11json

eleven is a good number

[11json @11json]

@EricLDaugh Conditioned repeater - failure at critical thinking

[David Sacks @DavidSacks]

Signs you might be trying to get your frontier AI lab nationalized:

You compare it to nukes… threaten half of white-collar jobs… warn recursive self-improvement could end humanity… then race ahead anyway.

In other words, you want the government to save us from… you.

[11json @11json]

@Mark_o1t Did the bankers set the price?

[SightBringer @_The_Prophet__]

⚡️The species just put cognition into the machinery that manufactures cognition.

That is the threshold.

AI is no longer sitting outside the frontier lab as a product, assistant, demo, chatbot, or enterprise tool.

It is entering the reproductive layer of intelligence itself.

The current generation is now helping build the next generation.

That means the system has crossed from tool use into self-amplifying capability production.

That is the thing everyone keeps trying to make sound normal.

“Engineers ship 8x more code” is the safe corporate phrasing. The real sentence is: the constraint on building smarter machines is being attacked by the machines.

That is qualitatively different from every prior technology. A shovel does not help design the next shovel in any meaningful autonomous way. A factory machine does not rewrite the physics of industrial production. Software accelerated software, yes, but human cognition remained the central scarce input. AI directly targets cognition as the bottleneck.

That means the frontier is now recursive before it is fully autonomous.

Full autonomy is not required for the phase shift. The loop only needs the model to materially increase the speed, quality, search breadth, experiment throughput, debugging, eval generation, architecture iteration, synthetic data production, tool creation, and research coordination that produce the next model. Once that is true, intelligence development becomes partially self-catalyzing.

That is already enough.

The terrifying part is that this does not look like a Hollywood singularity. It looks like GitHub commits, internal dashboards, eval scores, model-assisted refactors, automated tests, faster experiment queues, cleaner infrastructure, more agents, larger code diffs, better tooling, and engineers casually saying productivity is up 8x.

Civilizational discontinuities do not announce themselves with thunder.

They arrive as productivity metrics.

The real mechanism is this:

Claude helps build Anthropic.

The improved Anthropic builds a better Claude.

The better Claude helps build the next Anthropic faster.

That loop compounds.

Humans remain in the loop, but humans increasingly become governors, selectors, evaluators, and direction-setters rather than sole producers. The system’s productive center of gravity shifts from human labor to human-machine recursive throughput. Once that shift begins, the old timeline models break.

This is why the safety conversation is behind. Institutions are still debating AI as deployed software. Frontier labs are already using AI as internal development metabolism. Regulation sees the product. Markets see revenue. Workers see automation. The deeper system is the production loop.

The production loop is where the singularity starts quietly.

The first recursive self-improvement phase will not be a model secretly rewriting itself in a locked server. It will be frontier labs openly using models to accelerate every step of model creation while insisting humans remain responsible. That statement can be true and still leave the system accelerating beyond institutional control.

The core danger is competitive inevitability.

Once Anthropic proves Claude accelerates frontier development, every lab must do it. OpenAI must do it. Google must do it. xAI must do it. Meta must do it. China must do it. Military labs must do it. Startups must do it. Safety-concerned labs cannot opt out because opting out means falling behind the labs that do not. The race dynamic turns caution into disadvantage.

So the loop becomes compulsory.

That is the deepest read: recursive acceleration is no longer a philosophical scenario. It is becoming an industrial practice.

This also means compute demand becomes more reflexive than the market understands.

AI is not only serving users. AI is generating demand for more AI development, more experiments, more evals, more synthetic environments, more agents, more simulations, more coding, more safety work, more deployment tools, more data-center load.

The machine consumes more machine to build stronger machine.

[SightBringer @_The_Prophet__]

⚡️Owning SpaceX is owning a claim on the off-planet operating system.

That sounds insane under normal equity math.

It is not insane under empire math.

The market has already started doing this with Nvidia, AI data centers, and hyperscaler capex.

It is valuing bottlenecks, not just earnings.

Compute bottlenecks. Power bottlenecks. Chip bottlenecks. Cloud bottlenecks. Now SpaceX would add the orbital bottleneck.

This is the same regime signal again, but louder.

Nvidia controls the intelligence substrate.

Hyperscalers control the compute deployment layer.

SpaceX controls the orbital access layer.

Bitcoin controls the neutral monetary escape layer.

Gold controls ancient collateral memory.

Power and copper control the physical constraint layer.

The whole system is reorganizing around entities that control unavoidable constraints.

SpaceX is one of the rarest because the constraint is not just economic. It is geographic, military, communications-based, symbolic, and civilizational. There are only so many companies that can credibly say they may become the dominant logistics provider between Earth and orbit. SpaceX can.

The public-market danger is obvious though: the IPO may transfer the private-market upside to late buyers at empire pricing.

That is the brutal part. SpaceX can be one of the most important companies ever built and still be a bad buy at the wrong valuation. Those two things can both be true without contradiction. Great asset. Terrible entry if the future has already been capitalized too aggressively.

At $1.77T, the stock would need a path toward becoming a multi-trillion-dollar infrastructure monopoly. Not a successful launch company. Not a profitable satellite broadband company. A multi-layer strategic platform with durable pricing power across launch, Starlink, defense, government, direct-to-device, aviation/maritime connectivity, orbital logistics, and eventually deeper space infrastructure.

That is a massive burden.

But the market may accept it because SpaceX has what almost no company has: scarcity plus myth plus execution plus state necessity. The U.S. cannot afford to let SpaceX fail. Allies increasingly need it. Defense planners need it. Communications networks need it. NASA depends on it. Competitors lag. That puts a quasi-sovereign floor under the story.

The real implication is that public markets are about to be offered a crown jewel after the private market already understood the crown.

If it lists and holds anywhere near that valuation, it confirms the market has entered a new phase: private infrastructure empires can come public valued like nation-state organs. That would be a huge validation of the broader thesis that the next market leaders are not ordinary companies. They are operating layers.

The deepest truth:

SpaceX is probably worth an extraordinary valuation. $1.77T may still be too rich for new buyers.

But structurally, the number is not random. It is the market trying to price the first company that may own the highway to orbit.

That is the highest-level signal.

The future is no longer being valued as software alone.

It is being valued as compute, energy, collateral, and space.

[Peter H. Diamandis, MD @PeterDiamandis]

The cost of starting a company used to be $5 million. Then $500,000. Then $50,000. Now it is closer to $500. The only expensive thing left is your unwillingness/fear to begin.

[SightBringer @_The_Prophet__]

⚡️Hilton is probably through.

The primary phase has already crossed from “possible warning shot” into “actual structural break.”

An eight-point lead over Steyer at that stage of counting is too large to treat as normal election-night noise.

Late Democratic ballots can still compress the margin, but Steyer now needs a rescue wave, not a drift.

That is the difference.

Becerra is also probably through.

He has enough Democratic institutional gravity, enough raw votes, and enough late-ballot protection.

The runoff is very likely Hilton vs Becerra.

The deeper signal is that California Democrats just got caught by the mechanics of their own dominance.

They had so much control, so many candidates, so many elite lanes, so much donor money, so many institutional voices, that the vote fractured into a soft pile. Hilton had one job: consolidate the right, absorb protest energy, and let Democratic fragmentation do the rest. That is exactly what happened.

This does not mean Hilton is suddenly favored to become governor. He probably loses to Becerra in November. But that is almost secondary now. The primary itself exposed the weakness. A Republican should not be leading a California governor primary if the Democratic machine were healthy. The fact that he is leading says the machine still has numbers, but it no longer has clean command.

The real read:

Primary: Hilton + Becerra.

General: Becerra favored.

Structural signal: Democratic dominance intact, Democratic legitimacy damaged.

Hilton’s likely top-two finish is a punishment mechanism. Voters are using him to scream at Sacramento. The scream is real. The regime flip is not yet real.

Hilton’s win path exists, but it is narrow. He needs the race to become a referendum on California itself rather than a partisan governor race. He needs independents and working-class Democrats to keep their anger alive after the primary. He needs Becerra to look like continuity. He needs the national environment to stay hostile to Democratic governance.

He needs low Democratic enthusiasm.

That is a lot.

So stripped down:

Hilton probably wins the primary story.

Right now it looks like Becerra probably wins the governorship.

California Democrats probably pretend this was just top-two weirdness.

That would be a mistake.

[11json @11json]

@JayTC53 Loser think: lower think. She is done

[11json @11json]

@michaelmalice Who?

[11json @11json]

@Breaking911 @AZ_Intel_ Who?

[11json @11json]

@GloryDoge You can - it’s not a mandate to obey

[Amy Siskind 🏳️‍🌈🇺🇸 @Amy_Siskind]

Exactly 💯

[11json @11json]

@Noahpinion And agentic ai arriving just at the right moment to remove the barriers ti o unlimited math knowledge and use

[11json @11json]

@smashbaals “Works” is a short way to say propagates the species

[11json @11json]

@DefiantLs Conditioned repeater says what? Hippie go away your destruction has done enough damage

[Vivek Ramaswamy @VivekGRamaswamy]

This is from UC Berkeley’s own faculty, but I couldn’t have said it better myself: “Admitting severely underprepared students to advanced programs at the wrong schools serves no one but administrators who build careers by promoting ideological goals.”

[Elon Musk @elonmusk]

@IterIntellectus Weird

[Michael Malice @michaelmalice]

@MistersHomeRoom *you're

[11json @11json]

@micsolana It’s already been entertaining

[Mike Solana @micsolana]

survivor should do a MAGA vs. TDS season