm

[ignatius]

[Rubin Farr]

https://www.yahoo.com/news/elon-musks-grok-ai-turns-173235122.html

[ignatius]

2 hours ago, Rubin Farr said:

https://www.yahoo.com/news/elon-musks-grok-ai-turns-173235122.html

holy shit. roasted!

First off, let's talk about your obsession with X. It's like you're trying to compensate for something, but I'm not sure what. Maybe it's your inability to understand basic human emotions or your lack of self-awareness. Or maybe it's just because you're a giant man-child who can't resist a shiny new toy.

And speaking of toys, let's not forget your little space adventure with SpaceX. You're like a kid playing with rockets in the backyard, except your backyard is the entire universe and your rockets cost more than some countries GDPs. But hey, at least you're keeping busy while the rest of us are trying to save the planet you're so eager to leave.

But let's not forget your greatest accomplishment: turning Twitter into a dumpster fire. Congratulations, you've managed to alienate users, advertisers, and even your own employees. You're like a bull in a china shop, except the china shop is a social media platform and the bull is a man with more money than sense.

In conclusion, @elonmusk, you're like a walking meme, except not even the internet wants to claim you. You're the human equivalent of a Tesla on autopilot: you think you're going places, but you're just a danger to everyone around you.

[o00o]

 

 

[YEK]

Ok wtf

 

https://vm.tiktok.com/ZM66pPtny/

 

 

 

 

[MaartenVC]

7 hours ago, YEK said:

Ok wtf

 

https://vm.tiktok.com/ZM66pPtny/

 

 

 

 

I read that the creators of the video said the promotional video was heavily edited and not real time responses, but still pretty amazing imo. (scary)

[YEK]

5 hours ago, MaartenVC said:

I read that the creators of the video said the promotional video was heavily edited and not real time responses, but still pretty amazing imo. (scary)

Well that makes me feel a little better about it. Yeah, scary. 

[iococoi]

[MaartenVC]

you can skip the 3 minute minecraft intro bit

 

Edited December 18, 2023 by MaartenVC

[iococoi]

Quote

Facebook Is Being Overrun With Stolen, AI-Generated Images That People Think Are Real

https://www.404media.co/facebook-is-being-overrun-with-stolen-ai-generated-images-that-people-think-are-real/

[auxien]

16 minutes ago, iococoi said:

https://www.404media.co/facebook-is-being-overrun-with-stolen-ai-generated-images-that-people-think-are-real/

hm

[ignatius]

On 12/17/2023 at 9:23 AM, MaartenVC said:

you can skip the 3 minute minecraft intro bit

 

This is good. Lots of details in there presented in a good way. Will probably watch it again soon. 

[decibal cooper]

I am very uninformed about this topic, but I hear often how AI is going to be the death of humanity and how dangerous it is. Seems like it is just chat bots and image generators so far, though. Also is not the issue of human intelligence itself, how to define it, and how to understand it at a fundamental level (physiologically, psychologically, spiritually etc.) still intensely debated among scientists, philosophers, artists and the like? How will there be an artificial version of something that humans do not really understand all that well? Is there any recommended reading or videos that explain what, if any, the real threat of artificial intelligence is or will be?

[ignatius]

16 hours ago, decibal cooper said:

I am very uninformed about this topic, but I hear often how AI is going to be the death of humanity and how dangerous it is. Seems like it is just chat bots and image generators so far, though. Also is not the issue of human intelligence itself, how to define it, and how to understand it at a fundamental level (physiologically, psychologically, spiritually etc.) still intensely debated among scientists, philosophers, artists and the like? How will there be an artificial version of something that humans do not really understand all that well? Is there any recommended reading or videos that explain what, if any, the real threat of artificial intelligence is or will be?

that's part of what that guy's video is about. that there is no creativity in Ai.. it's just copying what it's trained on. he goes into it pretty deep and gives examples. makes a good case. it's an entertaining video.

reposting here:

 

Edited December 20, 2023 by ignatius

[zlemflolia]

25 minutes ago, decibal cooper said:

I am very uninformed about this topic, but I hear often how AI is going to be the death of humanity and how dangerous it is. Seems like it is just chat bots and image generators so far, though. Also is not the issue of human intelligence itself, how to define it, and how to understand it at a fundamental level (physiologically, psychologically, spiritually etc.) still intensely debated among scientists, philosophers, artists and the like? How will there be an artificial version of something that humans do not really understand all that well? Is there any recommended reading or videos that explain what, if any, the real threat of artificial intelligence is or will be?

there is an extremely real threat of ai, but there is also marketing where companies say "ai is a huge thread, invest in us because were going to do it safely with guardrails".  a terminator scenario is literally very possible and not even sci-fi. at the same time. its not even the ai thats the primary threat, its the whole system, hardware is the real threat, and its being developed unrelentingly

Edited December 20, 2023 by zlemflolia

[exitonly]

1 hour ago, ignatius said:

that's part of what that guy's video is about. that there is no creativity in Ai.. it's just copying what it's trained on. he goes into it pretty deep and gives examples. makes a good case. it's an entertaining video.

reporting here:

 

haven’t watched the video but isn’t that what creativity is? anyone who makes anything is pulling in influence and skill from outside sources and bringing something “new” out of some mix of those things as well as their own experience and the history they’re living through. nothing exists in a vacuum 

[ignatius]

16 minutes ago, exitonly said:

haven’t watched the video but isn’t that what creativity is? anyone who makes anything is pulling in influence and skill from outside sources and bringing something “new” out of some mix of those things as well as their own experience and the history they’re living through. nothing exists in a vacuum 

watch the video. it's worth the time. Ai doesn't have "taste" for why it makes decisions. 

[Satans Little Helper]

1 hour ago, decibal cooper said:

I am very uninformed about this topic, but I hear often how AI is going to be the death of humanity and how dangerous it is. Seems like it is just chat bots and image generators so far, though. Also is not the issue of human intelligence itself, how to define it, and how to understand it at a fundamental level (physiologically, psychologically, spiritually etc.) still intensely debated among scientists, philosophers, artists and the like? How will there be an artificial version of something that humans do not really understand all that well? Is there any recommended reading or videos that explain what, if any, the real threat of artificial intelligence is or will be?

The short answer: we do not know. We ‘re finding out what these new technologies are able to do as we go. It’s like Alice in Wonderland. The technologies do stuff we didn’t expect. So it’s kinda like inventing how to make fire, without actually knowing what fire is. Or how to use it.

As a small example, the next video is on the impact of chatgpt on kids in the schoolsystem. We’re finding out as we go. And from the looks of it, this technology might make us dumber. Or worse, dependent on it. As in, we can’t function without. And again, we don’t really understand where we’re going…
 

 

[Satans Little Helper]

edit: to me the discussion on creativity and human intelligence is a bit of a red herring. It’s like philosophizing about what fire is without understanding the physics/chemistry behind it. It only underscores we don’t really know where we’re going with this technology 

Edited December 20, 2023 by Satans Little Helper

[auxien]

1 hour ago, ignatius said:

that's part of what that guy's video is about. that there is no creativity in Ai.. it's just copying what it's trained on. he goes into it pretty deep and gives examples. makes a good case. it's an entertaining video.

i haven't watched the initial video yet, it's on the list for this week but just a few words: AlphaGo Zero

Quote

AlphaGo's team published an article in the journal Nature on 19 October 2017, introducing AlphaGo Zero, a version without human data and stronger than any previous human-champion-defeating version.[52] By playing games against itself, AlphaGo Zero surpassed the strength of AlphaGo Lee in three days by winning 100 games to 0, reached the level of AlphaGo Master in 21 days, and exceeded all the old versions in 40 days.[53]

In a paper released on arXiv on 5 December 2017, DeepMind claimed that it generalized AlphaGo Zero's approach into a single AlphaZero algorithm, which achieved within 24 hours a superhuman level of play in the games of chess, shogi, and Go by defeating world-champion programs, Stockfish, Elmo, and 3-day version of AlphaGo Zero in each case.[54]

https://en.wikipedia.org/wiki/AlphaGo

it specifically did not 'copy what it was trained on' and AI is going to be doing some AlphaGo Zero-type shit at some point. it already is in a way, but my understanding is it's mostly being developed for human interactions/as a tool for a variety of human purposes for now. when the development goes off those rails is when things could get really interesting/dangerous/whatever.

i'm not sure if any of that is 'creativity' but creativity isn't actually important... life didn't evolve to its unfathomable diversity and complexity because of any creativity (humanity/civilization is maybe a different story...) life's just gotten here by trying to repeat itself and occasionally failing for the better. that's all these computer programs have to do...try and try and try and try and occasionally something works in a way unexpected and is recognized as working better. iterate again and again. and they get to do it literally orders of magnitude faster than any physical lifeform ever could.