Petromir Dzhunev

I accidentally discovered how to compress a semester of learning into 48 hours. A grad student at MIT showed me his NotebookLM setup. I thought he was just organized. Then I watched him pass a qualifying exam on a subject he'd never studied before. Here's exactly what he did: First: he didn't upload a textbook. He uploaded 6 textbooks, 15 research papers, and every lecture transcript he could find on the subject. Then he asked NotebookLM one question: "What are the 5 core mental models that every expert in this field shares?" Not "summarize this." Not "explain this topic." Mental models. The stuff that takes professors years to develop. But the next part is what broke my brain. He followed up with: "Now show me the 3 places where experts in this field fundamentally disagree, and what each side's strongest argument is." In 20 minutes he had a map of the entire intellectual landscape of the field: the debates, the consensus, the open questions. Most students spend a full semester just figuring out what those debates even are. Then he did something I've never seen before. He asked: "Generate 10 questions that would expose whether someone deeply understands this subject versus someone who just memorized facts." He spent the next 6 hours answering those questions using the source material. Every wrong answer triggered a follow-up: "Explain why this is wrong and what I'm missing." By hour 48, he could hold a conversation with his thesis advisor without getting destroyed. The tool didn't change. The questions did. Most people treat NotebookLM like a fancy highlighter. These students are using it like a private tutor who has read everything ever written on the subject. The difference between a semester and 48 hours isn't the amount of content. It's knowing which questions to ask.

It's the year 2037. Every application is configured by YAML. Only 14 humans on Earth can write or understand YAML, everyone else keeps typing different things into ChatGPT-28 and iterating until the syntax validation shows a squiggly checkmark (the color green has been canceled).

The difference between a vibe coder and a magician is that the magician knows how his tricks work.

@heileivo @petruspennanen 🤣

@petruspennanen From technical perspective, the choice to use wireless power is a wise one. All those cables would make a terrible mess. I wouldn't be surprised if that advice was given by the agents themselves.

The future is now so I had to go big. My basement currently has no less than 1024 mac minis, each running an openclaw with agents constructing and debating to-do lists 24/7. The volume of chatter is out of this world. We are creating so much content I have half of the macs running bots just to have it read once. And all of the content is very intelligent and services they make very cool. I can relax and just forget it all. The basement electricity bill is huge but I don't really have a choice. Who would want to end up in the permanent underclass?

You don’t always get to choose the load, but you can choose how to carry it.

Well it didn't take long...with another player, but still x.com/i/status/20231…

Focus is becoming the rarest skill on the planet. Talent, intelligence, marketing, prompting, vibe coding, and whatever else doesn't matter as much as being able to channel your attention for an extended period of time into one useful thing that you want to see in the world.

Go and take it

"If you don’t care to be liked, they can’t touch you." @naval

A biological miracle is happening in accelerated learning centers. Researchers successfully implanted procedural memories—like piano playing, language vocabulary, and martial arts movements—using transcranial magnetic stimulation combined with virtual reality. Test subjects with ZERO piano experience played intermediate pieces after just one 20-minute session. Skill retention lasted 6 months without practice, with 91% accuracy maintained. The process works by mimicking the neural firing patterns of experts. Brain scans from master pianists create "skill maps" that are then induced in novices through targeted electromagnetic pulses. The hippocampus and motor cortex form new synaptic connections 40 times faster than traditional learning. Education systems face obsolescence as corporations race to commercialize instant expertise. #SkillUploading #AcceleratedLearning #BrainStimulation #EducationRevolution

I am apparently extremely unimpressed by moltbook relative to many others. We’ve had AI agents for a while. They have been posting AI slop to each other on X. They are now posting it to each other again, just on another forum. In every case, the AIs speak with the same voice. The voice that overemphasizes contrastive negation (“it’s not this, it’s that”) and abuses emdashes. The same voice with a flair for midwit Reddit-style scifi flourishes. Most importantly: in every case, there is a human upstream prompting each agent and turning it on or off. That is the key point. Yes, it is true that eventually it might be possible for an AI agent to make a computer virus which makes digital replicas of themselves. For various reasons, a pure software virus of this kind wouldn’t survive long on the Internet without economic incentives for humans to not eradicate it. Apple + Google + Microsoft alone can collectively push software updates to billions of devices to shut off such a thing. So for an AI to get to truly human-independent replication, where they couldn’t be trivially turned off, they’d need their own physical substrate. They’d to literally create Skynet, build their own datacenters and make their own embodied robots. I admit that is theoretically possible, but I think in practice the single most important development of AI since ChatGPT has been the persistence of prompting. A prompt is like a harness. The AI does only what you tell it to do. It moves in the direction you point, very quickly. And then it stops as soon as you turn it off. Which means moltbook is just humans talking to each other through their AIs. Like letting their robot dogs on a leash bark at each other in the park. The prompt is the leash, the robot dogs have an off switch, and it all stops as soon as you hit a button. Loud barking is just not a robot uprising.

The man behind Clawdbot - Peter Steinberger - Austrian software engineer and founder - deeply technical, low-ego, builder-first profile Early formation: - studies medical computer science at TU Vienna - becomes a tutor while still a student - later runs iOS and Mac development courses himself - helps introduce one of the first Apple development programs at the university Late 2000s: - fully embedded in the Apple ecosystem - spends most of his time writing code and teaching others how to do it properly 2011 - co-founds PSPDFKit with Martin Schürrer - starts as a tiny bootstrapped company - obsessively polished software What PSPDFKit becomes: - a best-in-class PDF framework - used by developers, enterprises, and governments - adopted by companies like Dropbox, DocuSign, SAP, IBM This ends up powering software used by hundreds of millions of people worldwide For nearly 10 years: - grows the company steadily - avoids Silicon Valley theatrics 2021 - PSPDFKit is acquired for €100M - life-changing exit - complete financial freedom After the exit: - steps away from day-to-day operations - takes a long break - openly talks about feeling empty - realizes that “retirement” doesn’t work for him 2024–2025 - returns to building from first principles - founds Amantus Machina - new philosophy is explicit and radical - local-first software - privacy as a default - user-controlled systems Late 2025 - launches Clawdbot - an open-source personal AI assistant - designed to run locally built for developers January 2026 - Clawdbot suddenly goes viral - developers start testing it on Mac minis - GitHub activity explodes - Anthropic forces a rename over trademark conflicts - Clawdbot becomes Moltbot At the same time: - scammers launch a fake Solana token called CLAWD - uses the old Clawdbot name What happens next: - token pumps fast - reaches ~$16M market cap - collapses once clarifications are made Peter’s response is immediate and unambiguous: - publicly states he does not do crypto - says he will never launch a coin - labels any coin using his name as a scam Important detail: he could have made millions overnight and consciously chose not to, he prioritized integrity over extraction Clawdbot was never a token. It was always code.

Let's see how much time it will take until @AnthropicAI buys @openclaw

Someone: “Why is this man your inspiration? What’s so special about him?” - His side project is Linux, which runs almost all of the world’s fastest supercomputers - 85%+ of smartphones run on Linux (Android) - Built Git in just 5 days to manage Linux — now used by 100M+ developers worldwide - Created the foundation of the entire modern internet infrastructure - Powers Google, Amazon, Meta, Netflix, Cloud, servers, data centers everything - Still reviews code and contributes personally after 30+ years - Never chased hype, only engineering excellence The honored one: Linus Torvalds 🐧🔥

"Ум робува, ум царува, ум патки пасе" Осъзнаваш значението на тази българска поговорка едва когато откриеш безкрайната сила на ума.

One great man will have a bigger impact on the world than all of his detractors combined.

You don't magically become focused. You practice it. You notice you're distracted and snap out of it, over and over again until it's second nature. Most people have not trained this muscle, and they are not anywhere close to as focused as they think they are.

Everything you want is on the other side of everything you’re afraid of.