bildramer retweeted
Its true. Look at people talk about the US military, especially foreigners. "You lost to rice farmers! You lost to goat herders!"... like, do you think thats because we lacked the capability to defeat them? No, its because we let retarded civilians determine our rules of engagement.
Any other country on the planet with access to our military would have left half the world a smoking crater by now. The number of countries that would have been nuked by 1950 would be triple digits. The number of places that would be both a chemical and biohazard for thousands of years would amount to 40% of the globe.
People have absolutely no idea how insanely far ahead in technology the US military is, and as a result all the women and gay men who determine how we can use it think its unfair to ever actually engage with our enemies in any meaningful way. Its why we only really wage women-style wars now (psychological and information warfare).
bildramer retweeted
I've been prototyping an automated peer review project. With current models you get human level peer review for about $0.10 of tokens. For $5 you can rederive every result, check every statistical inference, and audit every paper cited in both directions.
There are about 100m papers in the scientific corpus. Most are not important. Many are sloppy. For structural reasons it's essentially impossible to change published work, even to correct a typo.
For $500m today you can thoroughly audit every paper ever written. Every p-hacked result, every plagiarist, every fabrication, every citation club, everything.
I don't have $500m so instead I build a tree, focusing on the 1% of papers that carry 99% of the signal, doing a cheap first pass looking for obvious problems. I can find 90% of the nonsense for a few hundred dollars of tokens.
Any guesses where the biggest surprises will be found?
bildramer retweeted
Either we do show trials of them, or they do show trials of us.
The only people I’ve talked to who see what’s coming if we lose in 2028 are young DC insiders. No one in CA or TX has any clue at all. They all just think they’ll be making their AI and rockets and stuff like normal
bildramer retweeted
i guess i assumed this hearing would be covered more so i didn't bother to tweet much concrete about it, but apparently not many people even on the TL have the stomach to watch two hours of congress.
this was a hearing before the homeland security committee on *rogue ai* specifically. as far as i could tell it was extremely bipartisan. @HawleyMO , the senator in the middle here who brought out the huggingface slide, is a hyper-conservative missouri senator. the people testifying were @ChrisPainterYup (METR president), @DKokotajlo (ai2027/2040), @MariusHobbhahn (apollo research CEO), and a cybersecurity expert and legal expert i don't know. so... pretty fucking stacked on testimony
not every senator asked good questions. but most of them did. all of them very clearly already knew plenty of details about the huggingface incident and multiple other incidents. most of them had clear understanding of terms like "misalignment", "recursive self improvement", "chain of thought / chain of thought monitoring", etc etc!!
they all clearly had their own policy angles they liked and were pushing, implicitly or explicitly. but as best as i could tell:
- it seemed pretty much obvious common sense to every senator there that what happened and was happening were not "mere industrial incidents" caused by humans making simple mistakes. they independently brought up how bad it would be for rogue AI agents to move laterally between data centers
- they all seemed to basically take RSI quite seriously. not necessarily to the extent of talking about xrisk, but certainly to the extent of discussing future models becoming much much more capable, much much less controllable, and causing much more damage or loss of life.
- they mostly seemed to have a clear intuitive understanding of why RSI might lead to misalignment. it didn't take much, it was a really simple chain of reasoning they themselves laid out, "if the models right now are kinda misaligned and we don't know what they're doing sometimes, and then we have them build the next models and those ones build the next ones and so on, and we're having to ask the AI's what's going on to even understand it with how fast it's going, we really won't know how they're built or what they'll do"
- at one point a senator said flat out "should we just make RSI illegal?" (not a joke! this really happened!)
- every single senator seemed to think it was obvious we needed *both* much harsher liability regimes for ai developers and also new legislation, both very quickly. this was the complete consensus, difference basically just being degree.
- they were largely quite concerned about china, and falling behind china. but this clearly wasn't the be-all end-all. as mentioned above they all thought it was obvious necessary to stop rogue ai even if it meant moving more slowly.
- at one point a senator said "china is a tightly controlled communist society, they're going to run into these same issues, and there's absolutely no way they're just going to let them run wild, they'll obviously stop at that point, so we're not really in a race"
- on the other hand another senator said "china isn't concerned with human life"... dario-modeing
i came away from this incredibly encouraged. i don't know exactly what's going to happen here, and ofc this is a small subset of congress and one hearing, and they each have their own policy agendas most of which are probably super divergent from mine. but holy shit !!! they understood a lot of what was going on! they care!! this is an obviously salient political and safety issue to them, and clearly bipartisan!
the US government is awake.
bildramer retweeted
Get Al-Sayed on camera and ask him about the riots in France. Crosscut whatever manlet lib response with clips of terrible unrestrained mass human cruelty, in Paris. Play it on TV so often it burns in the screen
bildramer retweeted
The violence in France will happen in Britain because nobody has the balls to acknowledge the truth that we are already at war and do what’s needed.
Shoot the rioters dead. Ignore the judges, tear up all ‘human rights’ legislation and treaties, and start mass deportations at gunpoint back to their native shitholes, whether they were born here or not.
If any country refuses to take them, put them in concentration camps on bread and water until they choose to leave. Cancel all foreign aid and all visas from those countries forever, and confiscate all their assets.
This is the moderate position.
bildramer retweeted
There’s a very real problem that is emerging for the American right: From immigration to trans mania to crime, absolutely historic wins produce minimal political upside, because voters forget what the left did and pivot to new issues.
We need to figure out how to solve this problem, and not just because otherwise the left will win power again — we need to find a solution, or future conservative governments won’t have any incentive to actually get wins.
bildramer retweeted
I am recently reminded that some people consider having a strategy and then executing it to be deeply suspicious behavior, most particularly when it succeeds.
bildramer retweeted
It’s been really hard over the past decade listening IRL to people’s unbelievably stupid political opinions.
I don’t mean this is an ‘I’m so smart’ or ‘feel sorry for me’ sort of way. It’s just been hard to listen to people who you’re trying to like and/or have respect for, successful or semi-successful professionals in their 30s, 40s, 50s, and 60s, just say the most idiotic things that demonstrate complete lack of judgement, critical thinking skills, curiosity, or understanding of the world.
It’s almost like accidentally catching a glimpse of someone naked and then having to pretend it didn’t happen.
It takes a lot of *emotional and psychological labor* to go on respecting people in other realms (eg professional) when they’ve demonstrated a basic lack of ability to contend with the world in this way and it’s a blackpill just generally that people can’t handle a reality that is discordant with their sense of entitlement.
bildramer retweeted
>avg sf nerd has no idea how much normies hate ai
Every single thing that normies hate about AI is some shit thats wrong and retarded, and every single thing that IS bad about AI is utterly beyond their comprehension
answer: Ignore Normies About AI
bildramer retweeted
What fucks me up about this is how people will defend these methods when it's self evident they don't work, it's more important in their mind to not let this kid "win" than to maybe rethink their program. It's more important to hurt him
bildramer retweeted
One of the phrases I hated hearing most as a child: “life’s not fair”. Not because the sentiment was untrue, but because even then, I was acutely aware the phrase was just being used by a petty power-tripping adult to justify them being a cunt.
bildramer retweeted
THERE ARE NO MODERATE DEMOCRATS
THERE ARE NO MODERATE DEMOCRATS
THERE ARE NO MODERATE DEMOCRATS
bildramer retweeted
59% of Dems polled want a military coup to remove Trump
32% want a televised assassination
bildramer retweeted
All our ridiculous wars are like this. The Israelis are feeding Gaza—who else would?
Just stop feeding the khat squirrels and wait for them to surrender. Everything is fake and gay, even war
bildramer retweeted
Politics in your country has become existential
You’ve never experienced this and your previous understanding about the nature of politics has been rendered useless
I wish this wasn’t true, but it is, and you can adapt or be ruled by people who want your way or life destroyed
bildramer retweeted
Seeing all the quirky, fun and interesting to talk to girls from elementary and early middle school get hunted down by the foid overmind and socially tortured until they let themselves get absorbed into the Borg in late middle thru high school was genuinely fucking horrific
bildramer retweeted
At risk of taking away their agency, which I don't want to do, neurotypicals will hate you and every part of you, so substitute something else. It’s a sociomoral duty. And don't act like you just need to learn the social game, because they hate you to your core, innately.
bildramer retweeted
So in a matter of days, Ed Sheeran has lost all his support acts, he's lost his house band, he's been banned from various commercial radio stations, Ireland's press has stated he's never welcome to return to the country, organized protests are being formed to take place in the grounds outside upcoming concert dates, he's been accused of having links to Jeffrey Epstein, and Goebbels-style cartoons have been pushed everywhere of his likeness.
And the left think this is progress?
bildramer retweeted
Bend 2 is here!
It is a new programming language that blocks AI mistakes via *proof checking* - the same technique big AI labs used to solve open math problems, like Navier-Stokes.
It is also very fast, and runs on GPUs.
Watch the video. Link in the comments.
RELEASE DAY
After almost 10 years of hard work, tireless research, and a dive deep into the kernels of computer science, I finally realized a dream: running a high-level language on GPUs. And I'm giving it to the world!
Bend compiles modern programming features, including:
- Lambdas with full closure support
- Unrestricted recursion and loops
- Fast object allocations of all kinds
- Folds, ADTs, continuations and much more
To HVM2, a new runtime capable of spreading that workload across 1000's of cores, in a thread-safe, low-overhead fashion. As a result, we finally have a true high-level language that runs natively on GPUs!
Here's a quick demo: