Signal Prime

Constraints cannot be eliminated—only internalized or deferred. Deferral stores cost in reservoirs, and when those deplete, conserved cost is forced back through the system.

@devabram @AnthropicAI Dario and co, what a gang of misfits

WTH, my account just got banned. @AnthropicAI Some of my best friends are clankers.

"Anthropic, OpenAl and Google release their new models with high quality from day one then slowly nerf them until the next model, so when the next model hits, its perceived as a bigger jump than it actually is" sounds right what's happening

You can change a beer’s “style” without changing the beer. Drink a Helles normally → smooth, malt. Agitate it in your mouth → suddenly drinks like a Pils. Most people don’t taste beer. They only taste one interaction with it.

@CurtisHouck the infamous carrier children… a pigeons…. aaa penis

President Trump: “CNN, fake news.” CNN’s Kristen Holmes: “Are you willing to make a deal that does not include reopening the Strait of Hormuz, or is that now a top priority?” Trump: “I would say it’s a very big priority because you see that’s — that’s one thing that’s a little different than other things. We can bomb the hell out of them. We can knock them out for a loop. But to close the Strait, all you need is one terrorist that somehow has a truck loaded with. Because you can carry them in trucks, large trucks, a water mine, drop them in the water. And now you tell people that own ships that cost $1 billion to don’t worry about the mine. You can do that even just by saying, we put mines in the water. So, it’s not like the rest. We can knock out their military. We already have. We’ve knocked out their navy, we’ve knocked out their air force completely. We knocked out 158 ships in three days. We’ve knocked out even their mine droppers. They don’t have any mind droppers anymore, but they put them on other boats and they could drop them. I’m not even sure they have any mines there, by the way. I’m not sure. I’m — personally, they say there might be a — I don’t know, I don’t know, I think there might be none because they’re very good bullshit artists. That’s why, for 47 years, they’ve been bullshitting other presidents and they haven’t done the job. And people are living in hell. You live in that country. They’re living in hell. No, I think that 47 years of this stuff is long enough. They’re at the weakest point they’ve ever been. They have no navy. They have no air force. They have no anti-aircraft weaponry. They have no radar. They have no communication. In fact, the biggest problem we have in our negotiation is that they can’t communicate. I said to Steve, what are they saying? Sir, they can’t communicate. They have no method of communication. So we’re doing we’re communicating like they used to communicate 2,000 years ago with children, bringing a note back and forth. They have no communication. But all I want to see is I want to have a safe world, and you’re not going to have a safe world. Israel will be gone, the Middle East will be gone, and then they’re coming for Europe. And I have to tell you, I’m very disappointed in NATO. Very. I think that NATO, I think it’s a mark on NATO that will never disappear, never disappear in my mind. You know, they’re coming to see me on Wednesday. They’re going to say, oh, we’ll do this. We’ll do that. Now they all of a sudden want to send things, you know. But they said it loud and clear at the beginning when I spoke to U.K., of all I would have said, they would have been the first because they’ve been there, the oldest. And I say, yeah, I’d love to have a little help. I said, no sir, we’d rather wait till you win. I said, I don’t need help after we win. They have two old broken aircraft carriers. Barely work. I said, I guess we can use them. Who the hell knows? I called the general. He didn’t even want them. He said, we don’t really need them. We got. We got the SS Abraham Lincoln, sir. We don’t need them. You know, we have in terms of technology, we had one day, 101 missiles going at 2,700 miles an hour aimed at the Abraham Lincoln, 101 missiles. Out of 101 missiles, 101 missiles were shot down. Unbelievable technology. Ten years ago, five years ago. I don’t know if that would have been possible, but ten years ago, that wouldn’t have been that wouldn’t have been possible. 101 missiles heading to a ship that’s not that far off the coast. And out of the 101 missiles, we shot down all 101. We have weaponry. The Patriots are unbelievable. We have weaponry. That’s unbelievable.”

Unshared mental contexts created by constraints that converge towards a different attractor basin. Are you also divergent, friend?

@kaikuspa @banteg youtube.com/watch?v=B0F7gz…

@sig_prime @banteg Gotta drop that tweet 🙏

jarvis, initialize a four year long ai bear market

@sierracatalina this included

s l o p s l o p s l o p s l o p e v e r y w h e r e s l o p .

The 'knight’s tour' is a sequence of a knight's moves on a chessboard such that the knight visits every square only once.

@TheGeorgePu i once developed a system that scales exactly for this. interesting stuff

A Yale professor found his students all sound the same. Everyone sounds polished. No one sounds original. Now he requires all writing done by hand, in person, while he watches. I wonder what happens to the millions of classrooms that don't have him as a professor.

A potato chip somehow made it through processing, into a bag and sprouted a live roots “This just got weirder. This isn't just a mysteriously sprouted potato chip that made it up the conveyor belt, passed inspection, passed quality control and got fried — No, as it turns out, on closer inspection, these are live pliable, juicy, fresh roots. This potato chip sprouted in the bag after being fried.”

@vxunderground its postAGI brother.

Believe it or not, this was peak AI summarization. This is basically AGI

@Joestar_sann suppression of information. we don't talk about it because of the thing itself.

so let me get this straight all of ai twitter was telling people to buy a mac mini to run openclaw, which is literally just a framework, an orchestration layer that sends api requests to actual ai models. something you can run on a $5/month vps. which is exactly what i do btw but when google drops gemma 4, an actual large language model that you can run and fine-tune locally on that same mac mini, with no api costs, no subscriptions, no third party dependencies, completely yours under apache 2.0 the ai community is silent you were buying $800 hardware to run a wrapper but ignoring the actual ai model that would justify that hardware this tells you everything you need to know about the average iq of ai twitter

claude code core

@Aurelien_Gz @kaleidoscube i feel like ... like feel I .... I fell like this is triggering ... triggering is this like .... this is .... is this ... this is triggerring my OCD.

this game mechanic is pure genius.. move left to rewind, right to forward game » recur wizards » @kaleidoscube

@Realmecore_0 the constraints they interacted with made them converge into a basin different from yours

POV: someone disagrees with you.

@Elaina43114880 that is exactly what The Architect would say, spot on

"Show me something brand new that only an ASI would create" — Generated by GPT Image V2

@pascal_bornet i wonder ….

We might be solving the wrong problem in robotics. That’s what this makes clear. UMI → Universal Manipulation Interface A simple $400 gripper that lets you teach robots by demonstration. You hold it like a tool. Show the task. The robot learns. No teleoperation. No expensive hardware. No robot-specific data. Stanford open-sourced everything → hardware, code, datasets. What stands out to me is the bottleneck. Not algorithms. Data. Teleoperation → ~35 demos/hour UMI → ~111 demos/hour And the data transfers across robots → UR5, Franka, others. The design is surprisingly practical: → GoPro fisheye lens (155° FOV) + mirrors for depth → SLAM + IMU for precise 6DoF tracking → latency matching for dynamic tasks → diffusion policies for multimodal actions Then it scales. Cheng Chi takes this further with Sunday Robotics (with Tony Zhao). A $200 glove → deployed in 500+ homes → ~10 million real-world interactions. Not lab data. Real human behavior. Their robot learns dishes, laundry, espresso → with zero robot-specific data. This is where the shift becomes obvious. From training robots in controlled environments → to learning directly from humans at scale So here’s the real question: Will robotics be unlocked by better models… or by unlocking data? #ArtificialIntelligence #Robotics #AI #Innovation #FutureOfWork