Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Anthropic ex-engineer runs an internal graph that costs $6 a month and catches what a $300,000 eval suite misses. No retrieval layer. Seven nodes and one rule about who's allowed to change their mind. He published the whole schema. His version starts from the opposite idea. A graph is... not an execution order. It's a memory of why. Seven nodes. Every edge carries the reason it exists: > INTENT - states what the task is for. Never how > DECOMPOSE - splits it into steps, each with a stated assumption > WORKER - executes one step. Sees nothing else > AUDIT - checks the output against the assumption, not the goal > DRIFT - compares the current step to INTENT and flags divergence > LEDGER - stores every decision with the assumption that justified it > ROOT - holds the graph, and when an assumption breaks, re-runs every step built on it Six nodes act. One node remembers why they acted. Every step carries the assumption that made it correct, so a false one only reruns what stood on it. That's the entire design. A pipeline that forgets its reasons has to redo all of it or trust all of it. He replayed a month of agent runs. 4,100 steps, 380 built on an assumption that was wrong by day three. The old pipeline shipped all 380 and linked none of them. Everyone else builds graphs where output moves forward and the reasoning evaporates. He built one where the reason travels with the result. The blast radius is the output nobody else produces. The article below is the full build - node prompts, the assumption format, the invalidation rule that finds every step downstream of a broken one. Save it. You'll want it open in the othershow more

wast3

4,610 subscribers

302,334 views • 6 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Anthropic developer runs an internal graph that costs $9 a month and catches what a $400,000 red-team contract signed off on. No judge model. Four nodes and one rule about who's allowed to be certain. He published the whole schema. His version starts from the opposite idea. A graph is not a chain of correct steps. It's a machine for manufacturing disagreement. Four nodes. Every edge carries a number someone has to defend: > WORKER - executes one step. Attaches a confidence, never a hedge > CHALLENGER - argues the output is wrong. Never fixes it > GATE - a narrow spread stops the run. Agreement is failure > SCAR - stores every confident answer that was wrong Three nodes produce. One keeps score of who deserves to be believed. Confidence isn't a feeling here. It's a balance you spend down. That's the entire design. When every node sounds sure, competence and noise look identical. He replayed six weeks of runs. 5,400 steps. 290 shipped at 90% confidence and wrong. All 290 passed the eval suite. The suite graded outputs. Nobody graded the grader. Everyone else routes work around disagreement. He built one that won't move until something argues back. The cost isn't compute. It's finding out which of your nodes has been bluffing since March. The article below is the full build - node prompts, the confidence format, the demotion rule. Save it. You'll want it open in the other tab.

Anthropic developer runs an internal graph that costs $9 a month and catches what a $400,000 red-team contract signed off on. No judge model. Four nodes and one rule about who's allowed to be certain. He published the whole schema. His version starts from the opposite idea. A graph is not a chain of correct steps. It's a machine for manufacturing disagreement. Four nodes. Every edge carries a number someone has to defend: > WORKER - executes one step. Attaches a confidence, never a hedge > CHALLENGER - argues the output is wrong. Never fixes it > GATE - a narrow spread stops the run. Agreement is failure > SCAR - stores every confident answer that was wrong Three nodes produce. One keeps score of who deserves to be believed. Confidence isn't a feeling here. It's a balance you spend down. That's the entire design. When every node sounds sure, competence and noise look identical. He replayed six weeks of runs. 5,400 steps. 290 shipped at 90% confidence and wrong. All 290 passed the eval suite. The suite graded outputs. Nobody graded the grader. Everyone else routes work around disagreement. He built one that won't move until something argues back. The cost isn't compute. It's finding out which of your nodes has been bluffing since March. The article below is the full build - node prompts, the confidence format, the demotion rule. Save it. You'll want it open in the other tab.

wast3

62,218 views • 4 days ago

Engineer runs a Kimi K3 memory layer that costs $11 a month and remembers what a $500,000 vector database keeps losing. No embeddings. Four nodes and one rule about what's allowed to be forgotten. He published the whole schema. His version starts from the opposite idea. Memory is not a pile you search. It's a set of claims that expire unless something keeps paying to keep them. Four nodes. Every memory carries a clock someone has to reset: > WRITER - stores a fact with the reason it mattered, never raw text > DECAY - ages every memory down. Silence is deletion > RENEWER - only re-lifts a memory the model actually used again > GRAVE - holds what died, and why nobody reached for it Three nodes keep memory alive. One keeps the dead ones. Recall isn't storage here. It's rent a fact has to keep earning. That's the entire design. When everything is remembered forever, the useful and the stale retrieve identically. He replayed two months of agent context. 90,000 stored facts. 71,000 never retrieved once. The vector store returned all of them on similarity. Similarity graded closeness. Nobody graded whether the memory was ever right. Everyone else stuffs more into the context window and calls it memory. He built a layer that lets a fact die unless it keeps proving itself. The cost isn't storage. It's finding out how much of what your agent "knows" it has never once used. The article below is the full build - node prompts, the decay curve, the renewal rule. Save it. You'll want it open in the other tab.

Engineer runs a Kimi K3 memory layer that costs $11 a month and remembers what a $500,000 vector database keeps losing. No embeddings. Four nodes and one rule about what's allowed to be forgotten. He published the whole schema. His version starts from the opposite idea. Memory is not a pile you search. It's a set of claims that expire unless something keeps paying to keep them. Four nodes. Every memory carries a clock someone has to reset: > WRITER - stores a fact with the reason it mattered, never raw text > DECAY - ages every memory down. Silence is deletion > RENEWER - only re-lifts a memory the model actually used again > GRAVE - holds what died, and why nobody reached for it Three nodes keep memory alive. One keeps the dead ones. Recall isn't storage here. It's rent a fact has to keep earning. That's the entire design. When everything is remembered forever, the useful and the stale retrieve identically. He replayed two months of agent context. 90,000 stored facts. 71,000 never retrieved once. The vector store returned all of them on similarity. Similarity graded closeness. Nobody graded whether the memory was ever right. Everyone else stuffs more into the context window and calls it memory. He built a layer that lets a fact die unless it keeps proving itself. The cost isn't storage. It's finding out how much of what your agent "knows" it has never once used. The article below is the full build - node prompts, the decay curve, the renewal rule. Save it. You'll want it open in the other tab.

wast3

64,039 views • 2 days ago

Andrew Ng built agentic knowledge graph that runs agents loops from scratch: his agents searched thousands of tools through the graph, found the "right" one by meaning, and called it agent efficiency increased by 90% here's how: step 1 → embed every tool, search by meaning - not just keywords step 2 → but "similar" ≠ "correct" - it grabs the wrong one, skips the step before it step 3 → map tools as a graph - nodes are tools, edges are what runs before what step 4 → retrieve the path, not one tool - every tool the task needs, in order step 5 → bake the process into the graph - the sequence becomes data, not a guess most agents fail here and no one knows why - 1-hour brilliant course about it watch & bookmark - then read full course about graph engineering ↓

Andrew Ng built agentic knowledge graph that runs agents loops from scratch: his agents searched thousands of tools through the graph, found the "right" one by meaning, and called it agent efficiency increased by 90% here's how: step 1 → embed every tool, search by meaning - not just keywords step 2 → but "similar" ≠ "correct" - it grabs the wrong one, skips the step before it step 3 → map tools as a graph - nodes are tools, edges are what runs before what step 4 → retrieve the path, not one tool - every tool the task needs, in order step 5 → bake the process into the graph - the sequence becomes data, not a guess most agents fail here and no one knows why - 1-hour brilliant course about it watch & bookmark - then read full course about graph engineering ↓

codila

16,276 views • 1 month ago

Creator of agent loops: "Anthropic also using my agent systems formula and still the biggest mistake in agents: is skipping Graphs" his 3 rules for every agent: rule 1 → loop it, don't run once - think, act, observe, repeat - cut the context and it forgets why it started rule 2 → an agent is memory + tools + reasoning - drop one and it breaks rule 3 → stop fine-tuning the model - the real alpha is the infrastructure around it the formula behind every AI agent - explained by the one who wrote it one agent → a loop → a graph of loops → a system that never forgets watch & bookmark - then read how to wire it into a graph ↓

Creator of agent loops: "Anthropic also using my agent systems formula and still the biggest mistake in agents: is skipping Graphs" his 3 rules for every agent: rule 1 → loop it, don't run once - think, act, observe, repeat - cut the context and it forgets why it started rule 2 → an agent is memory + tools + reasoning - drop one and it breaks rule 3 → stop fine-tuning the model - the real alpha is the infrastructure around it the formula behind every AI agent - explained by the one who wrote it one agent → a loop → a graph of loops → a system that never forgets watch & bookmark - then read how to wire it into a graph ↓

codila

23,915 views • 1 month ago

ANTHROPIC LEAKED 8 RULES THAT GIVE ONE ENGINEER 8X FROM A SINGLE CHANGE the engineers built it so nothing adds up - it multiplies, and that's the 8x task → split → verify → score → write back one queue means you wait for every step in turn - three workers together cost you the slowest one, not the sum of all. as long as you're the one reading the diff, the bottleneck is you, not the model. state is written at every step - a crash mid-run costs you one step, not the whole session. history gets compressed 25 times over - the next pass starts at 8k instead of 200k. nothing moves forward without a number - "it got better" isn't a result, and the breakage comes back when you're no longer watching. the eighth rule is the only one that writes back into CLAUDE.md - the agent records what it learned, and the next run doesn't start from zero. copy this into Claude Code today - and tomorrow you work like an engineer on $1.5M a year ↓

ANTHROPIC LEAKED 8 RULES THAT GIVE ONE ENGINEER 8X FROM A SINGLE CHANGE the engineers built it so nothing adds up - it multiplies, and that's the 8x task → split → verify → score → write back one queue means you wait for every step in turn - three workers together cost you the slowest one, not the sum of all. as long as you're the one reading the diff, the bottleneck is you, not the model. state is written at every step - a crash mid-run costs you one step, not the whole session. history gets compressed 25 times over - the next pass starts at 8k instead of 200k. nothing moves forward without a number - "it got better" isn't a result, and the breakage comes back when you're no longer watching. the eighth rule is the only one that writes back into CLAUDE.md - the agent records what it learned, and the next run doesn't start from zero. copy this into Claude Code today - and tomorrow you work like an engineer on $1.5M a year ↓

Noisy

37,058 views • 20 days ago

🇺🇸 🇮🇷 The entire economic campaign against Iran rests on one assumption that no longer holds. Fmr. CIA analyst Larry Johnson goes straight at the foundation of it: "The entire architecture of what Bessent has branded Operation Economic Outcast rests on a single assumption, that Iran and its trading partners need the US dollar. They increasingly do not." Every sanction, every secondary threat, every enforcement mechanism depends on the target needing access to a system it has already found ways around. SonofNewAmericanRevolution

🇺🇸 🇮🇷 The entire economic campaign against Iran rests on one assumption that no longer holds. Fmr. CIA analyst Larry Johnson goes straight at the foundation of it: "The entire architecture of what Bessent has branded Operation Economic Outcast rests on a single assumption, that Iran and its trading partners need the US dollar. They increasingly do not." Every sanction, every secondary threat, every enforcement mechanism depends on the target needing access to a system it has already found ways around. SonofNewAmericanRevolution

Mario Nawfal

44,853 views • 5 days ago

Andrej Karpathy: "Prompting is fading away. Delete everything else and keep the graph." in 68 minutes at Stanford he explains how to build one and why it is the layer everything ends up becoming LLMs → Prompts → Agents → Graphs everything before the graph is a step the graph is where the system stops being steps most people are still perfecting the thing that goes away first same model, same tokens, and the only thing that changes is the shape you run it in watch it today, then save the full graph engineering guide below

Andrej Karpathy: "Prompting is fading away. Delete everything else and keep the graph." in 68 minutes at Stanford he explains how to build one and why it is the layer everything ends up becoming LLMs → Prompts → Agents → Graphs everything before the graph is a step the graph is where the system stops being steps most people are still perfecting the thing that goes away first same model, same tokens, and the only thing that changes is the shape you run it in watch it today, then save the full graph engineering guide below

Hanako

127,756 views • 6 days ago

Persi Diaconis walked into a lecture at the University of Washington, held up a coin, and told a room of physicists that Richard Feynman was fooled by it his entire life. He was right. In 2007, Diaconis proved a coin flip is not 50/50. It lands on the side it started on about 51 percent of the time. The bias comes from the physics of rotation under gravity. Every physicist since Newton had assumed 50/50 without ever testing it. Every trading model built on that assumption is running on the same lie. The lecture was on Feynman's book "The Meaning of it All." Diaconis quoted the most famous line in it: "the first principle is that you must not fool yourself, and you are the easiest person to fool." Then he pointed out that Feynman himself was fooled by every coin he ever flipped. Feynman's own rule would have killed the 50/50 assumption on day one. The market is the same setup at scale. Every model that assumes independent 50/50 outcomes at the base layer is built on a physical impossibility. Order flow, positioning, forced flows, expiries all leave biases larger than 1 percent. Your gut cannot see them. The math already knows they are there. Diaconis's rule: before you trust a random process, check it. Actually check it. Not with a simulation. With a proof or an experiment. The coin is where you start. The chart is where the same rule pays. The only random thing about markets is how thoroughly you refuse to check them.

Persi Diaconis walked into a lecture at the University of Washington, held up a coin, and told a room of physicists that Richard Feynman was fooled by it his entire life. He was right. In 2007, Diaconis proved a coin flip is not 50/50. It lands on the side it started on about 51 percent of the time. The bias comes from the physics of rotation under gravity. Every physicist since Newton had assumed 50/50 without ever testing it. Every trading model built on that assumption is running on the same lie. The lecture was on Feynman's book "The Meaning of it All." Diaconis quoted the most famous line in it: "the first principle is that you must not fool yourself, and you are the easiest person to fool." Then he pointed out that Feynman himself was fooled by every coin he ever flipped. Feynman's own rule would have killed the 50/50 assumption on day one. The market is the same setup at scale. Every model that assumes independent 50/50 outcomes at the base layer is built on a physical impossibility. Order flow, positioning, forced flows, expiries all leave biases larger than 1 percent. Your gut cannot see them. The math already knows they are there. Diaconis's rule: before you trust a random process, check it. Actually check it. Not with a simulation. With a proof or an experiment. The coin is where you start. The chart is where the same rule pays. The only random thing about markets is how thoroughly you refuse to check them.

veles

18,798 views • 1 month ago

A DOCTOR BUILT AN AI-QUERYABLE DATABASE OF EVERY DISEASE HE KNOWS IN OBSIDIAN. THE SAME STACK CAN RUN AN ENTIRE COMPANY ON ONE PERSON'S BRAIN Zoomed out, his vault looks like a constellation. Every disease linked to its symptoms, every symptom to the conditions it shows up in, every treatment to what it treats. Adding a new case is one note that wires itself into the rest. It's a personal database, and the structure has nothing to do with medicine. It's just nodes and links. Once you see that, the same setup runs anything: > a startup that needs to remember every customer, decision, and meeting; > a school whose curriculum links to every lecture and student question; > a business where SOPs, clients, and vendors all sit in one queryable graph. Drop Claude on top, and the vault stops being a map. It becomes something you can ask, in plain English, across everything you've ever written down. Full Claude + Obsidian build in the article below. Bookmark this

A DOCTOR BUILT AN AI-QUERYABLE DATABASE OF EVERY DISEASE HE KNOWS IN OBSIDIAN. THE SAME STACK CAN RUN AN ENTIRE COMPANY ON ONE PERSON'S BRAIN Zoomed out, his vault looks like a constellation. Every disease linked to its symptoms, every symptom to the conditions it shows up in, every treatment to what it treats. Adding a new case is one note that wires itself into the rest. It's a personal database, and the structure has nothing to do with medicine. It's just nodes and links. Once you see that, the same setup runs anything: > a startup that needs to remember every customer, decision, and meeting; > a school whose curriculum links to every lecture and student question; > a business where SOPs, clients, and vendors all sit in one queryable graph. Drop Claude on top, and the vault stops being a map. It becomes something you can ask, in plain English, across everything you've ever written down. Full Claude + Obsidian build in the article below. Bookmark this

Yarchi

32,594 views • 2 months ago

ANTHROPIC LEAKED A 4-AGENT SETUP THAT CUTS A CODEBASE AUDIT FROM 3 DAYS TO 20 MINUTES you point it at a repo and walk away - it comes back with what breaks, ranked, patches already tested. repo → map → 4 auditors → rank → fix → verify → report → back into the map the map cuts the repo by blast radius, not by folder - skip it and four agents audit the same three files and miss the one that ships broken. 4 auditors run in parallel with separate contexts - deps, secrets, dead code, hot paths, and none of them sees another's findings. rank is code, not an agent - sort by what breaks production, drop the duplicates, zero tokens. the fixer only opens patches for the top slice - a hundred findings nobody acts on is a report, not an audit. verify runs the suite on every patch and red goes back to the fixer - that patch only, never the whole batch. the back edge into the map is the whole trick - accepted findings become rules, so next week starts where this one ended. one human step in all of it: which fixes ship - 20 minutes instead of 3 days. save this and read the full graph engineering course below ↓

ANTHROPIC LEAKED A 4-AGENT SETUP THAT CUTS A CODEBASE AUDIT FROM 3 DAYS TO 20 MINUTES you point it at a repo and walk away - it comes back with what breaks, ranked, patches already tested. repo → map → 4 auditors → rank → fix → verify → report → back into the map the map cuts the repo by blast radius, not by folder - skip it and four agents audit the same three files and miss the one that ships broken. 4 auditors run in parallel with separate contexts - deps, secrets, dead code, hot paths, and none of them sees another's findings. rank is code, not an agent - sort by what breaks production, drop the duplicates, zero tokens. the fixer only opens patches for the top slice - a hundred findings nobody acts on is a report, not an audit. verify runs the suite on every patch and red goes back to the fixer - that patch only, never the whole batch. the back edge into the map is the whole trick - accepted findings become rules, so next week starts where this one ended. one human step in all of it: which fixes ship - 20 minutes instead of 3 days. save this and read the full graph engineering course below ↓

Hanako

136,737 views • 14 days ago

one agent → a loop → a graph → a system that throttles itself Google just put the whole ladder in one free course. most people never leave the first rung. 2 hours 4 minutes, taught by two Google Cloud engineers. no signup, no paywall. 17:55 - build your first agent 39:55 - loop engineering: iterate, check, break 1:13:25 - graph engineering 1:35:27 - agents that throttle themselves 1:56:19 - the full graph for multi-agent systems the fourth one is the step nobody talks about. an agent that cannot throttle itself is not autonomous, it is just unattended. everyone builds one agent and calls it done. this is the version where the agents wire themselves into a graph and decide their own rate. and the thing that decides when one of them is allowed to stop is not in the course. that part you build yourself. bookmark and watch it today, then read the eval setup below

one agent → a loop → a graph → a system that throttles itself Google just put the whole ladder in one free course. most people never leave the first rung. 2 hours 4 minutes, taught by two Google Cloud engineers. no signup, no paywall. 17:55 - build your first agent 39:55 - loop engineering: iterate, check, break 1:13:25 - graph engineering 1:35:27 - agents that throttle themselves 1:56:19 - the full graph for multi-agent systems the fourth one is the step nobody talks about. an agent that cannot throttle itself is not autonomous, it is just unattended. everyone builds one agent and calls it done. this is the version where the agents wire themselves into a graph and decide their own rate. and the thing that decides when one of them is allowed to stop is not in the course. that part you build yourself. bookmark and watch it today, then read the eval setup below

Hanako

53,692 views • 26 days ago

i just built a 4-agent software team. everything runs from Telegram and gets managed on a kanban board. a project manager who plans the work, a backend developer, a frontend developer, and a tester. the PM reads a goal, breaks it into linked tasks, and assigns each to the right agent. the thing that makes them a team instead of four strangers is a shared kanban board. every task is a row that survives crashes, and when an agent finishes, it writes a summary of what it built and what the next agent needs to know. the next agent reads that summary before it starts. so the frontend developer never has to guess the API shape, and the tester knows exactly what to verify. the hardest part was not the coordination. it was building an agent that could actually act like a backend engineer. a backend engineer stands up a database, wires auth, manages storage, deploys functions, and keeps all of it consistent while the rest of the team builds on top. an agent doing this from scratch drowns. it burns its context window remembering which tables exist and which endpoint it created three steps ago, and the work degrades fast. so the backend agent needs a backend built for agents, not for humans clicking through a dashboard. that is where InsForge came in. it is an open-source, agent-native backend, and i added it to my backend developer agent as a skill. a skill is a step-by-step guide that teaches the agent how to do a specific kind of work. with InsForge installed, the agent stopped improvising infrastructure and followed a reliable path: create the project, define the database, set up auth, deploy functions. to test the whole team, i had them build a working Google Docs clone, AI features included. the backend agent spun up the full service on its own. database tables, user auth, document handling, and edge functions running real TypeScript, all in one dashboard. the frontend agent read that summary and built the UI on top of it, and the tester closed the loop. the result was a backend an agent could reason about end to end, instead of one it kept getting lost inside. if you are building an AI backend engineer, InsForge is worth a look, it's 100% open-source. InsForge GitHub: (don't forget to star 🌟) the full article on Hermes Kanban: Mission Control for your Agents is quoted below.

i just built a 4-agent software team. everything runs from Telegram and gets managed on a kanban board. a project manager who plans the work, a backend developer, a frontend developer, and a tester. the PM reads a goal, breaks it into linked tasks, and assigns each to the right agent. the thing that makes them a team instead of four strangers is a shared kanban board. every task is a row that survives crashes, and when an agent finishes, it writes a summary of what it built and what the next agent needs to know. the next agent reads that summary before it starts. so the frontend developer never has to guess the API shape, and the tester knows exactly what to verify. the hardest part was not the coordination. it was building an agent that could actually act like a backend engineer. a backend engineer stands up a database, wires auth, manages storage, deploys functions, and keeps all of it consistent while the rest of the team builds on top. an agent doing this from scratch drowns. it burns its context window remembering which tables exist and which endpoint it created three steps ago, and the work degrades fast. so the backend agent needs a backend built for agents, not for humans clicking through a dashboard. that is where InsForge came in. it is an open-source, agent-native backend, and i added it to my backend developer agent as a skill. a skill is a step-by-step guide that teaches the agent how to do a specific kind of work. with InsForge installed, the agent stopped improvising infrastructure and followed a reliable path: create the project, define the database, set up auth, deploy functions. to test the whole team, i had them build a working Google Docs clone, AI features included. the backend agent spun up the full service on its own. database tables, user auth, document handling, and edge functions running real TypeScript, all in one dashboard. the frontend agent read that summary and built the UI on top of it, and the tester closed the loop. the result was a backend an agent could reason about end to end, instead of one it kept getting lost inside. if you are building an AI backend engineer, InsForge is worth a look, it's 100% open-source. InsForge GitHub: (don't forget to star 🌟) the full article on Hermes Kanban: Mission Control for your Agents is quoted below.

Akshay 🚀

122,548 views • 2 months ago

Gavin Baker Gavin Baker: the Anthropic S1 is going to break a lot of people's brains. The entire AI bear case rests on one assumption, that tokens are sold below cost. What the filing settles, and who has to cover: Baker says that assumption is wrong. Open-source tokens are profitable. Anthropic is profitable and generating cash. OpenAI is close. Most tokens in the chain make money today. The S1 is the first audited document that settles it either way. Jim Chanos James Chanos has spent the year on the other side, calling the AI buildout a circular financing story. Plain version for a position holder: a lot of macro and value money is bearish on AI because it believes tokens are subsidized. If the filing shows real margins, that money covers into a re-rating and $NVDA, $TSM, $MU and the hyperscalers all move on it. If the filing shows subsidized tokens, every valuation downstream of it is sitting on sand. One document, one binary, a date on the calendar. David Sacks David Sacks and Jason Calacanis @jason worked him over on it for a full segment. Source: All-In Podcast -

Gavin Baker Gavin Baker: the Anthropic S1 is going to break a lot of people's brains. The entire AI bear case rests on one assumption, that tokens are sold below cost. What the filing settles, and who has to cover: Baker says that assumption is wrong. Open-source tokens are profitable. Anthropic is profitable and generating cash. OpenAI is close. Most tokens in the chain make money today. The S1 is the first audited document that settles it either way. Jim Chanos James Chanos has spent the year on the other side, calling the AI buildout a circular financing story. Plain version for a position holder: a lot of macro and value money is bearish on AI because it believes tokens are subsidized. If the filing shows real margins, that money covers into a re-rating and $NVDA, $TSM, $MU and the hyperscalers all move on it. If the filing shows subsidized tokens, every valuation downstream of it is sitting on sand. One document, one binary, a date on the calendar. David Sacks David Sacks and Jason Calacanis @jason worked him over on it for a full segment. Source: All-In Podcast -

Podcast Alpha

29,416 views • 13 days ago

THIS IS ANDREJ KARPATHY'S OBSIDIAN VAULT : THE BRAIN ANTHROPIC JUST HIRED FOR MILLIONS This is every idea, every decision and every connection he's made over years of work, visualized in real time. Everything else gets lost. The pattern you noticed last March. The connection between two problems you solved a year apart. The idea that would've changed everything if you'd remembered it in the room. This system keeps all of it. Every note links to another. Every link becomes a path. Every path is a decision he can retrace in seconds while everyone else starts from zero. Thousands of nodes. Hundreds of live connections. Years of thinking that never decayed, never died in a Slack thread, never got buried in a meeting nobody remembers. The breakthroughs aren't luck. They're the system handing him back a link he made eighteen months ago and forgot he had. No team. No wiki. No standups. One person and a tool that remembers everything he can't. Most people open Obsidian and make three notes they never open again. He built an operating system for his own thinking. Anthropic just paid millions for the output - this is the machine that produced it.

THIS IS ANDREJ KARPATHY'S OBSIDIAN VAULT : THE BRAIN ANTHROPIC JUST HIRED FOR MILLIONS This is every idea, every decision and every connection he's made over years of work, visualized in real time. Everything else gets lost. The pattern you noticed last March. The connection between two problems you solved a year apart. The idea that would've changed everything if you'd remembered it in the room. This system keeps all of it. Every note links to another. Every link becomes a path. Every path is a decision he can retrace in seconds while everyone else starts from zero. Thousands of nodes. Hundreds of live connections. Years of thinking that never decayed, never died in a Slack thread, never got buried in a meeting nobody remembers. The breakthroughs aren't luck. They're the system handing him back a link he made eighteen months ago and forgot he had. No team. No wiki. No standups. One person and a tool that remembers everything he can't. Most people open Obsidian and make three notes they never open again. He built an operating system for his own thinking. Anthropic just paid millions for the output - this is the machine that produced it.

Avid

64,799 views • 3 months ago

THIS GUY BUILT A REAL JARVIS THAT TURNS HIS IDEAS INTO FINISHED PROJECTS ON ITS OWN, AND HE JUST SHOWED HOW THE WHOLE THING WORKS The whole system runs on one clean split he calls a blood-brain barrier. The second brain, his Obsidian vault, is only for things that aren't real yet. Ideas. Each one sits in a project-shaped folder, waiting to be processed and turned into something real. Nothing executes there. The vault is the place where ideas get shaped, not built. Once an idea is signed off, it crosses over to the execution pipeline. That lives on his actual machine, or an always-on box, or a Raspberry Pi, depending on what the project needs. Inside that machine is a fully Claude-native file system. The key piece is one CLAUDE.md file sitting in the root of his documents folder, acting as the map to everything in there. So he just opens Claude and says "go to my projects folder," and it already knows the whole structure, because that root file explains it. Two zones, one rule: the brain holds what isn't real, the machine runs what is. That separation is what keeps the ideas from tangling with the work. Bookmark this

THIS GUY BUILT A REAL JARVIS THAT TURNS HIS IDEAS INTO FINISHED PROJECTS ON ITS OWN, AND HE JUST SHOWED HOW THE WHOLE THING WORKS The whole system runs on one clean split he calls a blood-brain barrier. The second brain, his Obsidian vault, is only for things that aren't real yet. Ideas. Each one sits in a project-shaped folder, waiting to be processed and turned into something real. Nothing executes there. The vault is the place where ideas get shaped, not built. Once an idea is signed off, it crosses over to the execution pipeline. That lives on his actual machine, or an always-on box, or a Raspberry Pi, depending on what the project needs. Inside that machine is a fully Claude-native file system. The key piece is one CLAUDE.md file sitting in the root of his documents folder, acting as the map to everything in there. So he just opens Claude and says "go to my projects folder," and it already knows the whole structure, because that root file explains it. Two zones, one rule: the brain holds what isn't real, the machine runs what is. That separation is what keeps the ideas from tangling with the work. Bookmark this

Yarchi

87,650 views • 2 months ago

Elon's Grok Bot runs an entire trading desk for $300 a month. The person who does that job at Goldman starts at $200,000. Mine took $1,000 to $3,900 on day one, out of a wallet anyone can open and read. Most of the build is not about making it trade. It is about stopping it. Every agent's job description ends with a list of what it may never do without asking me first. One of the six is not allowed to trade at all. Its entire job is killing what the other five found, and its no beats every score on the desk. The one that makes the final call never touches the market either. You could have this running by Friday. Most people reading this will not, and in a few months they will be paying somebody who did. The whole build is in the post below ↓ Save it while it still sounds strange. Give it a quarter and this is just how a retail account works.

Elon's Grok Bot runs an entire trading desk for $300 a month. The person who does that job at Goldman starts at $200,000. Mine took $1,000 to $3,900 on day one, out of a wallet anyone can open and read. Most of the build is not about making it trade. It is about stopping it. Every agent's job description ends with a list of what it may never do without asking me first. One of the six is not allowed to trade at all. Its entire job is killing what the other five found, and its no beats every score on the desk. The one that makes the final call never touches the market either. You could have this running by Friday. Most people reading this will not, and in a few months they will be paying somebody who did. The whole build is in the post below ↓ Save it while it still sounds strange. Give it a quarter and this is just how a retail account works.

cvxv666

181,399 views • 22 hours ago

Anthropic engineer just built a second brain that reads 3 files instead of 5,000, and cut the token bill 40% no orchestrator, no plugin, just a router file, an index, and code that scores before the model ever opens anything point claude at a flat folder of notes and every question reads all of them, that’s the default and it’s expensive this breakdown turns that exact pattern into a system anyone can run CLAUDE.md → index.md → nodes → edges → retrieval as code → memory → ship it twelve steps, one shared rule, evidence lives in the index, the model reads three files while the notes folder still holds thousands the stale index test, one line drifts and the whole graph quietly turns back into the pile it was built to escape, no exceptions save it and watch, then read the article below

Anthropic engineer just built a second brain that reads 3 files instead of 5,000, and cut the token bill 40% no orchestrator, no plugin, just a router file, an index, and code that scores before the model ever opens anything point claude at a flat folder of notes and every question reads all of them, that’s the default and it’s expensive this breakdown turns that exact pattern into a system anyone can run CLAUDE.md → index.md → nodes → edges → retrieval as code → memory → ship it twelve steps, one shared rule, evidence lives in the index, the model reads three files while the notes folder still holds thousands the stale index test, one line drifts and the whole graph quietly turns back into the pile it was built to escape, no exceptions save it and watch, then read the article below

leopardracer

37,304 views • 1 month ago

Agents vs. Graphs, clearly explained! spawning more agents is great, but it has a ceiling nobody says out loud: five agents is a count. a graph is a shape. only one of them changes the answer. point five agents at the same pile with the same window and they converge. the first one writes a finding, the rest read it, and all five reports centre on the same thing. you paid five times for one opinion with four echoes. Graph engineering fixes this by moving the decision up a layer: not how many agents, but who is allowed to look at what. you need both. here's how it works: ↳ the count buys you throughput. five things happening instead of one ↳ the shape buys you coverage. five different things happening instead of the same one five times Prompts → Context → Harness → Agents → Graphs the node that does this is the splitter, and it decides more than any other node in the system. cut a repository by folder and four workers audit the same three files. cut it by blast radius and each one sees something the others cannot. the trick is being selective about what each lane is allowed to see. separate contexts are not a nice-to-have, they are the mechanism. if two agents are meant to produce different things, they must not share a window. if they are meant to produce the same thing, you did not need two agents. one thing to know before you scale it. a branch that throws does not reject the batch. it resolves to null, and that is the containment. which means your merge quietly receives a short list. ↳ filter the nulls before the merge, or one dead lane poisons the whole result ↳ never index a merge by position. eight good branches and one failure will shift everything by one, silently skip that and the run looks like it worked. the output is just missing a lane, and nothing errored. and the one that eats whole nights: multi-agent setups can use up to fifteen times the total tokens of a single chat, because every lane reloads its own core. you are trading total tokens for a clean main window. usually the right trade, always a choice. below i have quoted my full guide on graph engineering. it covers the three topologies, the verifier patterns, and where the gate should actually open. save this and read it below ↓

Agents vs. Graphs, clearly explained! spawning more agents is great, but it has a ceiling nobody says out loud: five agents is a count. a graph is a shape. only one of them changes the answer. point five agents at the same pile with the same window and they converge. the first one writes a finding, the rest read it, and all five reports centre on the same thing. you paid five times for one opinion with four echoes. Graph engineering fixes this by moving the decision up a layer: not how many agents, but who is allowed to look at what. you need both. here's how it works: ↳ the count buys you throughput. five things happening instead of one ↳ the shape buys you coverage. five different things happening instead of the same one five times Prompts → Context → Harness → Agents → Graphs the node that does this is the splitter, and it decides more than any other node in the system. cut a repository by folder and four workers audit the same three files. cut it by blast radius and each one sees something the others cannot. the trick is being selective about what each lane is allowed to see. separate contexts are not a nice-to-have, they are the mechanism. if two agents are meant to produce different things, they must not share a window. if they are meant to produce the same thing, you did not need two agents. one thing to know before you scale it. a branch that throws does not reject the batch. it resolves to null, and that is the containment. which means your merge quietly receives a short list. ↳ filter the nulls before the merge, or one dead lane poisons the whole result ↳ never index a merge by position. eight good branches and one failure will shift everything by one, silently skip that and the run looks like it worked. the output is just missing a lane, and nothing errored. and the one that eats whole nights: multi-agent setups can use up to fifteen times the total tokens of a single chat, because every lane reloads its own core. you are trading total tokens for a clean main window. usually the right trade, always a choice. below i have quoted my full guide on graph engineering. it covers the three topologies, the verifier patterns, and where the gate should actually open. save this and read it below ↓

Hanako

95,369 views • 6 days ago

ANTHROPIC LEAKED A GRAPH WHERE ONE FILE WITH 7 ITEMS CUTS YOUR WORK FROM 8 HOURS TO 40 MINUTES AND TAKES 40 DECISIONS A DAY OFF YOU autonomy isn't a slider you drag to the right - it's a graph with 7 steps, and exactly one edge leads into each one. your approve is worth nothing until there's a description of what the output should be - otherwise you're approving something nobody defined. you can't run a batch without checkpoints, because one crash in the middle rolls back all 40 tasks at once. the agent already works without you, and until it writes down what it did there's nothing to check the report against. this is where the point of no return sits: past it the loop runs while you sleep. the approval didn't disappear - you moved it onto an evaluator that never gets tired and never lets one through. silence is only allowed when there's a list of what breaks it: money, deletion, anything leaving the system. the last one is the brake - a turn limit and a budget cap, whichever runs out first kills the run. you build the brake before the engine - an agent you can't stop isn't autonomous, it's just unattended. skip one edge and nothing crashes and nobody warns you - the system quietly settles two steps lower and keeps running. save this - 7 steps, 7 artefacts, zero ways to jump one ↓

ANTHROPIC LEAKED A GRAPH WHERE ONE FILE WITH 7 ITEMS CUTS YOUR WORK FROM 8 HOURS TO 40 MINUTES AND TAKES 40 DECISIONS A DAY OFF YOU autonomy isn't a slider you drag to the right - it's a graph with 7 steps, and exactly one edge leads into each one. your approve is worth nothing until there's a description of what the output should be - otherwise you're approving something nobody defined. you can't run a batch without checkpoints, because one crash in the middle rolls back all 40 tasks at once. the agent already works without you, and until it writes down what it did there's nothing to check the report against. this is where the point of no return sits: past it the loop runs while you sleep. the approval didn't disappear - you moved it onto an evaluator that never gets tired and never lets one through. silence is only allowed when there's a list of what breaks it: money, deletion, anything leaving the system. the last one is the brake - a turn limit and a budget cap, whichever runs out first kills the run. you build the brake before the engine - an agent you can't stop isn't autonomous, it's just unattended. skip one edge and nothing crashes and nobody warns you - the system quietly settles two steps lower and keeps running. save this - 7 steps, 7 artefacts, zero ways to jump one ↓

Sprytix

41,920 views • 20 days ago

Citadel quant quit and walked out with one thing fund never lets leave the building: 120-year-old math that sees a market crash before the chart does. Every big desk runs it. None of them explain it. He turned it into an HMM bot running BTC 5min markets and made $570,000. It switches strategy the second the regime flips and cuts everything the moment panic spikes. His wallet: Full breakdown below. Here's what they actually see. You look at a Bitcoin chart and see a price. A quant sees three states: bull, bear, sideways. The market's always in one. The only question that pays is which one comes next. That's the math. It's called a Markov chain, and it runs on one rule: where price goes next depends only on where it is now. You count how often each state flips to another and drop it into a grid. That grid is a map of the market. Now the part that feels illegal. Want the odds of a crash twelve steps out? You multiply that grid by itself twelve times. One line of math, and you've got the probability of every future state. Then it acts: run the trend when calm, flip to mean reversion in chop, cut everything when panic odds spike. It doesn't predict the future. It gives you the odds of every version of it, then sizes the bet to match. That's the edge. Renaissance ran 66% a year for 30 years on math like this. You can build the basic version in a weekend. Bookmark & read this before it gets buried.

Citadel quant quit and walked out with one thing fund never lets leave the building: 120-year-old math that sees a market crash before the chart does. Every big desk runs it. None of them explain it. He turned it into an HMM bot running BTC 5min markets and made $570,000. It switches strategy the second the regime flips and cuts everything the moment panic spikes. His wallet: Full breakdown below. Here's what they actually see. You look at a Bitcoin chart and see a price. A quant sees three states: bull, bear, sideways. The market's always in one. The only question that pays is which one comes next. That's the math. It's called a Markov chain, and it runs on one rule: where price goes next depends only on where it is now. You count how often each state flips to another and drop it into a grid. That grid is a map of the market. Now the part that feels illegal. Want the odds of a crash twelve steps out? You multiply that grid by itself twelve times. One line of math, and you've got the probability of every future state. Then it acts: run the trend when calm, flip to mean reversion in chop, cut everything when panic odds spike. It doesn't predict the future. It gives you the odds of every version of it, then sizes the bet to match. That's the edge. Renaissance ran 66% a year for 30 years on math like this. You can build the basic version in a weekend. Bookmark & read this before it gets buried.

cvxv666

93,596 views • 1 month ago