ftrpoly @ftrpoly

USMNT fans seeing that Folarin Balogun can play against Belgium

In today’s @BostonGlobe

My interpretation of this: Right now, Anthropic and OpenAI are making a killing by selling enterprise FDE services to F500s, building workflows for them on top of proprietary models, then using the traces and context from this to build RL envs to improve the models. This is crazy amounts of leverage - instead of buying this data they're getting paid gigantic consulting fees to extract it. This also goes way beyond typical consulting in scope - organizations are effectively outsourcing key learning curves and domain knowledge to the AI labs. Despite that, it's so far been worth it for them because the value of skilled FDE is so high and the ROI so fast, and orgs are willing to pay a premium for competent AI implementation. But in the long run, one of two things happens: either orgs are gonna get hooked on this and end up paying for the model training that replaces their business, or they find a way to build and own their own model ecosystem. What that looks like is developing some combination of AI models, evals, RL envs, and workflows. Initially probably the model will still be an off-the-shelf frontier model from a top lab. But as firms build out more sophisticated eval / RL env (increasingly the same thing) infra, it starts to become viable to post-train an custom model on top of an OSS base. Cursor have done this successfully with their Composer model RL'd on top of Kimi. Sidenote, this is the same conversation that a lot of national governments in Europe are having in the past week. When we look at what the rhetoric about 'sovereign AI' in the UK actually boils down to, it's doing custom post-training on top of an OSS model, and then running it on local GPUs. Ultimately, the current feeding frenzy for AI services in all of its guises - FDE, AI consulting, etc - should raise questions about long-term sustainability. If consulting services are truly a value add and competitive advantage, then in the long term you want to in-house.

Satya Nadella @satyanadella

a month ago

x.com/i/article/2065…

3K 8K 42K 66.7M 56K

The only way this move is rational is if the U.S. admin is absolutely certain Chinese open source models are about to fall badly behind as Mythos level models employ anti-distillation. This is in fact more bullish for the AI trade than anything else including limiting international TAM. As long as America can export the products it creates with AI to the rest of the world the AI TAM is global. And the last year of tariff wars has demonstrated that the direction of travel is in favor of American export access. This is bullish AI. The biggest risk was always the Chinese.

Interesting snippet from Situational Awareness given the recent government interest in taking stakes in OAI/Anthropic

@BoringBiz_ Ice cold take, so many menial tasks that no longer require human input aside from a quality check after completion

TRUMP PUMP FACTOR cc: @lokoyacap *very rigorous quant process by my team…. 100 years of data, cross sectional, deep learning … or just his pressers YTD: “roofing” INTC PLTR DELL AAPL TMO MU GE BA NVDA IBM (latest edition)

right monitor is 20 codex instances. left monitor has situational awareness on autoscroll. center monitor is my word doc mainfesto. two keyboards, one for both hands. left airpod is dwarkesh x eric jang, 3x speed. right airpod tchaikovsky. meta quest 3 overlays my HUD: heart rate, words per minute, blood caffeine content. one assistant hooks me to an iv of chinese peptides, cocktail. the other feeds me kimchi. my unitree robot steps in when my posture slouches. blue light beams down on me in my herman miller chair. efficiency. no wasted movement. no wasted thoughts. think you can keep up with me? good luck. this is just for my morning emails.

Trump just flatly stated he believes $IBM stock is "gonna go up a lot more." While the retail crowd treats this like a passing headline, the smart money recognizes the administration is signaling the ultimate national champion of the post-silicon era. This is not empty rhetoric. The Commerce Department recently finalized a massive two billion dollar quantum computing package where the federal government is taking direct equity stakes in critical domestic infrastructure. $IBM single-handedly locked down half of that entire allocation. The company is taking its one billion dollar federal award and matching it with one billion of its own cash to spin off Anderon, a standalone pure-play quantum wafer foundry in Albany. When the state takes a literal equity position on your cap table and the president publicly backs the stock, traditional valuation models become completely obsolete. $IBM is not a slow-moving legacy enterprise play anymore. It has been designated as a state-underwritten chokepoint that will control the foundational fabrication rules and standards for the next century of computing. With fifteen billion dollars in annual free cash flow, $IBM never needed government survival capital. This is a calculated geopolitical alliance to monopolize the quantum supply chain before foreign adversaries can even close the gap. The structural bottleneck has been chosen, and the institutional re-rating is just getting started. NFA, do your own DD

I spent the last 5 years asking 400 people a single question. It changed how I think about generating an edge in trading. The question: Name a president that you don't think other people will name, keeping in mind that other people are also trying to name a president that others won't name. Take a moment to answer before reading on. The results map directly onto how I think about alpha. Bucket the 45 presidents into 6 tiers by name recognition with Tier 1 being the most famous and Tier 6 the most obscure. The results show a clear bias toward obscurity—selections rise steadily from Tier 2 through Tier 6—with a sharp reverse-psychology spike back up at Tier 1. People favor going obscure ("Who would think of Millard Fillmore?") but many deploy maximum reverse psychology ("Washington is too obvious. Nobody will say him"). In reality, Washington and Fillmore are the two most common answers. [sorry, @dumborambo_, it is true] The sweet spot is Tier 2: Clinton, JFK, Jefferson, both Bushes, Nixon, LBJ. Household names, but not top-of-mindnames. Out of 407 respondents, only 20 picked from this group. That is 16% of presidents, under 5% of selections. Most people are genuinely shocked to learn these are the best answers. [Maybe except for @Rigatohni, who picked the last unnamed president (George W. Bush)]. This is how I approach every market I've traded. Sports betting, options, commodities, prediction markets. Tier 1 alpha ("Google has the best search engine," "Nvidia makes the chips") is real but saturated. The edge is gone before you arrive. Tier 6 alpha (tracking flight delays to beat sportsbooks, exotic statistical arbitrage) sounds clever but is almost always fruitless for the same reason: everyone who thinks they're being original is being original in the same direction. Tier 2 alpha is the sweet spot. It is never complicated and does not require a Harvard PhD, but it is also never obvious enough that the crowd has caught on. The smartest people I know have failed in trading because they couldn't believe Tier 2 alpha existed. Anything approachable, they assumed, had already been found by someone smarter. When I showed them edges that were statistically irrefutable, they called me a blind squirrel. I used to half-believe them. But the more edges I find, the more the pattern holds. The best alpha is Tier 2. Once you see it, you'll wonder why you didn't see it sooner.

for all you know, this could be your last summer. as homer wrote, you will never be lovelier than you are right now. so tell her she has a pretty smile. get lean or become a unit. find wonder in boarding the airplane. drink more coffee than you need. walk on the beach until your feet hurt. write your girlfriend a note on a tuesday. waste money because its not all about retirement. pull over at the lookout. ask the uber driver about his life. tell the woman in the elevator you like her shoes and mean it. call your mother. ask the waiter his favorite thing on the menu and order that. theres so much life in the slow moments, and you sprint past it chasing the next thing. but at any time, in any situation, you can find the good. smile. be happy. recognize it for a one of one life experience. what good is walking around a curmudgeon when tomorrow could be curtains. for heavens sake, flirt with the world.

This is gonna get so much worse It’s an industry led by historically unpopular people, loudly saying everyone will lose their jobs, during a plutocratic corrupt time

Ron DeSantis @RonDeSantis

2 months ago

Not surprising.

471 706 5K 662K 182

PTJ still got it

Eric Wallerstein @ericwallerstein

3 months ago

Paul Tudor Jones calling out JPY as the next major market move on his radar is music to my ears

18 51 515 140K 290

@TRAVmoneyofmine @prpl8 Until they use AGI to solve energy

@based16z Fun corollary I just thought of to this is say the coin is heads the first two flips. What % of your port would you bet on P(Heads)>50

@based16z @thedigipo Yea exactly if someone told you a coin is rigged but they didn’t know how and flipped it in the air, you are both indifferent between calling heads and tails. No info about how it is rigged makes it equivalent to a normal 50/50 coin in this case

@based16z @thedigipo If neither person has any info about the probability of heads, then doing just a normal flip would be “fair”. If P(heads) in the rigged coin is a random unknown variable, whoever chooses heads has an equal chance of on the disadvantaged side as being on the advantaged side

Only way to limit coming AI backlash is to start shifting taxes from labor to compute. The average voter needs to see salient benefits from AI. Today we tax labor > compute (income vs corp tax, depreciation for machines not education, payroll tax, etc). This will have to invert.

BREAKING: Airstrikes have just targeted Fajr Petrochemical Complex in Iran the first energy site to be targeted today.