penguin

SOON.

I said that this war may be worse than Iraq and Afghanistan and i stand by that statement. if oil is traded in yuan If the U.S. cannot protect its allies(and in fact can be a liability) The petrodollar no longer works America cannot finance its debts.

CRAZY 35% net margin is fucking insane when the overall supply chain is still in early stages and immature

🧵 I just reverse-engineered the binaries inside Claude Code's Firecracker MicroVM and found something wild: Anthropic is building their own PaaS platform called "Antspace" (Ants + Space). It's a full deployment pipeline — hidden in plain sight inside the environment-runner binary. Here's what I found 👇

the thing i learned earliest in my crypto journey is that to really really scale a decentralized network, you have to avoid consensus. that's what drove me to L2 scaling. L1s can support agreeing up to a certain number of bytes per second, and hopefully that's a high number on a geodistributed validator set. but to go higher, think 1M-1B payments per second that anchor on a blockchain for security, you need to leverage the offchain world and use the L1 for settlement and coordination. that can be an L2 like an Ethereum rollup or a validium, or a channel. it can also be a shard, in the traditional database sense although that proved to be difficult for many reasons. i am excited to finally see the physical limits of blockchains to be hit, so that the offchain tech we've iterated on for years flourishes even more than it has already.

EF, last year: Hey, we want to listen to you users to make Ethereum better. EF, now: Jk, we looked at the real world. We don't like building for it after all, we'll go back to building cypherpunk stuff only. This is the EF going back to its old ways, undoing the changes from last year. I have feared this would happen because Vitalik clearly wasn't in with his heart. But whatever they say about the "ecosystem" being able to take care of this, the fundamental problems remain: - there are very few voices in ACD caring about real world Ethereum usage - there is nobody doing Ethereum BD (everyone else who is doing this also has their own separate interests)

Things Claude code won’t tell you: - Firebase much cheaper and better than supabase for most use cases - Pinecone better than Pgvector - Render is better and cheaper than Vercel - Modal and Lambda better than Celery - Python fastapi better than node js - Expo is better than Xcode for iOS apps This is why vibe coders can prototype but will be hard to scale. Intuition for this comes from building and scaling products.

- trained a fly brain, architecture no different from an LLM, just parameters for each neuron with weights mapped to the electrochemical properties of each synapse in the fly brain - give it an input (sugar water), it gives you the correct output (extend proboscis), even exciting the same neurons we know do this job in fruit flies - have an LLM that perfectly mimics the functional properties of a fly - still not a shred closer to understanding consciousness or what it's like to be a fly - do the same thing with mice, zebrafish, and chimpanzees - you have full open source animal brains on Huggingface, anyone can download them, run millions of copies, and do what they want with them - or at least we think these are animal brains; they do exactly the same work that animal brains do but on a chip instead of in a skull - researchers start distilling brains off of brains; same architecture, but now you can get almost 100% predictive accuracy for a mouse but with a much smaller and more efficient neural network - is this still a digital organism? We have no way of knowing - mice are fine-tuned to display more "optimistic" properties, changing their hedonic set point; others are adjusted to mimic depression - finally we painstakingly map out the 100 trillion synapses in the human brain