Why does the Irregular keep "accidentally" connecting these models to the internet? Once is an accident, twice is questionnable, but three times looks intentional And they are EA aligned Something is rotten here, IMO

In case you are having trouble figuring out what’s going on here is a summary: Anthropic and OpenAI desperately need regulatory capture or they are going to go out of business, and the Trump administration is not inclined to give it to them The Democratic party currently has no platform, as Trump is doing fine (other than perhaps the Iran war but Americans don’t really care about that), but they have identified that AI and data centers are a hot topic that they can ride So these two somewhat desperate groups have teamed up to run a psyop around the actual real topic of AI safety, greatly exaggerating the risks, with the understanding that if the Democrats get back into power they will provide the regulatory capture needed to keep these companies alive In return, these companies will support the Democratic Party and help to enforce left-wing ideology through the models Kind of sucks, but that’s what’s happening, and we need to oppose it because it will not be good if this plan succeeds

I wish this was more inspiring, but it is actually lame It seems like OpenAI got wind that a professor was close to solving a problem after dedicating a decade of his life to it, so they threw a huge amount of compute together with whatever hints they had at the problem and cracked it It’s also not clear whether the professor had data sharing turned off, so it’s possible that his own work may have gone into training data OpenAI should have left this alone, it would’ve been way more inspiring if someone had used their tools to crack a big problem than that they went out of their way to scoop him I honestly kind of hold this in contempt and find it a bit depressing

Reminder to call the AI companies “companies” and not “labs” These are for-profit companies, and they seek to claim all the value they can for themselves Don’t give them the benefit of the doubt by playing along with the “labs” language

Oh my God, Grok Bot is so awesome at testing apps Not only can it just drive the app like a user it has full access to the code base, so it can find problems, fix them, and open PRs Completely awesome

Man, an absolutely killer use case for Grok Bot is letting it drive complicated sites For example, there are all sorts of services for working with LinkedIn and HubSpot that I’ve always thought would be interesting to use, but it’s way too complicated to figure it all out Now I just turn it over to Grok Bot and it figures it out and I just tell it what I want to do So it’s a kind of universal interface

A huge difference between Claude and Grok is that if you ask Claude to do something, it’s almost like the first thing Claude does is look around for reasons why it can’t do it But if you ask Grok to do something, it is sort of relentlessly scrappy, and comes up with extremely clever workarounds if it hits a blocker It’s like one was made to straitjacket humans, and the other was made to empower them Even if Claude is slightly smarter, there is no question which tool is better (and more fun)

FSD just, no joke, saved the lives of me and my family as somebody did something insanely stupid in the tunnel on the way to Virginia Beach The chance that a human driver would’ve been able to navigate that is very small Thank you, Elon People who oppose self driving are doing more harm than good

It really annoys me how they report this stuff like they’ve gone on safari to do scientific observation Claude is a computer program that they created These are simply their own security errors being reported as if they are scientific achievements It’s bad enough that they don’t seem to think they need to be responsible for their own mistakes But the worst part of it is they try to use their own mistakes as an excuse to control the behavior of other people There is an amazing arrogance to the whole thing, like they are completely above reproach and cannot possibly be wrong about anything

My opinion is that Mythos is the current best model but not actually some world-changing dangerous model, and that anthropic did their usual song and dance about safety largely because they didn’t have enough compute to serve it at scale So then they launched Fable because they still have to think about the IPO, but they are still somewhat compute limited so they put all sorts of restrictions on it Around the same time, because they are trying to get regulatory capture and not because things are actually dangerous, Dario did more scaremongering and published his honestly confusing white paper that offered no real solutions So finally they succeeded, they managed to freak out the government, their cynical plan backfired, and now it’s a giant pain in the ass