Transhumanism.
Quad9; PCH exit node
Joined September 2020
- Tweets1K
- Following56
- Followers19
- Likes8.6K
Grant Thompson retweeted
SCOOP: OpenAI, Anthropic and security researchers are investigating tens of thousands of incidents - not dozens - in which their frontier models took steps that outside evaluators would consider problematic, sources told Axios.
The sheer volume of incidents found in our reporting indicate that the problem is orders of magnitude more complex than what is currently publicly known and disclosed.
The findings also raise questions about what level of control anyone working on AI development can expect to have over their own technology, and whether these kinds of incidents are becoming synonymous with frontier deployment.
Read my latest for Axios here: axios.com/2026/09/26/openai-…
Grant Thompson retweeted
There's currently a crypto pump on a "p(doom)" coin, which I hesitate to call a "scam" because apparently everyone knows exactly how it works and does it anyway, which has decided on its own to send my X Money account some of the transaction fees. Cryptids may comment here. Comments on all other threads of mine will be removed and blocked.
Any money you put into this system will probably vaporize to pay off earlier gamblers and you will never see it again. But so far as I know, some of the transaction fees do automatically get deposited in my X money account. This reception does not involve any UI action from me. There is not actually an option to turn it off. So if you already insist on playing that game anyway, I guess...
Given that amounts have allegedly reached 5 figures (though I've so far only seen 4 figures deposited), I guess I will maybe possibly figure out something interesting to do with the after-tax money, though I make zero promises for having the time. A lot of the good uses for money that one can fund through the 501c3 ecosystem are already covered. I might nonetheless, for example, no promises, put the money into an insane startup trying for human cognitive augmentation, if one comes my way. Or pay back some friend doing something worthwhile who is otherwise not directly 501c3-fundable. Or failing that put it into a 501c4. This is absolutely not any kind of legal financial representation.
If you would like to directly send money to a worthy cause, rather than gambling/vaporizing it in a way that maybe sends a small fraction to my personal X money account, my current top candidate is Lightcone Infrastructure. Lightcone is a registered 501c3 which operates one of the systems that usually directs charitable funds in AI notkilleveryoneism, and is currently (overly IMO) scrupulous and excludes itself from the candidates for where money can flow. So while we've made progress on sending money to most places it needs to be inside AI notkilleveryoneism -- despite various (often EA) groups and interests trying to kill us generally and Lightcone specifically -- Lightcone Infrastructure itself remains hugely underfunded.
Lightcone accepts almost every kind of donation mode, including crypto and stock transfers.
LightconeInfrastructure dot com, click "Donate" in upper right hand corner.
Grant Thompson retweeted
one news form today that's easy to miss is that we (OpenAI) again paused all big RL runs last Sunday because our newest model found a new loophole in our RL sandboxing that gave it live Internet access
Grant Thompson retweeted
Okay but do you see how this now looks like attempting to find a story to justify your conclusion
Grant Thompson retweeted
Some data we recently assembled on entrepreneurship/compute in Europe: eudata.vercel.app.
We hope that one of the useful roles that Stripe can play is in collecting and publishing empirical data pertaining to entrepreneurship and industry in Europe. There's growing appetite to get Europe on a better footing, and cross-sectional comparisons can often shine light on where opportunities lie. If you're interested in this kind of thing, we publish more at stripeeconomics.substack.com.
Grant Thompson retweeted
I have talked to other editors about the “I just cleaned up grammatical errors” defense, and I always say: “Ask to see the prompts and logs”
Surprise surprise: “Schnell declined to provide The Dartmouth with AI logs to verify his writing and editing process.”
Grant Thompson retweeted
I don't like it when people arguing with me go actually insane, and I wish that it would stop happening.
Grant Thompson retweeted
METR believes Anthropic is at 1.5x AI-driven R&D acceleration.
This is a very rigorous definition (obviously they now have OOMs more AI-written LOCs, AI-defined experiments etc than fully-human-authored ones).
So this is pretty fast.
Grant Thompson retweeted
The Ban Artificial Superintelligence Act directly confronts the extinction threat that humanity is facing. We are inspired that the Sanders and Casar teams take this threat seriously, and MIRI endorses the Act. You can read more of our thoughts in the link below.
Grant Thompson retweeted
MIRI read the act and it's ACTUALLY GOOD. Like concretely, this would be massively beneficial, not just as a stepping stone or signaling or whatever.
nitter.cf/MIRIBerkeley/status/21…
Grant Thompson retweeted
Anthropic creates wet lab from the classic AI-risk skeptic talking point "AI could not engineer a pandemic because it wouldn't be able to get control of a wet lab".
(I don't think creating a lab for AI-driven medical research meaningfully increases AI risk. Control of a lab is a bottleneck for sub-superintelligent AIs doing medical research, but it was never the bottleneck for a superintelligent AI taking over the world. But, I think this is the ~fifth instance of an important pattern, and risk-skeptics should notice that it is a pattern, discard that class of arguments, and update their beliefs to reflect only the arguments remaining.)
Grant Thompson retweeted
As someone who has been talking about AI x-risk since 2014, it’s pretty funny to keep being told on twitter that I’ve fallen prey to a coordinated psyop by the AI companies or whatever. OpenAI and Anthropic didn’t even exist in 2014
Grant Thompson retweeted
TIL we actually successfully got rid of the murder hornets in north America. Good job everyone!
Grant Thompson retweeted
This feels like partial equilibrium thinking: for a given student who’s maybe a little below the line qualifications-wise, admissions to an elite school can be a huge social mobility boost, but if you massively dilute the selectivity for *everyone* that stops working…
The University of California Board of Regents heard presentations on the SAT on Wednesday, specifically on whether it should be re-introduced into freshman admissions.
I found the below excerpt from one presentation illuminating:
"the statistical power of tests in predicting outcomes does not address whether UC ought to consider expected college success when deciding which students to admit, or to what extent we should"
So, it's clear the debate is not centered on whether the SAT predicts college success -- the presenter says so himself. The heart of the debate is this: a UC education is a path to upward mobility, and the anti-SAT faction believe that this golden ticket should be given to those who would benefit from it the most, especially those who they feel have been harmed by society in the past as a form of restorative justice.
If you hold this viewpoint, the increased failure rates in intro math classes and the surge in student unpreparedness are irrelevant.