Ask Anything on Yik Yak | AI is the biggest tech grift of all time. Does it work? Sure. But it should’ve stayed in research phase for 4 more years, the. Beta for an additional 4 years. Instead it’s going to collapse the global economy and remove a generation from the workforce.

Yik Yak

Download

Join communities on Yik Yak Download

Ask Anything 1d

AI is the biggest tech grift of all time. Does it work? Sure. But it should’ve stayed in research phase for 4 more years, the. Beta for an additional 4 years. Instead it’s going to collapse the global economy and remove a generation from the workforce.

Anonymous 1d

And the economy and jobs are feeding it just to fuck itself and get replaced

Anonymous 1d

Ai IS in research phase and people using it is part of how it gets improved. The government and companies have been researching AI and LLMs for decades already. This isn’t new just because your barely learning about it its been a thing. It’s largely being researched and improved by people using it thats why AI is operating at a loss for people to use because it needs people to use it to improved.

Anonymous 1d

My understanding is that modern AI requires large sums of money to fund to make more data centers. The main discovery in the field of AI over the last years is that additional scaling of compute, and number of parameters consistently decreases the disconnect for reality with models. I agree that the underlying technology is solid but it requires huge investment to improve models. Additionally if the technology had stayed in the lab for an additional 4-8 years only a few tech monopolies like…

Anonymous

#1 1d

My theory is that the government hyping it up solves two issues: building the surveillance state and holding up the economy from depression - which should’ve happened pre COVID.

Anonymous

#2 1d

Google and potentially Meta would have the funding to control a seemingly very important technology. I would also argue that the AI models are useful enough now to generate significant economic value. Right now the domain of coding is the most obvious area where value can be generated but I have personally found it to become more helpful in less verifiable domains as well. I understand that is anecdotal evidence but I think the general pace of progress is faster than needing an additional 4-8…

Anonymous

#2 1d

years. I believe releasing the current AI models to the public right now will also be beneficial because it will allow people to adapt to the technology while it is still imperfect and understand that where AI is useful and where it is less useful.

Anonymous

#1 1d