arepo

yap station - views are my own and my own only

[arepo @___arepo]

Micron is financing its own funeral for $250M

[arepo @___arepo]

AI Slop but it gets my idea nicely:

Game 1: AI-memory — three nested games stacked on top of each other
The supplier game: a repeated prisoner's dilemma that finally converged. Analogy: three petrol stations on the same street. Any one can cut prices to steal share (defect), but if all three cut, everyone bleeds. Classic theory says with only 3 players, infinite rounds, and perfectly observable moves (fabs take 3 years and everyone sees the construction cranes), cooperation is sustainable without ever talking — each player knows defection triggers mutual destruction, and they all still have 2023's $4.50/share losses tattooed on their memory. That's the "capacity discipline" everyone praises. Earnings calls are the coordination device: when Micron's CFO says "we will remain disciplined," that's not investor reassurance — that's a public signal to Samsung and Hynix: I'm cooperating, your move. Game theory calls it cheap talk that becomes credible through repetition. The June antitrust suit is the referee asking whether tacit coordination (legal, emergent) crossed into explicit coordination (illegal, intentional). The dark joke: the folk theorem says they never needed to actually collude to produce collusion-shaped outcomes.
The buyer game: a bank run in reverse. In a shortage, allocation goes proportional to orders. If you're a hyperscaler and order exactly what you need while rivals pad their orders, you get shorted on allocation. So the dominant strategy is over-ordering — even though everyone over-ordering creates the phantom demand that deepens the shortage. Everyone knows the order book is partly fiction; nobody can individually stop lying. That's why Micron extracts $22B of prepayments and take-or-pay terms — those are commitment devices to separate real demand from strategic padding. And note the CEO-level meta-game: for Zuckerberg or Nadella, over-spending on AI is a forgivable mistake, under-spending and losing AI is a career-ending one. Asymmetric payoffs → rational arms race regardless of actual ROI. The capex isn't (only) a demand forecast; it's insurance against being the CEO who missed it.
The Nvidia game: deliberately engineered circularity. Nvidia invests $2B in Coherent with purchase commitments, backstops neoclouds, partners with Micron — investing in its own customers and suppliers so the demand signal up and down the chain stays strong. That's not emergent. That's a player intentionally wiring the board so every other player's "independent" demand signal partially routes through Nvidia's balance sheet.

[arepo @___arepo]

@rickyho_1989

[arepo @___arepo]

@lucaxyzz he was too close to the sun

[arepo @___arepo]

@KnightDark19816 @zerohedge let me know when you buy next time blud

[Palantir @PalantirTech]

Our thoughts on the importance of AI sovereignty.

1. Your AI sovereignty dictates your institution’s future. Sovereignty is the precondition for choice. Relinquishing sovereignty transfers the future choices of your institution to others, who are likely to exploit it for their gain and your loss.

2. Data retention is your treasure. Transfer it at your own peril. Your ability to win is dictated by your ability to recognize and use your unique edges, and you keep winning by compounding the underlying data to generate new insights. Transferring that data hands over access to your pre-existing winning plays and yields the means of production for new ones.

3. Tokenmaxxing hijacks your value orientation and decreases your institutional fortitude and intelligence. The pursuit of high token usage incentivizes disposable scripts over robust software — with the addictive feeling of false progress. There is a reason why those selling tokens refuse to charge based on value.

4. Controlling your weights is controlling your fate. Weights are the distilled form of hard-won, accumulated institutional knowledge. If you let others control your weights, you are allowing them to migrate the alpha of your business to theirs.

5. There is no contradiction between sovereignty and alpha. The architecture that maximally preserves sovereignty is one that enables institutions to own their tribal knowledge, and to compound it as alpha.

6. Politicizing the technical issues involving sovereignty is what your adversary wants. Techno-politicization is the wellspring of false sovereignty. Techno-politicization drives decisions that seem to reduce dependency, but ultimately limit agency — especially on the battlefield in the West.

7. Real expertise is existential. Allowing politics or favoritism to determine your technical decisions rewards whoever is best at politics, not whoever is right. Listen to those closest to the problems, not those speaking most compellingly about them.

8. Learn from institutions that are winning or that have consistently delivered. Institutions facing existential threats do not have the luxury of making technical decisions based on political preferences.

9. Only listen to institutions, countries, and people who have a proven record of being right. A track record of correctness is the best and only signal for future correctness. Judging something as right or wrong based on who you like is exceedingly misguided.

[arepo @___arepo]

@theBuoyantMan @signulll Most of the companies I have worked with are deep in the Microsoft workflow. It is too much of a friction of them integrating another solution. Not mentioning cost.

[arepo @___arepo]

@signulll @theBuoyantMan your day to day uncs using teams and microsoft products are not on twitter buddy

[arepo @___arepo]

the one who saves him will win the 2029 election

[fin @awaaannnn__]

hasil putusan sidang vonis nadiem makarim :
- 10 tahun penjara
- pidana denda Rp 1M
- pidana tambahan pembayaran uang penganti senilai Rp 809 M sekian, apabila tidak dibayar dalam kurun waktu 1 bulan maka diganti dengan pidana penjara selama 5 tahun

[BULL OF BRITAIN @BULLOFBRITAIN]

Anyone selling Robotics is grifting hard.

This trade is way too early and will end the same way as CPO did.

> revenue isnt justifying the valuation because its too early, so stocks roundtrip.

These names are good to have mentally tabbed till the time comes.

But there has also been huge success in “retardmaxxing” on the hype so maybe I should just bid, despite it having no value at the minute.

[arepo @___arepo]

Bullish on hyperscalers

[Gergely Orosz @GergelyOrosz]

I talked with a few folks inside Anthropic and I am starting to understand what @karpathy is saying (and what lots of people are misunderstanding)

It's not about Slack, but about a cloud AI, hooked up to ALL internal company systems, that "just works." THIS is the breakthrough

[arepo @___arepo]

NVIDIA customers will in the long run be its competitors

[Brian McCormick @bjmtweets]

The best place to be in semi space is $AMZN $GOOG and $MSFT with your own cloud specific TPUs

Or $TSM since you fab everything

Neoclouds like $NBIS $CRWV and $IREN get the short stick. Will be dumped by hyperscalers as soon as token supply catches demand and contracts run out, and hyperscalers retain an edge by having cloud customers, more efficient energy, and more economies of scale.

[arepo @___arepo]

@thdxr @grok generate the normal distribution meme

[arepo @___arepo]

@Ajtonyothin if you zoom into it, it would still be similar still

[arepo @___arepo]

If China is an Engineering State and the US a Lawyerly Society, What is Indonesia?

[arepo @___arepo]

Start of the arms race

[arepo @___arepo]

@rickyho_1989

[Elon Murz | SVP of Meme Engineering @ecommurz]

With no meritocracy, "indonesia maju" is just wishful thinking