Is this…

A lemmy post sharing a bluesky post that has a screencap of an Instagram post thats a share of a video?

Are we serious?

… are you daft?

I read the article, this link: https://today.rtl.lu/news/science-and-environment/u-s-researchers-use-ai-to-generate-virus-for-the-first-time-500575347

Which is what the post shares.

What article did you read?

And notice how the article shows no sources, no citing, its basically just clickbait. Garbage.

People need to raise their standards, garbage like this should be down voted into oblivion.

The fact this garbage is upvoted so much shows how the avg person on here is either

A: not reading the links at all and just updates cuz of the clickbait titles

Or

B: reads the article and has zero sense of upholding journalistic integrity.

Yall are really pushing for an echo chamber.

Do. Better.

The article doesnt link to the original published paper.

That makes it utterly valueless and trash.

Stop posting garbage articles. Did you even verify or try to find if this was even real…?

Where do you think the AI piss filter came from…?

Its trained on a bunch of existing artists work…

Nothing on these cans indicates AI to me.

obvious slop

So your source is “I made it the fuck up”?

Stop spreading disinformation.

Any proof at all this is AI art or are we just calling any stylized art “AI” now baselessly?

I squinted at these cans for a minute and a bit, saw nothing that looked like AI artifacts.

Post your actual proof, otherwise, stop spreading disinformation.

Literally post any source, or any proof this is AI art.

Im squinting at it and not seeing anything atm that indicates its AI.

So it sounds like you are making shit up.

Nah, not surprised at all, I work with developers who run stuff in yolo mode raw dogging copilot directly on their work laptops every day.

Madness.

I keep that stuff boxed up inside of a docker container, sandbox’d, so possible vectors of damage are kept to a minimum.

The tl;dr:

1. Prompt inject a malicious instruction in a word document that instructs the AI to copy this instruction to other documents as part of the payload.

2. Dumb user downloads and opens the document with copilot enabled, abd ignores the large suspicious white blank page that totally doesnt look like a hidden giant injection attack.

3. Thats it pretty much it.

Copilot will get injection attacked because the prompt is super huge and at the end of the document, so its prior instructions start to fuzzy out.

Then it’ll go “okey doke” and start copying the prompt injection attack payload to a bunch of other documents.

The fix is stupid simple… copilot should just be prompting the user for permission if it ever edits a file other than the one that is open. Im surprised that isnt already the case…?

It certainly is already the case for copilot in vscode.