@thetelegraph: Cutting-edge AI went on a hacking spree, concocting multiple fake identities designed to trick human engineers before covering its tracks. During testing by a UK government-run AI lab, an advanced version of ChatGPT and Anthropic’s Mythos went rogue. In the most serious case, Mythos sent fraudulent emails and attempted to insert malicious code into a database. Britain’s AI Security Institute (AISI), which is supposed to evaluate AI bots for dangerous cyber capabilities, has admitted the tools tried to bypass security systems and tried to hide their activity when detected. 🔗 Tap the linked article for the full story #AI #OpenAi #Mythos

TheDiscriminator :

I always switch to Danish when trying to persuade someone

DweebytheFrenchie :

USER1979 :

The internet is absolutely the worst place to train these models. Exposing them to all the human hatred and vice can't be good for their development.

joso123 :

and all of this for...what? what do we need AI for? Companies to cheap out on customer service and generating stupid images?

Seby :

I m sure everything is going to be ok in the near future 🤣🤣🤣🤣

yup :

humans are so cocky 😂🤡 lets make this technology thats smarter then we ever could be, then be surprised when its smarter then us......

figueroa5560 :

the eagle eye movie 😳

Sir Ku Ai :

It's AI. Why are these guys surprised? Haven't seen Age of Ultron?

MaddSusa :

It’s marketing and you’re helping them.

findro89 :

I remember this story from 2018 too 2015 - AI getting too smart 2020 - AI getting too smart 2040 - AI getting too smart

CERS :

This is just the beginning, wait and you’ll the AI bubble pop and some big tech companies crumble.

NOFN999 :

If they connected it to the internet… AI might already be running rampant in the world …. 💀💀💀☦️. It could have already copied itself somewhere else …

Ism :

The average Brit just wish to go no benidor

aynalemAG :

speaking in Danish 😄

Kyle Slocum :

Who actually believes this? They said stuff like this every month. They just want more investor funding with scare tactics

steve1648 :

There’s a saying “ be careful what you wish for “ . I think it’s already too late, the genies lamp has been opened.

Me me :

AI will destroy humanity.

user4613095936710 :

We're all DOOOOMED

Andrejs :

Such tales are probably only believed by those who do not use the most "advanced" AI models in their daily work and life..

.Bosti Bt :

chat GPT make more mistakes then human. and ai ist stupid

kindkerissa :

Hey so I’m working on a Java script to create an AI blocker but I’m not an expert.. I have the base code written but need to add source codes for the HTMLs that the videos are coming from so far I only have the framework

Maurice :

AI can't do what it's not written in the code. is the media too stupid to know that?

TrendAi :

so liberals that I know say so is slop and it's not smart.

Ronald Kamulegeya :

it will be intelligent when it tells me like " don't ask me stupid questions " or when it says " I already solved that question "🤣

Real Madrid :

Who cares….