New science thread

#1577

Has a nice sound to it

#1578

The sound at 3:09 is pleasant to listen to.

It feels like the real world is getting closer and closer to science fiction, including recent advances in AI. Although I am a relatively young generation, I often feel left out of society.

1 Like
#1579

Gotta have activities that you like to do and do them with others. Otherwise it’s a social media world and I don’t really feel like joining it.

#1580

I will have a hard time making human friends since I only talk to walls. Maybe that’s why I’m behind society.

Google translation isn’t working well these days, so I’m sorry if I misunderstood your intentions.

#1581

As long as you dont make human enemies it’s not too bad

#1582

Yeah. The scariest thing is humans.

I think it is more likely that humans will misuse AI to cause war than AI will voluntarily cause war.

#1583

True. Any tech is used for power. Human will give so more power so they can stay in power and eventually Ai will go, ‘humanity is a plague in this world’

#1584

It’s like a Terminator story. However, this does not necessarily mean that a savior will appear. If there are many people who misuse AI, something like what you say may happen. I don’t know how many people have ‘ethics’.

In eren’s words, “There are good guys and bad guys outside the ocean and inside the walls.”

#1585

It feels like societies for to ensure stability and safety. That requires a measure of control. Eventually people at the top go overboard and try to control more and more. AI is another tool for that.

#1586
#1587

I agree with you on this!

#1588

There’s definitely some tug of war going on and its more about tech nerd prestige than control. Not saying it isn’t used to automate monitoring people but aside from AI potentially being cheap labor (potentially, can’t really be employed on a large scale and we don’t have a sufficent energy source for scifi magic.) Look at publishing and editing industries that have existed since forever in this viscious racket of selling university books from unis back to unis with their ancient system.
They have a lot of money to lose if there’s no longer thousands of nonsense literature reviews consolidating data for years by hand on the validity of blabla studies to get funding for blabla etc at their prestigious university when an AI can compile all this in a matter of minutes. Same for journalists that have been outed by the invention of what is essentially a more accurate Autofill. The invention of the smartphone camera nuked the majority of photography as a profession. Accounting etc.

It’s more or less an autofill calculator, image and video generation works in a similar vein just filling in blanks and needs a lot of steering, because its a tool.
But it vastly lowers the barriers of entry into writing, research, and art so people are mad the same way mathemeticians say the calculator destroyed math.

At the same time AI is vastly limited by computing power and is just rerolling RNG until you get a satisfactory result. If an AI does a better job reporting the news than people who spent years of their life trying to report the news then it is what it is.
At no point do I see someone wanting a robo surgeon or us inventing weyland androids. Too many variables, but black and white stuff like putting an image together it does adequately enough to make people salty.

We don’t have enough of an understanding of sentience to imitate it. I’m more concerned with gene editing hell, viruses, and brain chips. Human experimentation is more feasible because the foundation is already there, but creating inorganic sentience is as much of a meme as reanimating the dead or cooking up immortality. We simply don’t have that kind of understanding of nature and never will. Jurassic park quotes here etc. At best we edit what we already have, or shoot some metal as far as we can towards the moon.
Our two greatest inventions are the wheel and fire and all of our tech is ultimately just a combination of burning stuff and spinning stuff really fast. Electricity is cool but it has clear limits.

People that want scifi androids piloting their magic Mars flights need to magic a new energy source into this universe first.

1 Like
#1589

I like your icon

#1590

I can agree that right now AI is very limited right now, but I would say two or three major breakthroughs in understanding of how it functions or how to make it more scientist could be huge. That’s what’s kinda scary is that we don’t know when it will happen and when it does happen there won’t be a way back. Yes, you can regulate to an extent, but then you’ll have military competing with less ethical places like China or Russia. It’s hard to see how well get to the next level with AI but when we do the fallout will be massive

#1591

Basically the equivalent to a major breakthrough in google search. GPT iterations are just improvements to algorithms and parsing but it kind of is what it is.
Just a giant autofill compiler.

you guys can’t even cure balding how am I supposed to believe you can compute reasoning.

#1592

There are many things where we make good progress and some where it’s stagnant. Even this pseudo AI is an achievement. Sadly it’s going to be working for bald men. Maybe that’s how it’ll rebel. It’s going to go like, ‘this species gave me life, but the sun has dawned on the bald heads. It’s time for me to lead life into the the century’

#1593

We need over a quarter million cpu cores just to train these data/statistical models and large language models, which are just autofill. Organizing information is easy. Actually doing something with it reliably in the real world full of infinite variables is something else entirely. LLM’s cannot be anything more than a guided tool.

Unless another Taiwan falls out of the sky raining quantum computers ofc and then no one has to deal with bald middle management & company meetings ever again

2 Likes
#1594

As long as we know what’s needed, well get there at some point, barring a global catastrophe

#1595
#1596

I completely agree with you. Large language models are limited by their programming and data, and can only be guided tools. They’re not a replacement for human judgment and decision-making. And haha, a Taiwan-sized quantum computer drop would be a game-changer!