JvanJ1983
122 posts
Greatest bitcoin explanation of all time! With our founder Wesley Schlemmer ⚡
YOU ARE PLAYING THE FINANCIAL GAME COMPLETELY WRONG. Imagine a father who secures $100,000 in spot Bitcoin. He holds the asset until the valuation hits a massive $5,000,000. Liquidating the position directly triggers devastating taxes on $4,900,000 of pure profit. So he executes the perfect institutional maneuver instead. He locks the Bitcoin in a legal trust, takes out a collateralized loan against the stack, and lives off the borrowed liquidity. Because he never executed a sale, his tax liability remains at absolute zero. Upon his death, the heirs receive the Bitcoin with a brand new cost basis set exactly at $5,000,000. The government cannot legally touch a single cent of the accumulated gain. This is exactly how generational wealth is permanently secured.
This is not getting near the attention it deserves. Remember when @_sholtodouglas and others went on @dwarkesh_sp seemingly years ago and said the ultimate AI game is automating ML research? That's what Karpathy just did here. This will accelerate progress again.
I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of ~630 lines of code, then: - the human iterates on the prompt (.md) - the AI agent iterates on the training code (.py) The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc. github.com/karpathy/autor… Part code, part sci-fi, and a pinch of psychosis :)
A few people away from 50k followers Upon reaching 50k, a random follower who retweets this post will receive a free lifetime subscription to my blog posts
I might just spend the next year laughing so hard that I'm in danger of rupturing my abdomen. Yes. The evidence that Donald Trump was a Federal informant in the Epstein case has been trickling out for a couple of years now. I am too libertarian to be MAGA and by no means Donald Trump's biggest fan, but I noticed this. What are the Trump-hating media and the Left (but I repeat myself) going to do now that these court documents are going public? They've backed themselves into a corner by incessantly screaming "Trump is a pedo!" despite the fact that he never went to Little St James and cut times with Epstein after Epstein hit on a 14-year-old Mar-A-Lago employee. Absolutely brilliant move by Trump not to exculpate himself by saying in public "I helped the law take Epstein down". Now his enemies are in so deep that all of their choices are horribly self-destructive. Look for a whole lot of journalists to have heart attacks and aneurysms as they grapple with the concept of publishing anything about this. They've organized their world-view and their public appeal around "Bad Orange Man is baaaad!" for so long that acknowledging he was one of the heroes of the Epstein story rather than a villain would break their brains and be reputational suicide. But the alternative - letting the truth leak out via social media while trying to smother it with non-coverage - would be in some ways even worse. The credibility of the journalistic establishment is already in the toilet. Contemplating how much worse it's going to get if they try to stonewall this is the part that makes me laugh so hard it hurts. What will happen if they actually backpedal might be even funnier to watch. Utter chaos, lefty/liberal brains melting by the millions, cats and dogs raining from the sky...indeed shall many things come to pass.
