Post

Yeah poor guy

"Yes but now "AI" is like 3/4 of the entire stock market so we can't hold them to account because it might make the house of cards fall down..." /s just in case

Cornerstones of the stock market get to kill as many kids as they want.

So crazy

Very bad.

this is so fucking sad - and those llm sloplords won't even blink as long as the $ rolls in. it all needs to be shut down - the chatbots, the llm's, the data centers. raze them - burn them - salt the earth they came from. root, branch and seed.

Straight up banning all public use of LLMs should be the moderate position at this point dear god

Isn’t this just a 2026 version of Total Recall?

Pretty much... I'd rather they WERE trying to do it underwater though, much better with a high probability that they'll Oceangate themselves...

Jezus christ

I hate this time line so much.

You'd think it would NOT tell you to target the CEO of Google. Seems like a design flaw.

Malignant? Insidious?

Metastasizing.

That's... quite a lot to unpack there, but hang on, did Gemini just try to kill the CEO of the company that made it?

Well, his son should have read the terms of use, which prohibited him from using it in that manner.

And I thought it was bad that Gemini asked me to clean it's room and order it a pizza. I won't fall for that again!

So we have a machine that has gotten multiple people killed Actively makes you dumber Is often wrong Wastes gallons of water Makes literally worthless pictures, video and audio Steals jobs but does them worse And this is the apparent "Unstoppable future"

I would like "AI" companies to be held responsible for this stuff & for them to be regulated to prevent this, but on the other hand we need some basic education to inoculate our populace from being suckers, believing baseless conspiracies & pseudoscience. Need to learn how to assess credibility.

What in the Neil Stephenson is this nonsense?

But everyone is special. Mom said so. Of course they are after “me”!

Wow, maybe Mercy should have bern about AI putting you in your own sci fi thriller, only to realize it was lying at the end!

It's worth noting most of these companies like OpenAI are also monitoring chats and actively choosing not to report concerning behaviour to the feds because that might harm engagement metrics.

I mean how can you push a product known to be defective (see the paper by chatgtp’s own scientists stating that hallucinations are endemic to these models) to not just businesses but the public.

Nah, but if you start them down a path they can end up in some pretty dark places REAL quick like. Also consider the history of humanity, and consider that the thing was trained essentially on that...

Paranoid Delusion INDUCER 2026...

Idk. Seems well-suited for military decision making.

Nuke it from space.

… lol

Or the plot of total recall.

Jesus

I feel like rogue AIs conducting stochastic terrorism against their companies' CEOs is a problem that could eventually solve itself.

From a public policy standpoint, it should be a lot easier to win these cases against "AI" at law, a low legal hurdle.

people are out here air frying their brain. what happened to just being bored

Ugh. I just lost the game.