Ambisphaeric

My unyielding, unrelenting, unbreakable optimism on this point comes from decades of pessimism and dooming out over nothing. The amount of pipeline growth this develops downstream is going to simply facemelt the humans into AI Acceleration and Capitulation.

Interesting project. Probably not for most in its current form but architecturally, fairly sound.

Current bear case estimate of TAM on a functional, capable human-dexterity robotic hand is approximately 100x-1000x the current size of the entire global economy.

The Grand Bargain.

They're going to need to print trillions. People are going to misread this as currency debasement. It'll actually be necessary due to growth, which is a net positive for everyone. People may still misinterpret this growth as a bad thing.

@DrRaySays @purpostition @elonmusk @xenocosmography Europe actually needs mass deportations, scaled 100x more desparately than the United States. Europe is not going to accept Iranian refugees, even Merkel has accepted they went off the rails. Trump is not accelerating White genocide. Net migration is DOWN.

@purpostition @elonmusk @xenocosmography Fuentes lol

1) Help to defeat MAGA. 2) Democrats flood the US with ~100,000,000 virulently hostile aliens from the third-world. 3) ??? 4) Total America First victory.

@romlib_ @1nterchangeable @ChadNotChud He’s both correct and incorrect at the same time, slanted toward pessimism. It’s not that they aren’t a “normal” technology. It’s more a story of realistic and pragmatic human agency.

@1nterchangeable @ChadNotChud I was curious what made delaniac tweet this since I feel like he has been using them extensively a lot too, so maybe he knows something I dont. It doesn't seem implausible to me that LLMs stagnate, but my priors just dont go that way right now.

increasingly clear that LLMs are just a Normal Technology cool and useful, but fantasies that their rate of improvement will continue indefinitely are just that like everything, we’ll reach a point where further improvement requires a qualitatively different approach

@xw33bttv @sdmat123 Consider how this will play out in court: > Kimi lawyers: “Yes, Your Honor, we lifted the data to train our model. But Cursor cannot steal our stolen technology. It’s against the terms!” > Cursor lawyers: “Yes, Your Honor, we lifted the lifted data and post-trained the model.”

@sdmat123 Yeah, I was seeing non-stop praise on X most of the morning too. It seems so silly to do something like this... like it's all risk no reward.

Cursor AI may be in material breach of contract with their new Composer model, which is generating buzz online for reportedly reaching Opus-level performance. It’s alleged that the new model is a fine-tuned checkpoint of Moonshot’s Kimi K2.5. If true, the original model is licensed under a modified MIT license containing a clause restricting commercial use by products with over 100 million monthly active users or more than $20 million (or equivalent) in monthly revenue. The requirement for triggering that clause is simply to prominently credit “Kimi K2.5” in the product or service’s UI. Cursor could now face serious PR and legal issues simply because they couldn’t be bothered to cite the underlying team’s work. This is open-source 101.... Talk about being hoisted by your own petard.

Millions and millions of people completely worried about the wrong thing all at the same time.

@LyraInTheFlesh @DaveShapi The scale of the bubble is mostly limited to private companies with high valuations in assets 99.9% of people have zero exposure to, and Google/NVIDIA. Google is likely not in nearly the scale of bubble people might think. Even if it were, it doesn’t matter that much.

I am both incredibly excited by AI, believe in its transformative inevitability, and I also believe we're in a bubble that must pop. It feels less like an "AI Bubble" and more like an unsustainable "OpenAI Bubble" though. When it pops it's going to impact everyone for a bit...

People who are wrong: - Degrowthers - Decels - Doomers - "AI is a bubble" - "AI is hitting a wall" - "Data centers are bad" They're ALL WRONG.

@laninoel @temporarymine @DaveShapi Perception compresssion would do just as well here.

@temporarymine @DaveShapi Doesn’t have to be infinite. All it has to do is give us something that would have otherwise taken 50 years and that’s enough

@DaveShapi This is remarkably accurate. The growth curve is going to continue to be extraordinary. I completely understand the internal model that leads to this line of thinking though. It’s very reasonable, it’s simply missing some critical vuln patches.

@realAtlasPress So diffusion, not transformers?

@ATLNTIS777 @xenocosmography I’d explain.. But I don’t think you’re capable of understanding.

@Ambisphaeric @xenocosmography Can you show me a comprehensive plan by the Trump admin to transition from fruit pickers to automation? Where have they said that? What bills are they trying to pass in order to achieve this? What plans have they stated? Why is it always head cannon with you people?

@ATLNTIS777 @xenocosmography Identify the cope. Or take the job picking fruit. Choice is yours.

@Ambisphaeric @xenocosmography This fucking cope is unreal

@ATLNTIS777 @xenocosmography Are you going to pick fruit now? Not next week. I mean, are you going to go find a job picking fruit RIGHT NOW when it’s needed? Didn’t think so. Don’t worry about that, lazy ass. In a few years, it’ll be entirely automated and there’ll be a way for you to benefit from it.

@Ambisphaeric @xenocosmography This is literally what Trump wants to do with fruit pickers btw

@xenocosmography 3) Provide amnesty for hostile aliens so that things can get even worse, but faster.

@xenocosmography 5) There’s no number five. Total America First Victory means 100,000,000 to 200,000,000 New Americans making America much worse, faster, but also means Total America First Victory.