This is why I keep a good chunk of my steam library downloaded locally and don’t buy always-online required games.

Image recognition AI has been found to be vulnerable to prompt injection via text present in the image.

Nepenthies is a type of webpage that generates an endless chain of links and pages to bait in content scrapers and feed them garbage that poisons their training data.

If it makes you feel better that AI is almost certainly doing a horrible job that will burden your former employer with unfixable tech debt in just a few years.

You could piss off the company running them though, don’t these cameras use AI? Get some QR codes that lead to nepenthies traps full of their own AI generated images of nonsense license plates. Waste their money and poison their training data.

I have FTL and Into The Breach installed so even if the rest of my library vanishes I think I’ll be alright.

iirc the power is not very doable, You’d need hundreds of times as many solar pannels as are on the ISS to power a single modest data centre.

Just waiting for Trump to invade Panema and bring about the final collapse of the US.

There was kinda that whole bit in between where all the unemployed workers crowded into cities and everyone died of cholera though…

Do I get to include the gargantuan cooling system for the data centre and all the infastructure required to keep that going?

So my GPU is about 300 watts and a still blatantly stupid LLM can write a little faster than me. Take off 100w to bring that down to my own writing speed then make it 10x slower to turn that 200 watts into 20 watts. Even with that heavy bias in the LLMs favour (forgiving it the entire power cost of my PCs other components that it partially utilizes) what we get is something slow, dumb, and incapable of learning because any local model is statically weighted.

To be fair, run an LLM on a machine with an equivelent power requirement to the human brain and we might se some different results on that one.

I still have my suspicions about how that 49% is being measured, because if it counts any ‘use’ then technically using most search engines is using an LLM regardless of wether you want to or not, a lot of people are now also forced by their employers regardless of wether they want to or not.

Only loosing 3x more than they make!? That’s a gigantic success by modern standards.

Even if that’s true its not odds I’d take on anything that costs money or that could be potentially dangerous.

Been on linux full time for under a week, I’d say linux is more like.
“I have issue”
“Here’s 10 extremely techincal solutions that will take you hours to work through and may not work”
“I still have issue”
“Oh yeh there’s just a button you can press that fixes it instantly”

On the bright side thats more my own lack of familiarity than anything and once an issue is fixed it stays fixed and generally works a lot more cleanly than if I had a similar issue on windows where fixes are always kind of cludgey.

I keep hearing the AI will get cheaper to run because its only training that’s expensive, but at the same time tokens cost more than employees and are still being sold at a 10 fold loss, and we apparently still need infinitely more compute to run models that can just barely do a simple job so long as you don’t mind spending even more time eventually fixing their work than they supposedly saved.

I find that hard to believe, I recently had to uninstall co-pilot after it weaseled its way into my search bar. Its not an exageration to say that my PC literally ran cyberpunk 2077 with pathtracting better than it ran the fucking windows search bar with co-pilot.

But everyone loves AI, look how its use has doubled! ᴬᶠᵗᵉʳ ʷᵉ ᶠᵒʳᶜᵉᵈ ⁱᵗ ⁱⁿᵗᵒ ᵉᵛᵉʳʸᵗʰⁱⁿᵍ ᵃⁿᵈ ᵐᵃᵈᵉ ⁱᵗ ᵒᵖᵗ ᵒᵘᵗ ᵒⁿˡʸ

You can but you can:

• A, Use an excessive amount of computing power to run an LLM agent to do it
• B, Spend 5 minutes on youtube
• C, The dev could just program in the option to skip