Thread 514923681

THe Grief for Money from Big Corporation on A.I. LLM models stinks so much, it makes the use of said a.i. tierdsome.

>>514923681 (OP)
probably not, AI is in a huge bubble right now and as it turned out throwing more GPUs at it doesnt make the AI much smarter, the diminishing effect is already here.
they build massive AI datacenters everywhere but the usecase is still not clear and it still doesnt generate a profit, AI isnt replacing anyone and it sure wont replace every job like the hypemen said.
for that its just too stupid, cant be scaled up further, takes too much electricity, too much water and too much space.

>>514924001
you are wrong, corporation have their own interrest which ulitmative dictate the outcome, mostly influenced by govermants agenda.
Try to boil down the values of the west with the solution from the east on deepseek and watch the chat in front of your eyes being censored.

>>514924126
censorship is another reason why it wont suceed.
censorship makes the model dumber and more useless.
it also takes a lot of compute to ensure it stays retarded.

>>514923681 (OP)
bull signal js

>>514924001
LLMs are limited. AGI will be the most consequential achievement in human history. The energy problem will be solved, probably by AGI itself. There will be not a single thing that a human can do that AGI won't be able to do far better, faster, and cheaper. Nobody can predict when the crossover point of AGI being more intelligent than a human will be, it might not be in our lifetimes. Computers are less than 100 years old. But it will happen.

>>514924960
>AGI will be the most consequential achievement in human history
doesnt look like they are even close to it. it cant even tell you how many R's are in strawberrys
>The energy problem will be solved
it wont and its unlikely all these bottlenecks I listed will ever be overcome. especially not the ideological one where the AI is censored to protect jewish asses.
>it might not be in our lifetimes.
convenient for a jew who wants to steal your money today.

LLMs are not going to improve as fast moving fwd because they can only ingest so much data to train. Gen AI will requires intelligent people to incorporate into useful way to profit.
As for "replacement" layoffs, it is a fucking ruse. These companies have wanted to fire deadweight for ages. Now they have the almighty "fear of replacement" to hold over highly competant employees, preventing thrm from leaving. They are just being made to work multiple people's jobs now and longer hours to avoid loosing their job.

>>514925133
LLMs can't do a lot of basic stuff like counting r's in strawberries and won't take very many jobs. AGI isn't going to be an LLM.
It you think humans have peaked in technological achievement and won't go further. LLMs have only been around a few years. Think of what computers looked like 5 years after they were invented. AI technology will advance and change the world.

>>514923681 (OP)
You could have let an AI fix those spelling errors for you and not look like a total retard. But I guess AI is just too stupid, innit

>>514925776
>AGI isn't going to be an LLM
its a marketing term to get morons to invest money into the AI bubble.
then the AI bubble pops, there will be no AGI but the money is gone.

>>514925918
if your concern is more about the IQ of another person, what does it say about you?

>>514924960
>LLMs are limited. AGI will be the most consequential achievement in human history
Is this how big corpo fish for shareholders?