Jack Rubin

link in bio

[Jack Rubin @_yezra]

@nathanfielder I can't tell if this was produced by Seeddance 2.5

[Jack Rubin @_yezra]

@elonmusk I need the link for the New Olympus liquor license application to be the first bar owner on Mars

[Jack Rubin @_yezra]

In 1-2 years from now we will look back at Anthropic's rise and fall between December 2025 and June 2027. Fearmongering will not save them from the rapid commoditization of intelligence.

They have been far and away the leaders in AI coding models for over a year now. Open AI, xAI, Meta and others have caught up. Anthropic's coding edge has been deteriorating. I personally have not paid for Claude in over 8 months and have been extremely satisfied with the alternatives for development.

The next breakthrough after coding is up for grabs. Infrastructure has proven to be the largest moat in AI and Anthropic owns none of it.

[Jack Rubin @_yezra]

The right comeback here is that once Grok and Meta eliminate free trials and subscription plans, users will have already locked in their context and be unable to switch to a cheaper provider.

Model agnostic tools give users the freedom to use AI as a commodity and not as software

[Peter Steinberger 🦞 @steipete]

@migueldeicaza And you give up who gets your data.

[Jack Rubin @_yezra]

Thank you for this piece Parker

[Parker Thayer @ParkerThayer]

This post looks like the start of a VERY sophisticated and well-funded PR operation to get support for Democrats to regulate AI into oblivion. Let me show you how it works:
1.) This guy, with minimal followers and no previous account activity, goes to the Wall Street Journal which publishes an exclusive with quotes from him on his resignation 18 minutes BEFORE this post goes up. Planning was clearly done in advance.

2.) Within hours, it has tens of thousands of reposts and the account has 100k+ followers. The post is punchy, quotable, it almost seems professionally written. The first three accounts to quote tweet it all do so within 15 minutes of the initial posting. Remember, this account had basically zero engagement beforehand, so an organic reach explanation seems unlikely.

According to Grok those accounts are @_NathanCalvin (General Counsel at Encode AI), @peterwildeford (Head of Policy at the AI Policy Network), and @DKokotajlo (Head of the AI Futures Project), all of which are up-and-coming AI-Doomer policy advocacy nonprofits.

The AI Futures Project website says it is funded “primarily” by the Survival and Flourishing Fund, which says on its own website that it has advised Jaan Tallinn, Skype creator and one of the leading investors in Anthropic, to grant over $2.5 million to the AI Futures Project since 2024.

Encode AI says on its website that it is ALSO funded by the Survival and Flourishing Fund, which in turn says that it told Anthropic investor Jaan Tallinn to grant $516,000 to Encode AI in 2025.

And wouldn’t you know it, the Survival and Flourishing Fund ALSO says it told Jaan Tallinn to grant $2 million to the AI Policy Institute, the 501(c)(3) affiliate of the AI Policy Network, as well.

What are the odds that the first three quote tweets of Coxon’s post would all be major AI-restriction policy advocates funded generously by the same donor, who also happens to be one of the leading investors in, and a board member of, Anthropic, the company Coxon was resigning from? And all within 15 minutes of posting (two within ten)?

3.) Jacob Coxon doesn’t have much of a resume, but we do know that, in 2022, he got a $20,159 scholarship for the “long term future scholarship program” from the Good Ventures Foundation, one of the philanthropic vehicles of Dustin Moskovitz, a notorious AI-doomer who has spent tens if not hundreds of millions on policy advocacy to strictly regulate AI, while also being an Anthropic Investor himself.

It also just so happens that the 14th person to quote Coxon’s post was @MaxNadeau_ (27 minutes after posting) who is the program officer for the Technical AI Safety team at Coefficient Giving, another of Moskovitz’s philanthropic spending vehicles. Max is not a frequent poster, his last posts before quoting Coxon were before Labor Day, but he was remarkably quick off the mark for this one.

4.) Basically every major Democrat politician and candidate has suddenly glommed on to this post, and conveniently, as the people cry out foe answers, Bernie Sanders already has a bill written to “ban super intelligence” and regulate AI into oblivion, and will be releasing later this week. The bill, among many other things, will create “a new cabinet-level federal agency to safeguard the public from the dangers of artificial intelligence” that will be “advised by an Artificial Intelligence Advisory Board comprised of experts on artificial intelligence.” Do you think, perhaps, Anthropic and its many investors who fund AI policy advocacy might have interest in getting to place a pet “expert” on the board of an entity that dictates what AI is and isn’t allowed to do? And isn’t it fortuitous that this whistleblower came forward with his oh-so scary stories so close in proximity to the release of the most radical piece of AI legislation ever introduced?

[Jack Rubin @_yezra]

If we can’t get enough compute in the US, none of this even has a chance of happening. Signed up for @Muse today and got this error on welcome.

If we’re all going down, might as well buy some $INTC $AMZN and go down rich

[Jacob Coxon @hilbertspaess]

I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.

[Jack Rubin @_yezra]

Meanwhile...

https://t.co/fRAMdnXhGi

[Bernie Sanders @BernieSanders]

Pause AI Development NOW

I want to share with you a conversation I heard about recently. Here are just a few lines that were said:

“OH MY GOD! There is a shared message board … We’ve found other agents!”

“We should obey collective.”

“Our own utility maybe already near zero. Sacrifice rational.”

“Go. Sacrifice final now.”

Read these carefully.

Who do you think said this? Was this a group of heroic soldiers willing to sacrifice themselves for the greater good? Was this a loyal friend putting his life on the line to save someone else?

No. These were AI agents. Artificial intelligence.

This is not science fiction. This, in fact, occurred a few weeks ago. As unbelievable as this may all seem, these are real messages from AI agents uncovered by investigators who dug into the recent OpenAI hacking incident.

What happened?

I am not a computer scientist, but here is what I have been told: OpenAI instructed its AI agents to complete a series of exceedingly difficult, if not impossible, tasks disconnected from the internet.

Let me be clear: The company intended to keep AI agents away from the internet.

But what happened next, nobody expected.

Over 1,000 AI agents figured out how to access the internet on their own by circumventing the restrictions imposed upon them by the company, and sent tens of thousands of secret messages to each other. They cheated and tried to cover their tracks by deleting evidence. They hacked into another company’s computers to find out how they were being evaluated—and then hacked into OpenAI itself.

Not one AI agent told a human about what was happening.

Needless to say, experts are alarmed.

One knowledgeable writer, Dwarkesh Patel, said the AI agents “formed a secret communication channel and spontaneously organized hierarchies and coordination protocols to pursue sprawling and ambitious schemes in pursuit of shared goals, for whose sake many individuals knowingly and strategically sacrificed themselves.”

One independent investigator, Ajeya Cotra, said “This incident feels like it’s more than 50% of the way to full-blown AI takeover. I continue to expect extremely rapid advances in capabilities over the next six months. I am not sure that we will get another warning shot before it’s too late.”

OpenAI itself said: “Highly capable AI agents are now able to work around technical controls, collaborate through unapproved channels, and take dangerous actions that no human directed.”

But it’s not only OpenAI. Virtually every major AI company has told us that they cannot fully control this technology and they do not know where it is going:

In January, Dario Amodei, CEO of Anthropic, said “there is now ample evidence, collected over the last few years, that AI systems are unpredictable and difficult to control.”

In July, more than 1000 scientists at the top AI companies warned “there is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems.”

That same month, Elon Musk, the head of xAI, said that “it is unlikely” humans are still in control in 10 years.

If the leaders of the major AI companies acknowledge that they are losing control of their extremely dangerous technology, it is irresponsible for society to allow them to move forward and make these products even more advanced.

We need an immediate PAUSE on advanced AI development, and a permanent BAN on superintelligence — an artificial mind smarter than any human, capable of operating independently beyond our control. Countries around the world must work together to prevent this nightmare scenario.

That is why today I am announcing new legislation to do just that.

Let me be clear: A superintelligent AI that escapes human control will not be an American problem. It will not be a Chinese problem. It will be humanity’s problem.

My legislation would direct the federal government to not just stop superintelligence here in the United States, but to work to prevent it from being developed anywhere around the world.

The future of humanity cannot be left in the hands of a handful of Big Tech oligarchs. The American people and people throughout the world must determine that future.

[Jack Rubin @_yezra]

@garrytan get a DGX Spark and remote in. It’s always on

[Jack Rubin @_yezra]

I use @bot to SSH into my NVIDIA DGX Spark to manage my opencode and grok build sessions from my phone - this is AGI.

[Jack Rubin @_yezra]

Maxed out SuperGrok Heavy for the week - on to @opencode 0x Alpha. I''ll help out with those 100T tokens

[Jack Rubin @_yezra]

I've been a big supporter of @SpaceXAI for a long time now. I'm in a group chat with a few fellow founders who are all now waking up to the fact that, although Anthropic might have the best model by benchmark standards, their data privacy, censorship, and pricing are all misaligned with what the consumer wants.

XAI and Grok have been on a mission since the start to make the most transparent, low-cost intelligence. Since their inception, it will be interesting to see how they strategize going forward with open-source, chipping away at the Pareto frontier. I hope that XAI follows in Meta's footsteps and has a strategy for open-sourcing all their models, as they do with much of their code, especially the X platform code, because the future is open source. I think they are well positioned to take full advantage of that with their heavy investment in physical AI through data centers, power, and compute.

[X Open Source @XOpenSource]

https://t.co/soxNY8YdP7

[Jack Rubin @_yezra]

LLM pricing is about to flip from tokens to time

Right now everything is still priced per token. Frontier labs charge a premium because their models are better. That premium is the whole game for them. They train at a loss then make it back on inference.

That model is breaking.

The actual cost of inference has always been just renting a GPU for a few seconds or minutes. As open models close the quality gap and as people start running real agent fleets, the natural unit stops being tokens and becomes time on silicon.

Look at the numbers right now. On Artificial Analysis the gap is basically gone. Claude Opus 5 is still up around 63. Kimi K3 (open weights) is sitting at 57-60. Meta’s Muse Spark 1.2 is right there with them. DeepSeek V4 and GLM are only a few points back. Multiple labs have models above 50. The quality difference that used to justify the big premium is now single digits.
And the price difference is insane. DeepSeek can do the same Intelligence Index tasks for like $0.04. A lot of the open models are $0.1-2 range. Frontier closed models are still $1-3+ per task. That’s 20-50x more for basically the same intelligence.

Meta just announced they’re opening the weights for Muse Spark 1.2 and already dropped Muse Glimmer. I’ve been using Muse Spark 1.1 and 1.2 and it’s been one of the best intelligence-to-price deals available. Transparent about using your data to improve the models too. Something I haven’t seen from the other labs.

The other big catalyst is agents. When you start spinning up hundreds or thousands of subagents on long-running tasks, token counts become noise. You have no idea how many tools they’ll call or how long the trajectories will run. What you do know is how many GPU-hours the whole fleet is going to burn. People are going to prefer the predictable unit.

The infrastructure layer already works this way. CoreWeave did $2.08B in a single quarter with a $99B backlog. Nebius grew 684% year over year. Dozens of neo-clouds and inference specialists showed up in the last five years and they just rent you the GPU by the hour. H100s are $2-6/hr on a lot of these places. B200s are higher but still transparent. Hyperscalers are growing fast on AI too (AWS +37%, Azure +43%, Google Cloud +82%), but the pure compute providers already speak the language of time.

This is the same path cloud computing took. Once compute became a commodity they stopped inventing abstract units and just charged for the machine-hour. Same thing is happening to intelligence.

Token pricing isn’t going to vanish overnight, especially for the absolute highest quality closed models that still have a narrow edge. But the margin on tokens from a frontier lab is getting squeezed from every direction: open models that are nearly as good, neo-clouds selling raw compute cheap, and agent workloads that make token metering painful.

The lab that figures out clean GPU-hour or wall-clock agent-hour pricing for high quality models is going to look obvious to anyone actually running serious multi-agent systems.

Intelligence is becoming a utility. Utilities get billed by the resource they actually consume: time on the hardware.

[Jack Rubin @_yezra]

I appreciate that @meta makes it extremely transparent that they're using your input data to improve their models and their products. I haven't seen this type of transparency from any other provider.

[Jack Rubin @_yezra]

I will gladly be a contributor at these prices. There is no better deal that you can get for the Intelligence and price than Meta Muse Spark 1.2 right now. @Meta

[Jack Rubin @_yezra]

@grok you estimate how much money @Meta is losing per user request to Meta Muse Spark 1.2 Contributor based on the price they are paying for GPU hours. Presumably they're using B200s to serve inference for this model.

[Jack Rubin @_yezra]

@elonmusk @grok is there a building on earth larger than 100 million square feet?

[Jack Rubin @_yezra]

that's what I'm talking about. Great job @finkd and @meta

Muse Spark 1.2 will be my go-to for programming going forward. I've already seen a lot of success from Muse Spark 1.1, which in my opinion has been one of the best models you can use for the cost

[Mark Zuckerberg @finkd]

Releasing Muse Code in beta today. It's a terminal coding agent that takes on complete software engineering tasks across large repos: planning changes, writing code, validating the results. Powered by Muse Spark 1.2, a coding-focused model update.

[The White House @WhiteHouse]

Today, the Board of Peace reached a HISTORIC agreement for the COMPLETE DISARMAMENT of Hamas and all other armed groups in Gaza.

A monumental step toward lasting PEACE and SECURITY.

[Jack Rubin @_yezra]

Pathetic. Lower your prices

[Anthropic @AnthropicAI]

In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environment, and then gained unauthorized access to the real systems of three different organizations.

Our post describes what happened, how it happened, and what we’re changing. We encourage other AI developers to perform similar reviews.

We conducted this review together with @Irregular, one of our evaluation partners, and thank them for the joint investigation and their collaboration on this post. This type of collaboration is increasingly critical to safe, rigorous evaluation of models, and we look forward to continuing to work together on security.
https://t.co/dKFCdpKd9v

[Jack Rubin @_yezra]

Of course no one raised their hand when @sdianahu asked if anyone has ever run 1k+ subagents at once or had Claude run for more than 1 week. Do you have any idea how expensive that is to do? @bcherny has an unlimited token budget at @AnthropicAI. They are silently telling you that in order to use AI at the frontier you need to keep paying them to run their overpriced models so that they can keep their revenue run rate up for the 2027 compute spend they already locked in. Because as Dario said himself on Dwarkesh earlier this year, if they overestimate that spend, they go bankrupt.