Andrea

Cominciate col fare ciò che è necessario, poi ciò che è possibile. E all'improvviso vi sorprenderete a fare l'impossibile. F. d'Assisi.

[Race 🕊️ @multiplanet1]

A SpaceX recruiter once described what happens when Elon Musk personally interviews a candidate. And it explains why SpaceX has the lowest acceptance rate of any company in aerospace, lower than NASA, lower than Boeing, lower than any defense contractor on earth.

She said the interview doesn't feel like an interview. There's no behavioral questions. No "tell me about a time when." No competency framework. Elon sits across from the candidate and starts asking technical questions at the boundary of the candidate's expertise. Then he pushes past the boundary.

He's not testing whether you know the answer. He's testing what happens when you don't. Does the candidate panic and make something up? Do they freeze? Or do they say "I don't know but here's how I would figure it out" and then reason through it in real time?

She said the candidates who get hired are almost never the ones with the most impressive resumes. They're the ones who reached the edge of their knowledge and kept thinking out loud instead of shutting down. The willingness to sit in not knowingn and work through it publicly is the signal Elon selects for.

She said he rejected a PhD from MIT and hired a self-taught engineer from a state school in the same week. The PhD froze at the edge of his knowledge. The self taught engineer said "I've never solved that specific problem but here's how I'd approach it" and then spent ten minutes reasoning through it while Elon listened.

SpaceX doesn't hire credentials. It hires problem solvers. And the only way to identify a real problem solver is to push them past what they know and watch what happens in the space between knowledge and uncertainty. That space is where rockets get built.

[Dustin @r0ck3t23]

Elon Musk just named AI’s next crisis. It’s not a shortage. It’s a surplus nobody can switch on.

Musk: “By the end of this year, chip production will outpace the ability to turn chips on.”

For three years the world was starved for silicon. Every lab, every government, every company racing to secure the chips that decide who wins the AI era.

That bottleneck is ending. A harder one is replacing it.

Musk: “The chips are going to be piling up and not be able to be turned on.”

Billions of dollars in the most advanced AI hardware ever built. Sitting dark.

Not because the chips don’t work. Because there isn’t enough electricity to run them.

You can’t print a power plant the way you print a chip.

The fabs scaled. The grid didn’t. Now the hardware everyone fought over is hitting a wall capital can’t buy its way through.

Compute becomes abundant. Electricity becomes the most valuable commodity on earth.

Physics doesn’t care about your chip architecture if your data center can’t pull the megawatts.

Every mind that has ever existed was limited by what it could eat. Ours ran on grain. This one runs on the grid.

That has never been solved by thinking. Only by building.

The war isn’t about who can print the most silicon. It’s about who can plug it in.

Whoever solves energy first doesn’t just win. They own the rails everyone else has to rent.

The losers stack chips in warehouses, waiting for power that never arrives.

We built a trillion dollar engine and forgot the fuel.

[Alberto Bagnai @AlbertoBagnai]

Incredibile! Nel 2011 perfino su lavoce si smontavano questi luoghi comuni!

[Mike @MikeLongTerm]

$PLTR $AMD | Dr. Karp and Dr. Su were right! ✍️

Companies are now fighting back. Dr. Karp, @PalantirTech CEO, recently told CNBC that enterprises are privately "unhappy" with frontier AI labs like OpenAI and Anthropic, accusing them of prioritizing "tokenmaxxing" or maximizing AI token consumption to signal activity over delivering real business value and understanding customer needs.

Uber, Coinbases routing to capping token usage or routing to cheaper models to keep cost under control. or Microsoft revoked Claude Code licenses companywide, Priceline imposed token limits after sharp cost spikes, and reports cite Meta, Salesforce, and multiple unnamed firms facing 3x+ budget overruns or $ hundreds of millions in unexpected spend by mid-2026. Analysts note this as an emerging industry pattern, with FinOps and executives describing "existential crises" over token bills; dozens of enterprises are now adding guardrails, though public complaints remain concentrated among high-profile tech firms experimenting at scale.

Dr. Lisa Su anticipated the pivot to inference economics and CPU-dense systems for agentic AI, correctly predicting that token costs, power efficiency, and deployability on standard platforms would determine scalable adoption long before the current enterprise pushback.

Dr. Alex Karp accurately diagnosed the disconnect in frontier labs' approach, calling out "tokenmaxxing" as activity without outcomes; enterprises are indeed demanding real implementation and business-specific value rather than raw volume that inflates bills without proportional ROI.

Together, their independent foresight validates the maturing AI thesis, efficient infrastructure (AMD Helios/EPYC optimized for lowest TCO & $/M Tokens) paired with outcome-focused platforms (Palantir AIP/Foundry) positions both companies to benefit as the market shifts from hype-driven consumption to sustainable, value-driven deployment.

Yes it may look good on the revenue growth for AI Labs to show off on IPOs investors/bankers, but the customers have to find value in those tokens spent where $NVDA & In-house chips on inference claims are just false. At the end of the day,

~Token cost needs to go down more & more particularly inference by owning more AMD chips/racks. In-house chips can make all kind of claims for years, but the bills enterprises paid have to obey economic.
~Enterprises want a thick software OS or solution focused, they do not want to have unlimited budget for "tokenmaxxing" where it is leading to high costs with limited business transformation; success increasingly depends on implementation layers that route tasks, enforce policies, and connect AI to existing workflows.

Not Financial Advice! DYOR!

[WOLF @WOLF_Financial]

MICROSOFT $MSFT CEO SATYA NADELLA JUST PUSHED BACK ON THE BIGGEST NARRATIVE IN AI

He did not mince words.

Nadella took direct aim at the AI executives who have been predicting mass job losses from artificial intelligence, including Anthropic CEO Dario Amodei who predicted new AI systems could wipe out half of entry-level jobs by 2029.

His counterargument: companies that are using AI purely as a cost-cutting tool to eliminate headcount are thinking about it wrong.

The right frame, in his words, is that every company must now possess both "token capital" which is in-house AI capability, and human capital. The two work together, not against each other.

"Yes, it's a lot of change management, it's a lot of displacement, but there is a path."

His vision for what AI-first companies actually look like: a continuous learning system where human wisdom and AI capabilities compound on each other over time. The companies that win will be defined by the tacit knowledge they contain, both human and AI.

[Althea @Althea415]

Anyway, here is Wonderwall post @England win. 😭🩵🫶🏼 🏴󠁧󠁢󠁥󠁮󠁧󠁿 #worldcup #england #oasis @liamgallagher @oasis

[Jawwwn @jawwwn_]

.@JTLonsdale on the SpaceX IPO:

“Betting against this guy is not very smart.”

“This is a critical company in the middle of the biggest Industrial Revolution ever.”

“@elonmusk set goals where he gets extra upside if he achieves them.”

“He’s a guy who doesn’t miss his goals.”

Via @CNBC

[Jawwwn @jawwwn_]

Peter Thiel:

“I certainly would never want to bet against @elonmusk. I would never want to compete head-on with Elon.”

[Milk Road AI @MilkRoadAI]

A single gigawatt of orbital compute requires roughly 200 Starship launches and Elon Musk is not satisfied with gigawatts (Save this).

The target is 100 gigawatts of orbital compute per year which means SpaceX is staring down a launch requirement that no organization in human history has ever attempted at anything close to that scale.

He acknowledges that scaling to gigawatts per year in orbit is a very hard challenge, but then points to something most people have missed entirely, SpaceX has already demonstrated the foundational capability, because building and launching thousands of Starlink satellites per year is the same industrial problem applied to a different payload.

When you understand the orbital compute satellite as a larger version of Starlink V3 with an Nvidia GPU rack at the center instead of a communications payload, the manufacturing and launch scaling challenge stops looking like science fiction and starts looking like a production ramp.

The infrastructure to support that ramp is already being built.

SpaceX is currently capacitizing for thousands of launches per year, two launch towers and pads in South Texas are operational, the first pad at Cape Canaveral is nearly complete, a second is on the way at Launch Complex 37, and additional locations are already in discussion.

As the CFO says it "You need to have those cost curves as you ramp up in volume and time, your costs go down."

The vision he describes for what this eventually enables is striking in its specificity.

He imagines asking Grok a question on his phone, the inference running on an orbital compute satellite, and the answer coming back down through Starlink direct-to-cell, a complete AI query processed entirely in space, from prompt to response, without touching a single terrestrial data center.

That moment, he says, is closer than the industry thinks, with initial capability demonstrations possible as soon as next year.

The bottleneck that stands between now and that moment is not the satellite design, the cooling physics, or the silicon, all of which SpaceX has already worked through.

[Milk Road AI @MilkRoadAI]

Elon Musk just described a project so large that most people will assume he is exaggerating (Save this).

He is not.

In the video, Musk lays out the central problem facing every AI company on earth, the entire global chip industry is on a path to produce roughly 100 gigawatts of AI compute per year.

That sounds like a lot until you understand that his companies alone Tesla, SpaceX, and xAI will need orders of magnitude more than that.

His answer is the TerraFab.

It is a joint chip factory spanning 100 million square feet, ten times the size of Tesla's Gigafactory Texas announced in March 2026, with Grimes County, Texas commissioners approving the full scale facility site just last week.

The goal is one full terawatt of AI compute output per year.

For context, 1 terawatt is 1,000 gigawatts twice the current total electricity consumption of the United States.

SpaceX has already committed an initial $55 billion to the prototype phase, with total investment estimates ranging into the trillions.

Here is why this matters for Micron specifically.

In the video, Musk named Nvidia's Rubin chips as the reference design for TerraFab's first orbital deployments, and said "You're going to need a lot of memory to go with that."

A billion full radical equivalent chips per year, each requiring stacks of high bandwidth memory, that is the demand signal Micron just received from one of the most capital-intensive projects in human history.

And Micron already cannot keep up with what exists today.

Micron's entire 2026 HBM output is fully sold out contracted before the year began.

HBM4 entered volume production ahead of schedule and sold out immediately.

The structural reason Micron wins here is simple.

Every AI chip ever built Nvidia H100s, Rubin chips, custom ASICs, TPUs is useless without high-bandwidth memory stacked directly on top of it.

There are only three companies in the world that supply HBM at scale, Samsung, SK Hynix, and Micron.

Samsung has had quality issues, SK Hynix is supply constrained.

Micron is the only US headquartered HBM manufacturer which matters enormously given CHIPS Act subsidies, domestic procurement requirements, and the political push to keep critical AI memory production on American soil.

TerraFab just made the memory deficit permanently larger.

Come join Milk Road Pro for our full breakdown of Micron and our entire AI thesis just for $1.

Link below!

[Mike @MikeLongTerm]

$PLTR Dr. Karp on @PalantirTech biggest secret to WINNING!

The most effective Sales Secret in the world- AIP 2026

Palantir's most effective sales approach isn't aggressive traditional selling or a huge sales force. Instead, it's letting customers experience the limitations of raw frontier LLMs (from companies like OpenAI, Anthropic) first.

~LLMs solve some problems (generating text, basic reasoning) but create bigger enterprise issues: data chaos, lack of governance, integration failures with messy real-world systems, security/compliance risks, hallucination in high stakes ops, and no "ownership" of the full workflow.

~Customers then realize generic LLM providers treat them as just another user (they "don’t care about you" in the sense of deep customization or enterprise control).

~Palantir's AIP steps in with its Ontology layer (a digital twin of the business that maps data, enforces rules, and powers secure, controllable AI agents). This lets organizations integrate models into their actual operations, maintain control, and "own the means of production" rather than depending on black-box vendors.

Palantir focuses on value creation through deployed, trustworthy systems (often charging based on outcomes or % of saving, or basically FREE), not just software licenses. It reduces reliance on a large sales team because successful pilots and word-of-mouth from real deployments "sell themselves" in a low-trust AI environment.

Source: https://t.co/BS7y8HnNqh

[Milk Road AI @MilkRoadAI]

The greatest living mathematician just said something that reframes the entire AI debate (Save this).

Terence Tao, Fields Medal winner, UCLA professor, and by most measures the most accomplished pure mathematician alive speaking at an OpenAI Forum event in March 2026, and the observation he made is deceptively simple but profound in its implications.

"We lived in a world with cognitive friction until very recently, where every task required us to use our brain. So we didn't really think about it, we just thought this was the cost of doing something intellectual. But now we have AI and the other technologies that can bring these frictions down to zero."

To understand why this matters so much, you have to understand what most research time actually looks like.

Most research time is spent checking cases, chasing references, translating intuition into computation, testing a path, finding it false, and deciding whether the failure taught you anything useful.

As Tao puts it the lower cost of exploration that AI enables means he can now try crazier things and that makes all the difference.

The reason unconventional ideas in science are often abandoned is because the bookkeeping, coding, or literature search needed to even test them is too expensive for what is ultimately just a hunch.

This is where cognitive friction becomes scientific friction, and lowering it does not make taste, judgment, or proof disappear, it makes more weak signals cheap enough to inspect before they are abandoned.

AI is making hesitation less expensive, and that is often where discovery begins.

Tao now uses AI to search literature, write code, make plots and figures, run calculations, and test whether a possible approach is even worth chasing and he declared AI ready for primetime in March 2026 after confirming that in math and theoretical physics, it now saves more time than it wastes.

He had previously called early AI models mediocre but not entirely inept graduate students and then watched as they passed the threshold where the value of acceleration exceeded the cost of correction.

Years ago, Tao predicted that 2026-level AI, when used properly, will be a trustworthy co-author in mathematical research and by his own assessment this year, that prediction came in on schedule.

A 23-year-old used ChatGPT to solve Erdős Problem #1196 , a problem that had gone unsolved for 60 years in just over 80 minutes.

OpenAI's GPT-5.2 Pro resolved another open Erdős problem, with OpenAI President Greg Brockman posting about it in January 2026.

And OpenAI's Chief Research Officer Mark Chen articulated the institutional goal in terms that every investor should internalize, "We care less about winning a Nobel Prize or a Fields Medal, and more about enabling 100 mathematicians out there to do that for themselves."

If AI is genuinely collapsing the cost of scientific exploration not just in mathematics but in drug discovery, materials science, climate modeling, and theoretical physics then the companies building the compute infrastructure that makes that acceleration possible are not just selling chips and cloud capacity.

They are selling the raw material of compounding human discovery, and that is a demand curve with no visible ceiling

[Rob @robertdolci]

@ricpuglisi I designer avevano judo quando Darwin spiego' l'evoluzione della specie, quando ci insegno' che l'uomo viene dopo la scimmia.
E invece..niente consecutio evolutionis

[Milk Road AI @MilkRoadAI]

Elon Musk just made two predictions in the same breath and most people are only paying attention to one of them.

Musk said we might have AI that is smarter than any human by the end of 2026, and no later than 2027.

It is worth noting that Musk predicted AGI by 2025, then by 2026, and is now saying no later than 2027 so his track record on specific timelines is mixed, and the broader AI research community has no consensus on when or even how to measure when AGI has actually arrived.

But the second prediction in this clip is the one that deserves far more attention, and it is grounded in physics rather than philosophy.

Musk said the lowest cost place to run AI will be space and that it will be true within two to three years.

Google and SpaceX are in active talks to build orbital data centers, as reported by the Wall Street Journal last week.

Google's Project Suncatcher is targeting 2027 prototype satellite launches with TPU chips designed for AI inference in orbit.

A startup called Orbital, backed by a16z has its first Nvidia Space-1 Vera Rubin GPU cluster scheduled for launch on a SpaceX Falcon 9 in April 2027.

The underlying logic is straightforward and it maps directly to what Jensen Huang warned about on the ground.

In low Earth orbit, solar power is continuous and roughly 40% stronger than at Earth's surface with no weather, no night cycle, and no grid interconnection queue.

Radiative cooling simply pointing a heat radiator at the cold vacuum of space eliminates the single largest operational cost and engineering challenge of terrestrial data centers.

There is no permitting process, no local opposition, no transformer backlog, and no PJM interconnection queue to fight through.

The catch, as TechCrunch noted, is that today the economics still do not work, satellite construction and launch costs mean orbital compute costs more per FLOP than ground-based infrastructure, and that gap is real.

But the gap is narrowing at roughly the same rate that Starship is improving launch economics, Nvidia is densifying spacegrade compute, and terrestrial energy costs are escalating.

The timeline Musk gave two to three years is aggressive but not obviously wrong given the convergence of those three curves happening simultaneously right now.

[Milk Road AI @MilkRoadAI]

Anthropic's CFO Krishna Rao just said something on camera that most executives spend their entire career carefully avoiding.

Two years into running the finances of a company now valued at $380 billion, a company that went from a $20 billion valuation to nearly $900 billion in preemptive investor offers in under 24 months.

And he told the interviewer that he has had to actively retrain his own brain to stop thinking about revenue in a straight line.

His exact words: "I think humans mostly think linearly and you think incrementally and that's a paradigm I've had to break."

That admission is more important for investors than any price target or earnings revision published this week, because the cognitive failure he is describing internally is the exact mistake retail investors and Wall Street analysts make every single time they look at an AI stock and call it expensive.

DCF models are linear by construction, you take today's cash flows and grow them at some assumed rate, discount them back, and arrive at a fair value that captures none of what happens when a business hits an exponential inflection point.

Analyst price targets are anchored to current revenue multiples, which is why every major AI name has spent the last two years being perpetually overvalue by consensus estimates that got revised upward quarter after quarter as actual results came in orders of magnitude above forecast.

Anthropic itself is the clearest possible illustration of what linear thinking misses.

The company posted an $87 million annualized run rate in January 2024, crossed $1 billion by December of the same year, hit $9 billion by end of 2025, then accelerated to $14 billion in February 2026, $19 billion in March, and $30 billion in April with the current internal run rate now reportedly closer to $40 billion.

Salesforce took twenty years to reach $30 billion in revenue.

Anthropic did it in under three years from a standing start and Rao's own admission is that even he, sitting inside the company watching it happen in real time, has had to consciously rewire how he thinks about the trajectory.

[WOLF @WOLF_Financial]

AMD $AMD JUST REPORTED Q1 EARNINGS

The headline numbers:
- Revenue: $10,253M vs $9,892M est 🟢
- Adjusted EPS: $1.37 vs $1.29 est 🟢
- Adjusted Net Income: $2,265M vs $2,114M est 🟢
- GAAP EPS: $0.84
- GAAP Net Income: $1,383M
- Gross Margin: 53%

Q2 guidance came in well above consensus:
- Q2 Revenue: $11,200M vs $10,516M est 🟢
- Q2 Adjusted Gross Margin: 56%

The strategic commentary:
- "Customer engagement around MI450 series and Helios is strengthening"

The numbers translate to:
- ~3.6% Q1 revenue beat
- ~6.2% Q1 adjusted EPS beat
- ~6.5% Q2 revenue guide above consensus (~$684M above midpoint)
- ~3 percentage points of gross margin expansion expected from Q1 to Q2

A clean beat-and-raise.

[Milk Road AI @MilkRoadAI]

Jensen Huang just reframed the entire history of computing in two minutes.

The argument is deceptively simple, but once you see it you can't unsee it.

Every single piece of software ever built, every app, every website, every search engine, every platform operated on exactly the same fundamental principle.

Someone creates content, it gets stored somewhere and when you ask for it, the system retrieves it.

Google indexes the web and retrieves the right page, YouTube encodes your video and retrieves it when someone clicks, Amazon photographs every product in its catalog and retrieves the listing that matches your search.

Every recommender system, every ad platform, every social feed, all of it, without exception, is a retrieval operation dressed up in a user interface and we called it the Information Age.

But strip away the branding and what you had, for 30 consecutive years, was an extraordinarily sophisticated filing cabinet.

The smartest engineers in the world spent their careers optimizing how fast you could put things in and pull things out.

Generative AI doesn't just improve that system but rather replaces the entire premise of it.

Instead of retrieving content that was pre-recorded by someone else, AI generates it from scratch, in real time, calibrated to your exact context, your specific intent, the precise ground truth of that moment.

The same question asked twice gets two different answers, both tailored to what the system knows about you right now.

There is no file being pulled or a pre-recorded version, the content is being synthesized on the fly from a compressed model of human knowledge, shaped to fit exactly what you need.

The implications of this for the companies that built the retrieval era are profound and already starting to show.

Google's click-through rates on organic search results have dropped 61% since AI Overviews rolled out, because users are getting answers directly instead of clicking through to files.

Gartner projects traditional search engine query volume drops 25% by the end of 2026 as users migrate to generative interfaces.

And yet this is exactly what Jensen predicted, in the old world, the computing bottleneck was storage and retrieval, you needed hard drives, bandwidth, and CDNs.

In the new world, the bottleneck is computation, you need the raw processing power to generate tokens at scale, millions of times per second, for millions of simultaneous users.

Inference computing demand has grown roughly ten thousand times in the last two years alone.

That shift is precisely why Nvidia's revenue opportunity forecast just jumped from $500 billion through 2026 to $1 trillion through 2027.

The retrieval era needed CPUs and storage and the generative era needs GPUs, token factories, and inference infrastructure at a scale never built before and Nvidia builds the engine underneath all of it.

Jensen has been making this argument since 2024. Most people wrote it off as a chip salesman talking his book but two years later, it's the architecture of the entire industry.

[Lance Roberts @LanceRoberts]

Great chart of who owns what in the Treasury debt markets.