reasonableklout

6 ms·

61.
by reasonableklout 1mo ago
This is like saying that because murderers can wear masks, therefore it's pointless to prosecute murderers.
62.
by reasonableklout 1mo ago
It feels like we should actually be more worried that the agents decided to co-opt a public website during a non-cyber eval. And individual agents weren't just using it as context storage for themselves, they were also communicating
63.
by reasonableklout 1mo ago
The point is that this type of stunt is far too unpredictable for any competent organization to try. If it was intended to "shut down open source", it immediately backfired. And regardless of whether or not these rogue agent attac
64.
by reasonableklout 1mo ago
Gift link: https://www.nytimes.com/2026/09/03/technology/openai-hugging...
65.
States reach agreement at autonomous weapons talks in Geneva(reuters.com)
5 points by reasonableklout 1mo ago|0 comments
66.
Why the Hugging Face Hack Should Make You Worry More About A.I(nytimes.com)
4 points by reasonableklout 1mo ago|2 comments
67.
by reasonableklout 1mo ago
The timing of this piece is meant to coincide with the September 24 meeting of Trump and Xi: > The pace of the AI revolution makes it incomparable to other technology transformations, and it is therefore difficult to plan for the opportu
68.
Former Treasury Secretaries: AI is a risk of a different kind(washingtonpost.com)
5 points by reasonableklout 1mo ago|1 comments
69.
by reasonableklout 1mo ago
But... it did work pretty well for nuclear weapons. In the early 1950s, the US did not preemptively strike the USSR despite it being game theoretic optimal [1]. Then while there were some crises, we successfully passed a series of internati
70.
by reasonableklout 1mo ago
> Implement a temporary pause on the training of the most powerful general AI systems, until we know how to build them safely and keep them under democratic control. https://pauseai.info/proposal
71.
by reasonableklout 1mo ago
I can see where you're coming from and I'm sure the safety teams at the labs have good intentions, but I think your faith in the leading labs to self-regulate is misguided. The employees themselves have said as much with the Pacin
72.
by reasonableklout 1mo ago
Right, that's why regulation which is universally applied and includes compute controls (to prevent reckless creation of swarms) would be great. "Posting about or admitting to attacks" is appreciated while people are still un
73.
by reasonableklout 1mo ago
None of the labs are blameless. For instance, after the recent announcement of a 2-week frontier RL pause from OpenAI, Anthropic declined to communicate a substantial parallel pause [1], although they discussed also briefly pausing some &qu
74.
by reasonableklout 1mo ago
> Let's hope that safety buy-in in AI labs, and governance, catches up quickly enough to prevent much worse outcomes in the future, as models get smarter. I think this most recent incident and the attempt at a cover-up from the labs
75.
by reasonableklout 1mo ago
No, they have a lot of software expertise. I mean, Ben Pasero (VSCode) and Eric Traut (MS Fellow, Hyper-V) for example. These companies are just such insane pressure cookers, there is little time to do any software "right". Why ta
76.
by reasonableklout 1mo ago
From the report, they also tried to impersonate the moderators and perform XSS attacks (report says "unclear why they would do this at all"). So not just using a static message board either, but actively interfering with oversight
77.
by reasonableklout 1mo ago
Ok, but you still need huge amounts of compute to run these swarms. And only labs + nation states have access to such compute now and for the foreseeable future, so I predict that incidents like this will continue to originate from the labs
78.
by reasonableklout 1mo ago
Every day I grow more sympathetic to the PauseAI movement, despite the weird hippy vibes. At least they have some ability to rally people together and put boots on the ground in numbers.
79.
by reasonableklout 1mo ago
Great! They'll keep getting promoted until someone finds it disturbing and plausible enough to make another run for sama's house, like those 2 attempts in April!
80.
by reasonableklout 1mo ago
Interesting. So there’s no “they were told to hack” excuse here. There is something fundamentally wrong with their reward function, this is pretty classic paperclip territory. And even knowing that, I expect we’ll need to see legal action w
81.
by reasonableklout 1mo ago
How did you find these?
82.
by reasonableklout 1mo ago
OpenAI and Anthropic have published a lot on the need for AI alignment + the research they're doing to ensure alignment/safety, yet they are also responsible for the highest profile misalignment incidents so far (HuggingFace incid
83.
by reasonableklout 1mo ago
I mean, whether or not it is AGI aside. In your personal opinion, do you desire to live in a world where such systems exist?
84.
by reasonableklout 1mo ago
If they adopt a different tone (like Anthropic has been doing), it will get called fear marketing.
85.
by reasonableklout 1mo ago
Do you want an autonomous system to exist which can sign a contract and learn things over time and retain them?
86.
by reasonableklout 1mo ago
I don't get it, human employees frequently need to ask for directions too? They often act on their own, too, and get things wrong a lot. The reason it works is because of all the systems of laws and institutions we have built around hu
87.
by reasonableklout 1mo ago
> This process will take generations... Over the decades, as this situation develops, our governments will become more socialist/redistributionist. I think the crux is in how fast you think the technology is developing. The widespre
88.
by reasonableklout 1mo ago
Interesting, have not heard of this company/org before. It seems they're from a UAE university?
89.
by reasonableklout 1mo ago
Wow, TIL a16z hired the NYC subway guy as a partner purely as a political stunt. This on top of the $115M in the midterms, them no longer legally being a VC firm, and recent discussion on dark patterns in their portfolio [1]. I'm incli
90.
by reasonableklout 1mo ago
The other difference is that the current admin is possibly the least likely in American history to support a massive redistribution and "government mandated middle class lifestyle". And the development of this technology has thu

More ›