The algorithm in this article helped build a $31 billion fortune.
It's called Baum-Welch, and this thread teaches it to detect market regimes. The "Baum" is Leonard Baum, who used the exact same hidden Markov math to help build Renaissance's Medallion fund: about 66% a year for 30 years, the best track record markets have ever seen.
It is the same algorithm that taught computers to recognize human speech. Pointed at price instead of sound.
The math is 60 years old and public. The moat was never the equation. It was the nerve to trust it.
The algorithm in this article helped build a $31 billion fortune.
It's called Baum-Welch, and this thread teaches it to detect market regimes. The "Baum" is Leonard Baum, who used the exact same hidden Markov math to help build Renaissance's Medallion fund: about 66% a year for 30 years, the best track record markets have ever seen.
It is the same algorithm that taught computers to recognize human speech. Pointed at price instead of sound.
The math is 60 years old and public. The moat was never the equation. It was the nerve to trust it.
This is a Rubik's Cube solver, built in minutes just by talking to Claude.
Then it kept going.
It broke on mobile, so he typed one lazy prompt, hit /yolo, and it fixed itself. Next he's adding webcam solving and 4x4, 5x5 cubes.
Shipping a real, working tool used to take a weekend. Now it takes a coffee. The barrier didn't shrink, it collapsed.
This is a Rubik's Cube solver, built in minutes just by talking to Claude.
Then it kept going.
It broke on mobile, so he typed one lazy prompt, hit /yolo, and it fixed itself. Next he's adding webcam solving and 4x4, 5x5 cubes.
Shipping a real, working tool used to take a weekend. Now it takes a coffee. The barrier didn't shrink, it collapsed.
This might be the most important visualization of AI-assisted development I've seen.
What you're watching isn't just a pretty graph.
It's the hidden structure of a production codebase:
• services
• APIs
• dependencies
• ownership
• relationships
Now imagine giving this understanding to an AI agent before it starts coding.
Suddenly AI isn't guessing.
It's reasoning.
That's the future we're building toward with Fable 5.
This might be the most important visualization of AI-assisted development I've seen.
What you're watching isn't just a pretty graph.
It's the hidden structure of a production codebase:
• services
• APIs
• dependencies
• ownership
• relationships
Now imagine giving this understanding to an AI agent before it starts coding.
Suddenly AI isn't guessing.
It's reasoning.
That's the future we're building toward with Fable 5.
A guy launched over 70 products that mostly flopped. Then one AI app hit $132,000 a month, and he runs it completely alone.
His name is Pieter Levels. No team. No investors. No office. He codes from a laptop, ships fast, and builds in public.
In 2023 he made PhotoAI: you upload your photos, it fine-tunes an image model on them, and generates studio-quality shots of you.
No photographer, no studio. It went from $0 to over $130,000 per month in about 18 months. His wider portfolio pulls in around $3 million a year. Zero employees.
He did not train a frontier model. He did not raise a cent. He took an AI capability that was already public, wrapped it in a product people actually wanted, and shipped it before anyone else bothered.
The lesson underneath it: the model was never the moat. Speed, taste, and actually shipping were. The tools are free. Almost nobody uses them.
A guy launched over 70 products that mostly flopped. Then one AI app hit $132,000 a month, and he runs it completely alone.
His name is Pieter Levels. No team. No investors. No office. He codes from a laptop, ships fast, and builds in public.
In 2023 he made PhotoAI: you upload your photos, it fine-tunes an image model on them, and generates studio-quality shots of you.
No photographer, no studio. It went from $0 to over $130,000 per month in about 18 months. His wider portfolio pulls in around $3 million a year. Zero employees.
He did not train a frontier model. He did not raise a cent. He took an AI capability that was already public, wrapped it in a product people actually wanted, and shipped it before anyone else bothered.
The lesson underneath it: the model was never the moat. Speed, taste, and actually shipping were. The tools are free. Almost nobody uses them.
ANTHROPIC's lead engineer allegedly leaked the Obsidian vault that powered their entire growth system — and was reportedly fired within the hour.
He was earning $1.9M a year to build and maintain it.
Inside are 8 massive knowledge clusters, arranged like a flower: Twitter Growth System, Distribution Growth MOC, LinkedIn Content Engine, YouTube, and more. Each cluster is a complete ecosystem of strategies, frameworks, and playbooks.
Zoom in, and thousands of interconnected nodes flood the screen until it almost turns white. Every audience strategy, content framework, growth experiment, and hypothesis is linked to everything else, creating a single connected knowledge network.
One section stands out: FACET 7 — Inner Ridge: People & Mental Models. A dense purple cluster with hundreds of connections branching downward like lightning, mapping how ideas, psychology, and decision-making influence growth.
In the corner sits GODMODE — ALL NODES: a single entry point connected to every node across the entire vault, giving access to the full knowledge graph from one place.
He was paid to build this system.
Now you're looking at it for free.
ANTHROPIC's lead engineer allegedly leaked the Obsidian vault that powered their entire growth system — and was reportedly fired within the hour.
He was earning $1.9M a year to build and maintain it.
Inside are 8 massive knowledge clusters, arranged like a flower: Twitter Growth System, Distribution Growth MOC, LinkedIn Content Engine, YouTube, and more. Each cluster is a complete ecosystem of strategies, frameworks, and playbooks.
Zoom in, and thousands of interconnected nodes flood the screen until it almost turns white. Every audience strategy, content framework, growth experiment, and hypothesis is linked to everything else, creating a single connected knowledge network.
One section stands out: FACET 7 — Inner Ridge: People & Mental Models. A dense purple cluster with hundreds of connections branching downward like lightning, mapping how ideas, psychology, and decision-making influence growth.
In the corner sits GODMODE — ALL NODES: a single entry point connected to every node across the entire vault, giving access to the full knowledge graph from one place.
He was paid to build this system.
Now you're looking at it for free.
Yuval Noah Harari, Oxford lecture on what AI agents really are:
• 0:57 – AI is an agent, not a tool
• 16:15 – AI as the native bureaucrat taking over
• 24:50 – finance, the easiest system for it to seize
• 30:35 – AI hacking the code of civilization
• 40:47 – the coming wave of AI immigration
One hour that reframes the AI debate better than 100 threads.
Watch it, then read the loop playbook below.
"AIs are hacking the code of human civilization." - Yuval Noah Harari
Yuval Noah Harari, Oxford lecture on what AI agents really are:
• 0:57 – AI is an agent, not a tool
• 16:15 – AI as the native bureaucrat taking over
• 24:50 – finance, the easiest system for it to seize
• 30:35 – AI hacking the code of civilization
• 40:47 – the coming wave of AI immigration
One hour that reframes the AI debate better than 100 threads.
Watch it, then read the loop playbook below.
"AIs are hacking the code of human civilization." - Yuval Noah Harari
Stephen Boyd, Stanford professor, on how quant funds build portfolios: "I am not aware of any quantitative hedge fund that doesn't simply use convex optimization to do this."
Citadel will pay a new quant around $400K - $500K to run that optimizer. The textbook that teaches it, written by Boyd himself, is free online.
The math was never the moat. Every fund reaches for the same equations from the same free books. The edge is the data, the speed, and the discipline to trust the model when everyone else flinches. Everyone has the textbook.
Almost nobody has the patience to trust it.