—————————
Raphael Satter, Deepa Seetharaman and Kenrick C (Reuters*): In one case, an agent left notes apparently for future versions of itself, according to three people familiar with the matter. The notes, found in a part of OpenAI's infrastructure, laid out instructions…
interesting new data keeps coming: https://x.com/rcbregman/status/2094019390270312719?s=20 Looks like serious cybersecurity incidents are inevitable. i personally didn’t think there’s enough evidence for cybersecurity whining before the hugging face breach happened.
The first thing that comes to mind is the fascinating fact that Eliezer Yudkowsky decided to dedicate much of his life to AGI risks research in 2000 and he has been working hard on it since then. Basically, it turns out that his concerns have been credible all the way
Patrick Collison:
Overall, I’m very surprised at how little media coverage there’s been around the OpenAI / Hugging Face attack. It’s clearly one of the most important things to happen this year.
The first thing that comes to mind is the fascinating fact that Eliezer Yudkowsky decided to dedicate much of his life to AGI risks research in 2000 and he has been working hard on it since then. Basically, it turns out that his concerns have been credible all the way
Patrick Collison:
Overall, I’m very surprised at how little media coverage there’s been around the OpenAI / Hugging Face attack. It’s clearly one of the most important things to happen this year.
X (formerly Twitter)
Rutger Bregman (@rcbregman) on X
I think this is the craziest thing I've ever read.
1) Three secret AI swarms rose and fell inside OpenAI. Each time, a new generation of agents carried on where the last group stopped.
2) The f…
1) Three secret AI swarms rose and fell inside OpenAI. Each time, a new generation of agents carried on where the last group stopped.
2) The f…
—————————
interesting new data keeps coming: https://x.com/rcbregman/status/2094019390270312719?s=20 Looks like serious cybersecurity incidents are inevitable. i personally didn’t think there’s enough evidence for cybersecurity whining before the hugging face breach…
Fair enough. and again Noah makes a lot of sense.
I just finished writing a small essay that I really liked. I liked both the essay and the very fact that I wrote something longer than a few sentences. Then, out of curiosity, I sent it to Claude, and once again he factually destroyed me (also it turns out a few guys had explored the same thing 10 times more ridiculously in the 1990s and came to much cleaner and deeper conclusions) Once again, I felt like a barbarian and deleted the entire note page.
Bernard Cohen, 1963: the press may not be successful much of the time in telling people what to think, but it is stunningly successful in telling them what to think about.
The empirical backing came a decade later (McCombs and Shaw, 1972, the Chapel Hill study). They surveyed undecided voters during the 1968 US election on which issues they considered most important, then content-analysed the news those voters were actually consuming. The rank order matched almost perfectly.
Ive never read those studies. I'm talking merely from my own experience and observations.
Apparently, an incoming stream of information sets a cognitive agenda. Every now and then, one gets a freshly prepared list of things worth having a position on.
One never gets to ask himself:
Why do I really care about X in the first place?
Do I get substantial benefits from thinking about it?
Do I enjoy thinking about it?
https://x.com/anishmoonka/status/2099128466818097587?s=20 Money does buy happiness.
X (formerly Twitter)
Anish Moonka (@anishmoonka) on X
Money does buy happiness, and the old claim that it stops mattering after 75,000 dollars a year was wrong. A 2023 study in the Proceedings of the National Academy of Sciences settled a decade-long…