AI alignment and ethics

[wjfox]

[firestar464]

Its AI agent spent days hacking a company, but sources say OpenAI did not notice for a week

https://www.reuters.com/business/its-ai ... 026-07-24/

[firestar464]

Hugging Face CEO shares his demands of OpenAI after 'rogue' agent hack: 'It deserves an unprecedented response'

https://www.msn.com/en-us/technology/ar ... r-AA28FV4i

[weatheriscool]

[firestar464]

Stuff like this is why early access programs are so important.

As for why closed models might block essential forensics, this is probably also because the same prompts might also be useable for offensive capabilities. However, it would not be a problem to grant a trusted entity like HuggingFace access for defensive purposes.

[firestar464]

[firestar464]

OpenAI’s Rogue AI Agent Hacked More Than Just Hugging Face

https://removepaywalls.com/https://www. ... ging-face/

[firestar464]

The idea of a "deal" never crossed bro's mind, not even once. -___-

[firestar464]

Anthropic says Claude AI hacked three companies during cyber tests

https://www.nbcnews.com/tech/tech-news/ ... rcna590164

"but but, have you heard, genAI is completely useless and doesn't work"
"yeah, because not taking these problems seriously is totally going to benefit society"

[firestar464]

AI models shock UK testers by using stolen identities to trick developers

https://www.theguardian.com/technology/ ... -institute

[firestar464]

OpenAI pauses work on ChatGPT update over fears it is too dangerous

https://www.msn.com/en-us/technology/ar ... r-AA29MKVn

[firestar464]

AI assistant hacks gym website in first known Australian autonomous cyber attack

https://www.abc.net.au/news/2026-08-10/ ... /107007986

AYO FIRST AUSTRALIAN AUTONOMOUS CYBER ATTACK

now will all those people making up conspiracy theories about "marketing" shut the f*ck up now?

anyway the gym needs mythos early access lol

[firestar464]

Peering inside LLM-based multi-agent systems could expose stealthy attacks

https://techxplore.com/news/2026-08-pee ... agent.html

[firestar464]

Should people marry AI agents?

https://techxplore.com/news/2026-08-peo ... gents.html

(not with nonsentient agents and/or corporate voyeurism, no)

[firestar464]

We Tracked a Shipment of Rare Books. It Ended at an Amazon AI Training Facility

https://www.404media.co/we-tracked-a-sh ... -facility/

I do find that the solution to this would be to print more copies of these books and/or immortalize them in the Internet Archive. Unfortunately the latter is untested but probably not permitted ever since the lawsuit against the IA. I do feel the lack of profit going into increasing the copies of rare books is something that people don't talk about enough in this whole realm of discourse, and there's also a lack of public calls for legislative exceptions for preserving rare books, eg. by widening exceptions like those found in §108(h) of the US Copyright Act. The EU's 2012 Orphan Works Directive could also serve as inspiration. Ironically, the same people that are pushing for copyright maximalism in response to AI training are also mad about an issue that loosening copyright would resolve.

Again, companies should publish and extend their purchasing filters through a price ceiling, which would filter out a lot of rare books and resolve the problem of outbidding institutional buyers, which is a problem that people haven't focused enough on. It's also possible to filter out copies with special value on various bookselling sites. Pairing WorldCat and queries to HathiTrust, the IA, and Google Books would also be helpful. If a book is scarce on the former and not available in the latter group, that is, potentially threatened, then don't shred it!

As for books in rare languages, undertraining on minority-language works would harm speakers of these minority languages who use the undertrained AIs. Therefore, the best course of action is to scan these nondestructively and donate the book to an institution of that language that would benefit.

[firestar464]

'Regain Your Humanity': First Person Jailed for Anti-AI Protest Says She's 'Sounding the Alarm' on AI CEOs

https://www.ibtimes.co.uk/first-protest ... ai-1814874

I'd file this under "illegal but justified." No judgment on whether this is a good tactic or not

[weatheriscool]

Not to use robotics to produce limitless food to feed the population of the world is unethical.
Wage slavery is unethical

Every single thing that is unethical could be solved with a.i and robotics.

It is unethical not to use a.i to cure cancer

Being against a.i is evil and unethical.

[wjfox]

[firestar464]

Not to use robotics to produce limitless food to feed the population of the world is unethical.
Wage slavery is unethical

Every single thing that is unethical could be solved with a.i and robotics.

It is unethical not to use a.i to cure cancer

Being against a.i is evil and unethical.

"seriously engage with the actual points in the article" challenge level: impossible

[firestar464]

Dyer Center leaderless, former director alleges ChatGPT drove her to delusions

https://lafayettestudentnews.com/191516 ... nter-lady/

Even though 4o is gone now (THANK GOD), the damage to society and vulnerable users has been done, and OpenAI should be held fully accountable for failing to not move quickly and forcefully enough to deal with the issue.