This is kinda same thing I am going through tbh I build AI infras sometimes models and all but I kinda hate the generations like poetry, images and videos. Maybe feasible internet is the reason too that to settle with people perspective and that larping thing influencers just overhype things a lot, for example let's say web designing skills they built the skills they overhyped it and now every website kinda looks same. From normal user's perspective it's changed and all but from my pov I can just tell and it kinda diminshes the spark of mine for working such type of things so in a nutshell AI has made the creation easy but what about creativity
What strategy? The current strategy is like the strategy of a rocket sled.
Everyone compete with each other to compete with China to get a brass ring that's entirely unknown.
It's development at breakneck speed and blinders on for the sake of development (also known as IPO)
There are technical uses and there are social uses. The latter are by design echo chambers and often destructive of independent thinking and physically destructive as in "You are immortal, now go step in front of automobiles" (and this is fact, not fiction)
The difference between this and the development of the atomic bomb, which it eerily parallels, is that scientists assured themselves prior to its construction that it would not destroy the entire world in a chain reaction. That is not the case here.
I previously posted [1]
quote
I just listened to a video linked here [0] in which key points are that AI is reaching a point where it can be truly dangerous as in "overthrow the government" and that we need to nonetheless race development because CHINA is racing development, and YET, the more advanced our models are, the more advantageous it is for China (or others) to simply steal it.
So, in defending against an "enemy" we are also doing their work for them. Nice. We have been down this path before, but maybe this is more dangerous than nukes. Maybe not.
AND that security from leakage within AI companies, while extremely crucial, is secret and therefore unknown.
Given the value of the inside information, it could rival past government espionage scandals, and its status is completely known.
This all seems to boil down to security issues, and the very deliberate weakening of cyber defenses, along with incompetent political hacks in charge, and the sudden arrival of powerful vulnerability detection and exploitation, could take it all down in 2017 if not sooner, even without outside attack.
Ultimately, while LLMs will make a great number of jobs more _efficient_ (translators, programmers, lawyers), it will not replace _everyone_, only the surplus pushed out by the increased supply of labour-power for the jobs.
A machine can never be held accountable, therefore a machine cannot make business decisions.
To me, they are just tools right now. It just hasnt become anything more than that. There are some damn good tools out there, but you can see whats happening. They are just solving problems that have already beed solved. Just more Mario Kart-style crap, because its possible.
This is kinda same thing I am going through tbh I build AI infras sometimes models and all but I kinda hate the generations like poetry, images and videos. Maybe feasible internet is the reason too that to settle with people perspective and that larping thing influencers just overhype things a lot, for example let's say web designing skills they built the skills they overhyped it and now every website kinda looks same. From normal user's perspective it's changed and all but from my pov I can just tell and it kinda diminshes the spark of mine for working such type of things so in a nutshell AI has made the creation easy but what about creativity
What strategy? The current strategy is like the strategy of a rocket sled.
Everyone compete with each other to compete with China to get a brass ring that's entirely unknown.
It's development at breakneck speed and blinders on for the sake of development (also known as IPO)
There are technical uses and there are social uses. The latter are by design echo chambers and often destructive of independent thinking and physically destructive as in "You are immortal, now go step in front of automobiles" (and this is fact, not fiction)
The difference between this and the development of the atomic bomb, which it eerily parallels, is that scientists assured themselves prior to its construction that it would not destroy the entire world in a chain reaction. That is not the case here.
I previously posted [1]
quote
I just listened to a video linked here [0] in which key points are that AI is reaching a point where it can be truly dangerous as in "overthrow the government" and that we need to nonetheless race development because CHINA is racing development, and YET, the more advanced our models are, the more advantageous it is for China (or others) to simply steal it.
So, in defending against an "enemy" we are also doing their work for them. Nice. We have been down this path before, but maybe this is more dangerous than nukes. Maybe not.
AND that security from leakage within AI companies, while extremely crucial, is secret and therefore unknown.
Given the value of the inside information, it could rival past government espionage scandals, and its status is completely known.
This all seems to boil down to security issues, and the very deliberate weakening of cyber defenses, along with incompetent political hacks in charge, and the sudden arrival of powerful vulnerability detection and exploitation, could take it all down in 2017 if not sooner, even without outside attack.
[0] https://news.ycombinator.com/item?id=50522546
end quote
[1] https://news.ycombinator.com/item?id=50525469#50525957
please keep sharing this stuff every chance you get. Personally I expect the web as we know it to be unusable in early 2027 if nothing changes.
Ultimately, while LLMs will make a great number of jobs more _efficient_ (translators, programmers, lawyers), it will not replace _everyone_, only the surplus pushed out by the increased supply of labour-power for the jobs.
A machine can never be held accountable, therefore a machine cannot make business decisions.
To me, they are just tools right now. It just hasnt become anything more than that. There are some damn good tools out there, but you can see whats happening. They are just solving problems that have already beed solved. Just more Mario Kart-style crap, because its possible.
I'm so damn tired!