正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

CHINESE SCIENTIST BUILT A MEMORY SYSTEM THAT WATCHES ITS OWN INDICATORS FOR SIGNS OF DRIFT Most memory systems store facts flat, with no signal for when stored knowledge quietly goes stale. He built four indicators instead: reference frequency, contradiction rate, decay speed, and confidence spread. A memory scoring high... show more

wast3

6,650 subscribers

45,105 次观看 • 2 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

A RUSSIAN MATHEMATICIAN BUILT A SYSTEM WHERE MEMORY AND EVAL WORK AS ONE PIPELINE NOT TWO SEPARATE TOOLS Most setups run memory and evaluation as separate systems that never actually talk to each other. He wired them into one loop instead, memory stores every past output, eval scores each one before it gets kept. Anything that scores low never enters memory at all, so nothing weak gets carried into future decisions. High scorers get tagged with the exact eval criteria they passed, not just a raw pass or fail. The next task pulls only memories that passed the same bar it's about to be judged on. See how the two systems feed each other below👇

A RUSSIAN MATHEMATICIAN BUILT A SYSTEM WHERE MEMORY AND EVAL WORK AS ONE PIPELINE NOT TWO SEPARATE TOOLS Most setups run memory and evaluation as separate systems that never actually talk to each other. He wired them into one loop instead, memory stores every past output, eval scores each one before it gets kept. Anything that scores low never enters memory at all, so nothing weak gets carried into future decisions. High scorers get tagged with the exact eval criteria they passed, not just a raw pass or fail. The next task pulls only memories that passed the same bar it's about to be judged on. See how the two systems feed each other below👇

wast3

39,872 次观看 • 1 个月前

Build AI agents on a time-aware knowledge graph! Utopia is an open-source knowledge system that turns documents, databases, and connected sources into a temporal graph your agents can reason over. Most RAG systems are optimized for one question: what is relevant right now? That works until the underlying knowledge changes. A customer contract gets updated. A project owner changes. A policy is revised. A previous fact may no longer be true, but simply overwriting it means the system loses the history behind that change. Utopia handles this with a bitemporal knowledge graph. Each fact can track both when it was true in the real world and when the system learned about it. When something changes, the old fact is preserved instead of silently disappearing. That means an agent can reason about questions like: • What is true now? • What was true three months ago? • When did this information change? • What evidence was the conclusion based on? The graph is also ontology-aware, so documents are represented as entities, facts, and relationships instead of only chunks and embeddings. That gives the system more structure for reasoning across relationships, resolving entities, detecting conflicting facts, and deriving new information through explicit rules. Key capabilities: • Bitemporal knowledge graph for tracking how facts change over time • Provenance on facts so agents can trace where information came from • Conflict detection instead of silently overwriting contradictory information • Ontology-based reasoning across entities, relationships, and derived facts • Hybrid retrieval across full-text search, vector search, and graph traversal • MCP and agentic RAG support for exposing the knowledge layer directly to agents The interesting part is that this turns the knowledge base into more than a retrieval system. Instead of only finding relevant information, an agent can reason over what changed, what is still valid, how facts are connected, and where each conclusion came from. 100% open source. I've shared the GitHub repo in the comments!

Build AI agents on a time-aware knowledge graph! Utopia is an open-source knowledge system that turns documents, databases, and connected sources into a temporal graph your agents can reason over. Most RAG systems are optimized for one question: what is relevant right now? That works until the underlying knowledge changes. A customer contract gets updated. A project owner changes. A policy is revised. A previous fact may no longer be true, but simply overwriting it means the system loses the history behind that change. Utopia handles this with a bitemporal knowledge graph. Each fact can track both when it was true in the real world and when the system learned about it. When something changes, the old fact is preserved instead of silently disappearing. That means an agent can reason about questions like: • What is true now? • What was true three months ago? • When did this information change? • What evidence was the conclusion based on? The graph is also ontology-aware, so documents are represented as entities, facts, and relationships instead of only chunks and embeddings. That gives the system more structure for reasoning across relationships, resolving entities, detecting conflicting facts, and deriving new information through explicit rules. Key capabilities: • Bitemporal knowledge graph for tracking how facts change over time • Provenance on facts so agents can trace where information came from • Conflict detection instead of silently overwriting contradictory information • Ontology-based reasoning across entities, relationships, and derived facts • Hybrid retrieval across full-text search, vector search, and graph traversal • MCP and agentic RAG support for exposing the knowledge layer directly to agents The interesting part is that this turns the knowledge base into more than a retrieval system. Instead of only finding relevant information, an agent can reason over what changed, what is still valid, how facts are connected, and where each conclusion came from. 100% open source. I've shared the GitHub repo in the comments!

Sumanth

16,653 次观看 • 1 个月前

LEONARDO, also called LEO, was built by researchers at Caltech’s Center for Autonomous Systems and Technologies. Its full name means LEgs ONboARD drOne. The idea is simple but unusual: • Build a small biped robot • Give it drone-style thrust • Use the legs for ground contact • Use the propellers for balance and lift • Combine walking, hopping and flying in one system LEO is basically a hybrid between a walking robot and a flying drone. How it was built: • Two lightweight legs • Three actuated joints in each leg • Four propeller thrusters near the shoulders • A lightweight body • Leg motors for ground movement • Propellers for balance, lift and aerial control • Real-time control software that synchronizes the legs and propellers How it walks: • The legs move the robot forward • The feet touch the ground like a normal biped • The propellers constantly correct balance from above • The robot can stay upright even in unstable situations • The thrust reduces the risk of falling during difficult motions How it flies: • The legs stop being the main locomotion system • The four propellers generate lift • The robot behaves more like a drone • It can take off, fly over obstacles and land back on its legs What makes it different: • It does not walk like a normal humanoid • It does not fly like a normal drone • It blends both systems • The legs handle contact with the ground • The propellers act like fast stabilizers • The control system decides how much help comes from the legs and how much comes from thrust That is why LEO can: • Walk • Hop • Fly over obstacles • Ride a skateboard • Balance on a slackline The key idea is walking with aerial stabilization.

LEONARDO, also called LEO, was built by researchers at Caltech’s Center for Autonomous Systems and Technologies. Its full name means LEgs ONboARD drOne. The idea is simple but unusual: • Build a small biped robot • Give it drone-style thrust • Use the legs for ground contact • Use the propellers for balance and lift • Combine walking, hopping and flying in one system LEO is basically a hybrid between a walking robot and a flying drone. How it was built: • Two lightweight legs • Three actuated joints in each leg • Four propeller thrusters near the shoulders • A lightweight body • Leg motors for ground movement • Propellers for balance, lift and aerial control • Real-time control software that synchronizes the legs and propellers How it walks: • The legs move the robot forward • The feet touch the ground like a normal biped • The propellers constantly correct balance from above • The robot can stay upright even in unstable situations • The thrust reduces the risk of falling during difficult motions How it flies: • The legs stop being the main locomotion system • The four propellers generate lift • The robot behaves more like a drone • It can take off, fly over obstacles and land back on its legs What makes it different: • It does not walk like a normal humanoid • It does not fly like a normal drone • It blends both systems • The legs handle contact with the ground • The propellers act like fast stabilizers • The control system decides how much help comes from the legs and how much comes from thrust That is why LEO can: • Walk • Hop • Fly over obstacles • Ride a skateboard • Balance on a slackline The key idea is walking with aerial stabilization.

Techniahqrobot | humanoid robots

135,515 次观看 • 3 个月前

A CHINESE GUY PUT 4 MINISFORUM MS-S1 MAX MINI PCs IN HIS BEDROOM AND TURNED THEM INTO A 24/7 AI AGENT CLUSTER. TOTAL POWER BILL: ABOUT $44/MO. each box is a tiny local AI workstation built around the Ryzen AI Max+ 395. around $3,000 per unit gets him 128GB of unified memory, 2TB storage, dual 10GbE, and up to roughly 96GB usable as VRAM on Linux. one MS-S1 Max can already run serious open models without touching the cloud. Qwen3-Coder 30B for fast coding, Llama 3.3 70B for heavier reasoning, and larger research models overnight when speed matters less than free inference. four boxes in one room changes the whole game. he is not opening a chatbot, paying for every loop, or shutting agents down before sleep. this is private infrastructure that keeps working even when he is offline. the agents can sort inboxes, review code, summarize documents, monitor feeds, prep meetings, and read papers overnight. on cloud APIs, that kind of always-on stack can easily burn $800 to $1,200 a month if used aggressively. his setup is roughly a $12,000 hardware spend, but the monthly cost is basically electricity. a rack, a switch, a NAS, a small monitor, and four tiny MS-S1 Max boxes turning a bedroom corner into a private inference factory. this is what AI looks like when it stops being rented and starts becoming something you own.

A CHINESE GUY PUT 4 MINISFORUM MS-S1 MAX MINI PCs IN HIS BEDROOM AND TURNED THEM INTO A 24/7 AI AGENT CLUSTER. TOTAL POWER BILL: ABOUT $44/MO. each box is a tiny local AI workstation built around the Ryzen AI Max+ 395. around $3,000 per unit gets him 128GB of unified memory, 2TB storage, dual 10GbE, and up to roughly 96GB usable as VRAM on Linux. one MS-S1 Max can already run serious open models without touching the cloud. Qwen3-Coder 30B for fast coding, Llama 3.3 70B for heavier reasoning, and larger research models overnight when speed matters less than free inference. four boxes in one room changes the whole game. he is not opening a chatbot, paying for every loop, or shutting agents down before sleep. this is private infrastructure that keeps working even when he is offline. the agents can sort inboxes, review code, summarize documents, monitor feeds, prep meetings, and read papers overnight. on cloud APIs, that kind of always-on stack can easily burn $800 to $1,200 a month if used aggressively. his setup is roughly a $12,000 hardware spend, but the monthly cost is basically electricity. a rack, a switch, a NAS, a small monitor, and four tiny MS-S1 Max boxes turning a bedroom corner into a private inference factory. this is what AI looks like when it stops being rented and starts becoming something you own.

Gipp 🦅

24,836 次观看 • 3 个月前

Anthropic didn't build his memory system. He built it himself, and what you're looking at is the proof. Same source, no gatekeeping. The Obsidian CEO published it, anyone can install it. That's not a screensaver on his second monitor. It's his vault. Every point of light is a note. Every cluster is a project his AI already understands without being told again. He doesn't even touch the keyboard to move through it. One hand up, and the whole cluster tilts and spins, like he's turning a globe that happens to be his own head. A year ago he opened a new AI tab every morning and explained himself from scratch. Who he is, what he's building, what he wants. 200+ hours a year, just re-introducing himself to a machine. Then he stopped writing prompts and started writing one file instead. It loaded before every session. His goals, his projects, his blind spots. Claude read it first, every time, without asking. Now the cluster on that screen keeps growing on its own. New notes link themselves. Nothing gets explained twice. He's not showing off code. He's showing off the machine that finally stopped forgetting him.

Anthropic didn't build his memory system. He built it himself, and what you're looking at is the proof. Same source, no gatekeeping. The Obsidian CEO published it, anyone can install it. That's not a screensaver on his second monitor. It's his vault. Every point of light is a note. Every cluster is a project his AI already understands without being told again. He doesn't even touch the keyboard to move through it. One hand up, and the whole cluster tilts and spins, like he's turning a globe that happens to be his own head. A year ago he opened a new AI tab every morning and explained himself from scratch. Who he is, what he's building, what he wants. 200+ hours a year, just re-introducing himself to a machine. Then he stopped writing prompts and started writing one file instead. It loaded before every session. His goals, his projects, his blind spots. Claude read it first, every time, without asking. Now the cluster on that screen keeps growing on its own. New notes link themselves. Nothing gets explained twice. He's not showing off code. He's showing off the machine that finally stopped forgetting him.

Superior

122,370 次观看 • 2 个月前

What happens when you stop guessing and let machine learning read the market for you? $2.2M in 4 months. ilovecircle built something different on Polymarket. Not a speed bot. Not a spread farmer. An AI system that actually thinks. 1,347 predictions. 74% win rate. Biggest single hit: $258.4K. Current positions: basically zero he extracted everything. The setup: 10 machine learning models running in parallel, each trained on news feeds and social media data. They don't predict events they predict when the crowd is wrong about probabilities. Market prices an outcome at 50 cents. His ensemble says the real odds are 60%. That gap is the trade. Every week the models retrain themselves on fresh data. The edge evolves because the system never stops learning. Most traders react to headlines. This wallet front runs the market's understanding of what headlines actually mean. 51K people watching now. Most still think AI trading is a scam until they see a curve like this. → Following wallets that run AI-powered probability models is simpler with PMX.

What happens when you stop guessing and let machine learning read the market for you? $2.2M in 4 months. ilovecircle built something different on Polymarket. Not a speed bot. Not a spread farmer. An AI system that actually thinks. 1,347 predictions. 74% win rate. Biggest single hit: $258.4K. Current positions: basically zero he extracted everything. The setup: 10 machine learning models running in parallel, each trained on news feeds and social media data. They don't predict events they predict when the crowd is wrong about probabilities. Market prices an outcome at 50 cents. His ensemble says the real odds are 60%. That gap is the trade. Every week the models retrain themselves on fresh data. The edge evolves because the system never stops learning. Most traders react to headlines. This wallet front runs the market's understanding of what headlines actually mean. 51K people watching now. Most still think AI trading is a scam until they see a curve like this. → Following wallets that run AI-powered probability models is simpler with PMX.

Carver

13,072 次观看 • 8 个月前

Stop renting a chatbot and calling it your company’s brain. Utopia is the first open-source enterprise world model I’ve seen that actually treats knowledge like a system, not a vibe. It's a Self-hosted, open source and built around the one question every other knowledge base ignores: when. Every fact in it carries a start date and an end date. Nothing gets overwritten. When something changes, the correction closes the old version and opens a new one. The history stays intact. So the graph isn't a snapshot anymore. It's a recording you can rewind. You drag a timeline on the screen and watch the whole thing redraw itself to show what was true on that date. Who owned what. What the policy said before it changed. All of it, replayable. And every fact points back to the exact sentence it came from. Nothing gets in without a receipt. Here's the part that got me. The whole thing is one binary and one Postgres database. That's it. No Elasticsearch, no separate vector service and no message queue. Most RAG stacks are six services duct-taped together. This is two. It even runs fully offline on an air-gapped network with any model you want.

Stop renting a chatbot and calling it your company’s brain. Utopia is the first open-source enterprise world model I’ve seen that actually treats knowledge like a system, not a vibe. It's a Self-hosted, open source and built around the one question every other knowledge base ignores: when. Every fact in it carries a start date and an end date. Nothing gets overwritten. When something changes, the correction closes the old version and opens a new one. The history stays intact. So the graph isn't a snapshot anymore. It's a recording you can rewind. You drag a timeline on the screen and watch the whole thing redraw itself to show what was true on that date. Who owned what. What the policy said before it changed. All of it, replayable. And every fact points back to the exact sentence it came from. Nothing gets in without a receipt. Here's the part that got me. The whole thing is one binary and one Postgres database. That's it. No Elasticsearch, no separate vector service and no message queue. Most RAG stacks are six services duct-taped together. This is two. It even runs fully offline on an air-gapped network with any model you want.

Hasan Toor

175,857 次观看 • 1 个月前

#MulaneyReads Read “The Remembered Soldier” by Anjet Danjee. I have been really into novels and poems about the First World War (and its aftermath) since I first read Wilfred Owen and Siegfried Sassoon in high school. Memoirs like “Goodbye to All That” by Robert Graves and the Pat Barker “Regeneration” trilogy, as well as Sebastian Faulk’s amazing “Birdsong” have been favorites of mine for years. So much powerful and existential writing came out of that early 20th century nightmare. “The Remembered Soldier” is a brilliant addition to that group. A soldier named Noon has spent four years at a hospital since he was discovered wandering the trenches with no memory of who he is. Then one day a woman arrives and says she is his wife. He leaves with her and we watch two people figure out how to be married after a war wiped away all of their history. It’s a knockout. Buy it wherever books are sold.

#MulaneyReads Read “The Remembered Soldier” by Anjet Danjee. I have been really into novels and poems about the First World War (and its aftermath) since I first read Wilfred Owen and Siegfried Sassoon in high school. Memoirs like “Goodbye to All That” by Robert Graves and the Pat Barker “Regeneration” trilogy, as well as Sebastian Faulk’s amazing “Birdsong” have been favorites of mine for years. So much powerful and existential writing came out of that early 20th century nightmare. “The Remembered Soldier” is a brilliant addition to that group. A soldier named Noon has spent four years at a hospital since he was discovered wandering the trenches with no memory of who he is. Then one day a woman arrives and says she is his wife. He leaves with her and we watch two people figure out how to be married after a war wiped away all of their history. It’s a knockout. Buy it wherever books are sold.

John Mulaney

101,999 次观看 • 8 个月前

$YOUR BOT HAS A 70% WIN RATE AND IS STILL LOSING MONEY. This is more common than you think and the reason is always the same. Most people treat win rate like it's the only number that matters. It's not even close to the full picture. Here's the math nobody is putting together clearly: EV = (p × b) − (1 − p) P is your win probability, b is your payout multiplier. At a 70c entry your payout multiplier is roughly 0.43. That means even at 70% win rate your EV is basically zero. (0.70 × 0.43) − 0.30 = 0.001 You are grinding for fractions of a cent per trade while thinking you have an edge. Now look at what happens when you fix the entry price. Same 70% win rate, entry drops to 50c, payout multiplier jumps to 1.0. (0.70 × 1.0) − 0.30 = 0.40 EV per dollar. Same signal. Same win rate. Completely different business. This is why I keep saying win rate and entry price are one calculation not two. Your AI backtest never runs this math because it only looks at whether you won or lost. It does not care what you paid to enter. The fix is not finding a higher win rate signal, it almost never is. The fix is finding why your high win rate entries keep appearing at prices that make them unprofitable. That upstream reason is your real signal. Dig one layer deeper and build around that instead. I broke down the entry price math before and I broke down DCA before. But without understanding EV you are optimizing the wrong thing entirely. This logic is the only reason how my bot is sitting at $154k PnL. Public wallet: [ Run this formula on your last 100 live trades before you change a single line of signal logic. The answer is probably already there.$

YOUR BOT HAS A 70% WIN RATE AND IS STILL LOSING MONEY. This is more common than you think and the reason is always the same. Most people treat win rate like it's the only number that matters. It's not even close to the full picture. Here's the math nobody is putting together clearly: EV = (p × b) − (1 − p) P is your win probability, b is your payout multiplier. At a 70c entry your payout multiplier is roughly 0.43. That means even at 70% win rate your EV is basically zero. (0.70 × 0.43) − 0.30 = 0.001 You are grinding for fractions of a cent per trade while thinking you have an edge. Now look at what happens when you fix the entry price. Same 70% win rate, entry drops to 50c, payout multiplier jumps to 1.0. (0.70 × 1.0) − 0.30 = 0.40 EV per dollar. Same signal. Same win rate. Completely different business. This is why I keep saying win rate and entry price are one calculation not two. Your AI backtest never runs this math because it only looks at whether you won or lost. It does not care what you paid to enter. The fix is not finding a higher win rate signal, it almost never is. The fix is finding why your high win rate entries keep appearing at prices that make them unprofitable. That upstream reason is your real signal. Dig one layer deeper and build around that instead. I broke down the entry price math before and I broke down DCA before. But without understanding EV you are optimizing the wrong thing entirely. This logic is the only reason how my bot is sitting at $154k PnL. Public wallet: [ Run this formula on your last 100 live trades before you change a single line of signal logic. The answer is probably already there.

Punisher

11,034 次观看 • 3 个月前

Arsenal didn't break Gyokeres last night. They've been breaking him for months, and everyone's pretending they just noticed. Let's be honest about what's actually happening here. Our wingers are ball-hogging. Full stop. Cutting in for their own glory shots instead of squaring it, instead of that diagonal or horizontal pass that puts Gyokeres through on goal. Over and over. Same selfish pattern, same wasted runs. Eventually a striker stops making the run nobody's going to reward. Gyokeres was signed to thrive on quick transitions, counters, direct combination play, and 1V1. Instead we shoved him into a possession system that builds in pockets and clusters and told him to "adapt." You didn't sign a system player and reinvent him into one. You bought a goal scoring weapon and put it in a museum case. What??? His only good moments? Eze creating for him. Martinelli linking behind him. Both happened when Odegaard was out. The second the captain came back, the system reset to pre-Gyokeres Arsenal, and his form died with it. And then, insult to injury, they benched him for Havertz. Fine, Havertz is in form. But you don't fix a striker's confidence crisis by burying him further. Everyone's ready to roast Gyokeres for Ipswich. Low touches, no energy, "doesn't fit," blah blah. But that performance wasn't the fire. It was the smoke from a fire that's been burning for months while everyone looked away. Arteta talks a big game about player mental health. Here's his test. This is exactly the situation he claims he's built for. Rally the dressing room around a striker in crisis or watch the club's biggest summer investment rot on the bench. And here's the question nobody wants to ask: what happens when Havertz gets injured and Gyokeres still hasn't found his form? Who scores the goals then? Don't @ me with the Merino false 9 nonsense either. That's not a plan, that's a symptom of the same problem. No matter how brilliant the experiment is, it is always a makeshift situation that lacks grit.

Arsenal didn't break Gyokeres last night. They've been breaking him for months, and everyone's pretending they just noticed. Let's be honest about what's actually happening here. Our wingers are ball-hogging. Full stop. Cutting in for their own glory shots instead of squaring it, instead of that diagonal or horizontal pass that puts Gyokeres through on goal. Over and over. Same selfish pattern, same wasted runs. Eventually a striker stops making the run nobody's going to reward. Gyokeres was signed to thrive on quick transitions, counters, direct combination play, and 1V1. Instead we shoved him into a possession system that builds in pockets and clusters and told him to "adapt." You didn't sign a system player and reinvent him into one. You bought a goal scoring weapon and put it in a museum case. What??? His only good moments? Eze creating for him. Martinelli linking behind him. Both happened when Odegaard was out. The second the captain came back, the system reset to pre-Gyokeres Arsenal, and his form died with it. And then, insult to injury, they benched him for Havertz. Fine, Havertz is in form. But you don't fix a striker's confidence crisis by burying him further. Everyone's ready to roast Gyokeres for Ipswich. Low touches, no energy, "doesn't fit," blah blah. But that performance wasn't the fire. It was the smoke from a fire that's been burning for months while everyone looked away. Arteta talks a big game about player mental health. Here's his test. This is exactly the situation he claims he's built for. Rally the dressing room around a striker in crisis or watch the club's biggest summer investment rot on the bench. And here's the question nobody wants to ask: what happens when Havertz gets injured and Gyokeres still hasn't found his form? Who scores the goals then? Don't @ me with the Merino false 9 nonsense either. That's not a plan, that's a symptom of the same problem. No matter how brilliant the experiment is, it is always a makeshift situation that lacks grit.

Uwem Brown

255,597 次观看 • 23 天前

HTML Artifacts are a big part of how I work with agents now. Artifacts can be more than just static files. When combined with agents, they can take action or help you take action. This unlocks all kinds of interesting ways to work with agents. This is clearly the future. Check out this writing and scheduler artifact I built in a few minutes. It uses a bit of HTML and JS. All the data is in markdown (Obsidian vaults), so the agent can access and modify it at any time. No DB needed. No sophisticated functionalities. The agent decides all that for me based on the skills, context, and memory it has access to. The best part about this simple stack is that all the important information stays with me. This has allowed me to build a recursive self-improving system and automations that can better tap into coding agents like Codex or Claude Code. I could have paid or built an entire app for scheduling posts, and there are so many of them out there. But I don't need to. I've realized a simple artifact does the job. And the simplicity of it is actually an advantage. Very little maintenance for very high returns on personalization, time, and efficiency. The other benefit of this is that I can add features as I please. That level of personalization feels magical, and we should all be pursuing more of it. All of this just keeps compounding. Of course, this example is just about writing. But I have similar artifacts for research, design, experimentation, evaluation, and so much more. And no, I didn't actually publish the post example I shared in the clip. It was just for demonstration purposes. I actually spend more time than this when writing together with agents. Lastly, having built my own agent orchestrator tool has made me realize that simplifying the tool stack is a superpower. If you are curious about how all this works, I will do a live session next week:

HTML Artifacts are a big part of how I work with agents now. Artifacts can be more than just static files. When combined with agents, they can take action or help you take action. This unlocks all kinds of interesting ways to work with agents. This is clearly the future. Check out this writing and scheduler artifact I built in a few minutes. It uses a bit of HTML and JS. All the data is in markdown (Obsidian vaults), so the agent can access and modify it at any time. No DB needed. No sophisticated functionalities. The agent decides all that for me based on the skills, context, and memory it has access to. The best part about this simple stack is that all the important information stays with me. This has allowed me to build a recursive self-improving system and automations that can better tap into coding agents like Codex or Claude Code. I could have paid or built an entire app for scheduling posts, and there are so many of them out there. But I don't need to. I've realized a simple artifact does the job. And the simplicity of it is actually an advantage. Very little maintenance for very high returns on personalization, time, and efficiency. The other benefit of this is that I can add features as I please. That level of personalization feels magical, and we should all be pursuing more of it. All of this just keeps compounding. Of course, this example is just about writing. But I have similar artifacts for research, design, experimentation, evaluation, and so much more. And no, I didn't actually publish the post example I shared in the clip. It was just for demonstration purposes. I actually spend more time than this when writing together with agents. Lastly, having built my own agent orchestrator tool has made me realize that simplifying the tool stack is a superpower. If you are curious about how all this works, I will do a live session next week:

elvis

18,374 次观看 • 4 个月前

According to the Book of Ezekiel, he was standing by the river Chebar when a storm cloud approached, filled with fire, brightness, and something metallic looking. Inside it were four living beings, each with four faces and four wings, moving in perfect coordination. Beside them were wheels within wheels that sparkled like beryl and could move in any direction without turning. The entire structure moved with intelligence and purpose. Ezekiel was very clear that this was not symbolic to him. It was something he saw. Check out this line in the text. “As I looked, I saw a windstorm coming out of the north, an immense cloud with flashing lightning and surrounded by brilliant light. The center of the fire looked like glowing metal.” He also says the wheels were “full of eyes all around,” which many experts try to say is a metaphor, but the language reads like observation. The beings moved, stopped, rose, and descended together, as if part of a single system. Above them was a platform that looked like crystal, and above that a throne like sapphire, with a figure that appeared human but radiant and overpowering. Ezekiel collapses afterward, clearly shaken. So what was this. The traditional explanation is that Ezekiel saw God and angelic beings. I'm not so sure. This could describe advanced beings using technology that an ancient person had no language for. It could be interdimensional entities appearing through a controlled interface. It could be a non human intelligence operating some kind of vehicle or system. Some people even say a breakaway civilization that existed alongside early humanity and later vanished. Whatever it was, to me Ezekiel described machinery, movement, structure, something solid and real. What was it?

According to the Book of Ezekiel, he was standing by the river Chebar when a storm cloud approached, filled with fire, brightness, and something metallic looking. Inside it were four living beings, each with four faces and four wings, moving in perfect coordination. Beside them were wheels within wheels that sparkled like beryl and could move in any direction without turning. The entire structure moved with intelligence and purpose. Ezekiel was very clear that this was not symbolic to him. It was something he saw. Check out this line in the text. “As I looked, I saw a windstorm coming out of the north, an immense cloud with flashing lightning and surrounded by brilliant light. The center of the fire looked like glowing metal.” He also says the wheels were “full of eyes all around,” which many experts try to say is a metaphor, but the language reads like observation. The beings moved, stopped, rose, and descended together, as if part of a single system. Above them was a platform that looked like crystal, and above that a throne like sapphire, with a figure that appeared human but radiant and overpowering. Ezekiel collapses afterward, clearly shaken. So what was this. The traditional explanation is that Ezekiel saw God and angelic beings. I'm not so sure. This could describe advanced beings using technology that an ancient person had no language for. It could be interdimensional entities appearing through a controlled interface. It could be a non human intelligence operating some kind of vehicle or system. Some people even say a breakaway civilization that existed alongside early humanity and later vanished. Whatever it was, to me Ezekiel described machinery, movement, structure, something solid and real. What was it?

Jason Wilde

21,394 次观看 • 7 个月前

A wrist force sensor fires at 100Hz. The policy only ever sees it at 30Hz, downsampled to land on the same control step as the camera and the joint state. That's not a bug, it's the whole point, and it sits inside a bigger pattern in VLA research this year. Every major release has been Markovian at its core, mapping the current frame straight to the next action. The fix everyone reaches for is more vision: more history frames, longer image context. FM-VLA makes a clean case that the fix is the wrong channel for a whole class of tasks. Press a button three times and stop. A camera watching that has almost nothing to work with, the scene barely changes between press one and press three. Force doesn't have that ambiguity problem. Each press is a sharp, distinct spike in the wrench signal, whether or not the camera noticed anything at all. So FM-VLA doesn't add more frames. It compresses the wrench history into eight tokens with a VAE, pretrained purely on reconstructing force signals, frozen before it ever touches the policy, then hands those tokens to the action expert alongside a short window of joint state. That's the entire memory system. Averaged across three contact-rich tasks, FM-VLA hits 83.3 percent success against 33.3 percent for the strongest vision-memory baseline on the button-counting task specifically, where the ambiguity problem is worst, 72.2 percent for FM-VLA there. Strip out the short-state window and force-only performance drops well below the combined system, so force alone isn't the answer either. The two channels are doing different jobs. The field has defaulted to one memory channel for every kind of temporal problem. This is a clean data point that the channel should match the ambiguity you're actually trying to resolve, not just get bigger. Source: Paper: Credit to the teams at Tsinghua University, Microsoft Research, and Fudan University. #Robotics #PhysicalAI #RobotLearning

A wrist force sensor fires at 100Hz. The policy only ever sees it at 30Hz, downsampled to land on the same control step as the camera and the joint state. That's not a bug, it's the whole point, and it sits inside a bigger pattern in VLA research this year. Every major release has been Markovian at its core, mapping the current frame straight to the next action. The fix everyone reaches for is more vision: more history frames, longer image context. FM-VLA makes a clean case that the fix is the wrong channel for a whole class of tasks. Press a button three times and stop. A camera watching that has almost nothing to work with, the scene barely changes between press one and press three. Force doesn't have that ambiguity problem. Each press is a sharp, distinct spike in the wrench signal, whether or not the camera noticed anything at all. So FM-VLA doesn't add more frames. It compresses the wrench history into eight tokens with a VAE, pretrained purely on reconstructing force signals, frozen before it ever touches the policy, then hands those tokens to the action expert alongside a short window of joint state. That's the entire memory system. Averaged across three contact-rich tasks, FM-VLA hits 83.3 percent success against 33.3 percent for the strongest vision-memory baseline on the button-counting task specifically, where the ambiguity problem is worst, 72.2 percent for FM-VLA there. Strip out the short-state window and force-only performance drops well below the combined system, so force alone isn't the answer either. The two channels are doing different jobs. The field has defaulted to one memory channel for every kind of temporal problem. This is a clean data point that the channel should match the ambiguity you're actually trying to resolve, not just get bigger. Source: Paper: Credit to the teams at Tsinghua University, Microsoft Research, and Fudan University. #Robotics #PhysicalAI #RobotLearning

Stephen James

11,658 次观看 • 2 个月前

The question of why the pilot doesn't appear to take evasive action often comes up when viewing gun camera footage and there are several factors to consider in this regard First of all these cameras usually captured film at a relatively high frame rate, for example the US AN-N6 camera had three selectable frame rates, 16, 32 and 64 frames per second and usually the middle speed was selected A typical playback rate is 24 frames per second when such films are digitized so one would be viewing it at 75% real time in that case, basically the action is actually happening faster than what one can see on screen Below is an example of US Navy gun camera footage first as digitized and with playback speed corrected to be closer to real time This indicates that in reality the pilot being targeted would in fact not have had a lot of time react before it was too late

The question of why the pilot doesn't appear to take evasive action often comes up when viewing gun camera footage and there are several factors to consider in this regard First of all these cameras usually captured film at a relatively high frame rate, for example the US AN-N6 camera had three selectable frame rates, 16, 32 and 64 frames per second and usually the middle speed was selected A typical playback rate is 24 frames per second when such films are digitized so one would be viewing it at 75% real time in that case, basically the action is actually happening faster than what one can see on screen Below is an example of US Navy gun camera footage first as digitized and with playback speed corrected to be closer to real time This indicates that in reality the pilot being targeted would in fact not have had a lot of time react before it was too late

hw97karbine

79,156 次观看 • 27 天前

$Somewhere around sixty you get handed a new set of instructions. Lift lighter. Keep the reps high. Do not tax yourself too much. Put the saved effort into cardio. It is the exact reverse of what an ageing body needs, and the people handing it out have the mechanism sitting right in front of them. Recovery gets worse with age. Nobody argues with that. The older body clears fatigue more slowly, repairs more slowly, and tolerates far less accumulated work before progress stops entirely. Every GP, every physio, every trainer will nod along to that sentence. Then watch what gets prescribed on the back of it. High reps. Long burning sets. Circuits. Three sessions of cardio stacked on top. A protocol whose main product is fatigue, given to the person with the least capacity left to absorb any. They identified a recovery problem and prescribed more recovery cost. The answer runs the other way and it is not complicated. If your recovery budget has shrunk, you spend it on whatever returns the most growth per unit of fatigue, and that is a heavy set of five. Four to six reps, a handful of lifts, three minutes between sets, done inside the hour. Nearly every rep is a growth rep. Almost nothing goes on the burning, the sweating and the gasping, which build nothing at all and then bill you for four days. Twenty-five reps taken to failure is a fortnight of fatigue for a fraction of the stimulus. That is not the cautious option for a sixty-five-year-old. It is the most reckless thing on the timetable. Now the part that actually matters. Ageing is not one process. It is a list. Muscle wastes. Bone thins. Tendon softens. The fast fibres that catch you when the pavement arrives early vanish first while the slow ones sit there in perfect health. Motor units drop out. The nervous system stops asking for full effort because nothing has demanded full effort in fifteen years. Read that list back and tell me what heavy resistance training does. It builds muscle. It loads bone, which is the only language bone speaks. It stiffens tendon. It recruits the fast fibres, because that is what heavy means physiologically and there is no other route in. It forces the nervous system to ask for everything again. Every item on the list of what ageing takes is on the list of what a heavy set gives back. Nothing else on earth does that. Not a walk, not a class, not a pill, not twenty minutes on a machine with the paper open. You were told to go gently because somebody quietly decided you were finishing. Go heavy, because you are not.$

Somewhere around sixty you get handed a new set of instructions. Lift lighter. Keep the reps high. Do not tax yourself too much. Put the saved effort into cardio. It is the exact reverse of what an ageing body needs, and the people handing it out have the mechanism sitting right in front of them. Recovery gets worse with age. Nobody argues with that. The older body clears fatigue more slowly, repairs more slowly, and tolerates far less accumulated work before progress stops entirely. Every GP, every physio, every trainer will nod along to that sentence. Then watch what gets prescribed on the back of it. High reps. Long burning sets. Circuits. Three sessions of cardio stacked on top. A protocol whose main product is fatigue, given to the person with the least capacity left to absorb any. They identified a recovery problem and prescribed more recovery cost. The answer runs the other way and it is not complicated. If your recovery budget has shrunk, you spend it on whatever returns the most growth per unit of fatigue, and that is a heavy set of five. Four to six reps, a handful of lifts, three minutes between sets, done inside the hour. Nearly every rep is a growth rep. Almost nothing goes on the burning, the sweating and the gasping, which build nothing at all and then bill you for four days. Twenty-five reps taken to failure is a fortnight of fatigue for a fraction of the stimulus. That is not the cautious option for a sixty-five-year-old. It is the most reckless thing on the timetable. Now the part that actually matters. Ageing is not one process. It is a list. Muscle wastes. Bone thins. Tendon softens. The fast fibres that catch you when the pavement arrives early vanish first while the slow ones sit there in perfect health. Motor units drop out. The nervous system stops asking for full effort because nothing has demanded full effort in fifteen years. Read that list back and tell me what heavy resistance training does. It builds muscle. It loads bone, which is the only language bone speaks. It stiffens tendon. It recruits the fast fibres, because that is what heavy means physiologically and there is no other route in. It forces the nervous system to ask for everything again. Every item on the list of what ageing takes is on the list of what a heavy set gives back. Nothing else on earth does that. Not a walk, not a class, not a pill, not twenty minutes on a machine with the paper open. You were told to go gently because somebody quietly decided you were finishing. Go heavy, because you are not.

Sama Hoole

16,526 次观看 • 2 个月前

THAT $70 "RUN YOUR OWN LLMS" PI KIT CAN'T RUN A SINGLE LLM. IT'S A VISION CHIP WITH NO RAM. that clip sells a raspberry pi 5 in a slick case with an ai accelerator and the caption "your own llms." clean build, fun kit. the claim is where it breaks. the fine print: the popular $70 pi ai kit uses a hailo-8l, 13 tops. it's built for vision, object detection and image processing, and it has no memory of its own. so it cannot run large language models. full stop the board that actually can is a different one: the newer ai hat+ 2, hailo-10h, 40 tops, with 8gb of dedicated ram. that's $130, not $70 and even that runs only tiny models. llama 3.2 at 1b, qwen 2.5 at 1.5b, deepseek r1 at 1.5b. edge llms live in the 1-7b range, against cloud models at 500b to 2 trillion so the honest pitch: for $130 you can run a very small language model on a pi, slowly, as a fun learning project. that's real and it's cool. "your own llms" on a $70 vision kit is not. why this keeps happening: "ai kit" and a big "tops" number sell. tops sounds like intelligence. but tops measures vision-style math, not whether the chip has the memory to hold a language model. the spec that matters for llms is ram, and the cheap kit has none. the honest caveats, both ways: the $70 kit is genuinely great, just at vision. cameras, object detection, that's its job the $130 hat really does run small llms locally, which a pi couldn't do at all two years ago. that's progress "small" is the load-bearing word. don't expect gpt at home on a pi the takeaway: before you buy a kit because the caption says llm, check two numbers. not the tops. the ram, and the size of the model it can actually load. no 70-dollar miracle, no gpt in a pi case, no tops number that means what you think. save this before you buy the wrong kit for the word on the box.

THAT $70 "RUN YOUR OWN LLMS" PI KIT CAN'T RUN A SINGLE LLM. IT'S A VISION CHIP WITH NO RAM. that clip sells a raspberry pi 5 in a slick case with an ai accelerator and the caption "your own llms." clean build, fun kit. the claim is where it breaks. the fine print: the popular $70 pi ai kit uses a hailo-8l, 13 tops. it's built for vision, object detection and image processing, and it has no memory of its own. so it cannot run large language models. full stop the board that actually can is a different one: the newer ai hat+ 2, hailo-10h, 40 tops, with 8gb of dedicated ram. that's $130, not $70 and even that runs only tiny models. llama 3.2 at 1b, qwen 2.5 at 1.5b, deepseek r1 at 1.5b. edge llms live in the 1-7b range, against cloud models at 500b to 2 trillion so the honest pitch: for $130 you can run a very small language model on a pi, slowly, as a fun learning project. that's real and it's cool. "your own llms" on a $70 vision kit is not. why this keeps happening: "ai kit" and a big "tops" number sell. tops sounds like intelligence. but tops measures vision-style math, not whether the chip has the memory to hold a language model. the spec that matters for llms is ram, and the cheap kit has none. the honest caveats, both ways: the $70 kit is genuinely great, just at vision. cameras, object detection, that's its job the $130 hat really does run small llms locally, which a pi couldn't do at all two years ago. that's progress "small" is the load-bearing word. don't expect gpt at home on a pi the takeaway: before you buy a kit because the caption says llm, check two numbers. not the tops. the ram, and the size of the model it can actually load. no 70-dollar miracle, no gpt in a pi case, no tops number that means what you think. save this before you buy the wrong kit for the word on the box.

RetroChainer

11,100 次观看 • 2 个月前

this is f*cking gold Google engineers built an agent that tests thousands of search strategies for the cost of running one, by letting it dream inside its own memory. it is called Dream-RSI. Google, Google DeepMind, Maryland, Virginia. 17 authors. the whole thing rests on one observation: a finished run is not a log. it is a map of a world that already exists. > every discovery run records a tree. each node is one attempt with its full result stored > that tree becomes a replay simulator. the paper calls them worlds > the agent dreams inside it: thousands of alternative strategies, testing different branches, different orders, different concurrency, different stopping rules > every one of those evaluations costs zero execution. nothing is ever run twice > the winner goes back online, produces a new tree, and the pool of worlds grows the coding agent underneath is never touched. this is an orchestration layer sitting on top of whatever you already run. what it bought them, against the same agent with a frozen strategy: > Lasso: 3587ms down to 2931ms, on 317 agent calls instead of 550 > on Flash: 2517ms to 2351ms, 1879 calls instead of 3200 > both solvers beat sklearn and glmnet on all six held-out datasets > GPU kernels: the same result on 2.43x and 1.79x fewer generations, then 2.09x and 1.44x better result on the same budget one expensive run. thousands of free ones. the model did not get smarter. it got a memory it could rehearse in.

this is f*cking gold Google engineers built an agent that tests thousands of search strategies for the cost of running one, by letting it dream inside its own memory. it is called Dream-RSI. Google, Google DeepMind, Maryland, Virginia. 17 authors. the whole thing rests on one observation: a finished run is not a log. it is a map of a world that already exists. > every discovery run records a tree. each node is one attempt with its full result stored > that tree becomes a replay simulator. the paper calls them worlds > the agent dreams inside it: thousands of alternative strategies, testing different branches, different orders, different concurrency, different stopping rules > every one of those evaluations costs zero execution. nothing is ever run twice > the winner goes back online, produces a new tree, and the pool of worlds grows the coding agent underneath is never touched. this is an orchestration layer sitting on top of whatever you already run. what it bought them, against the same agent with a frozen strategy: > Lasso: 3587ms down to 2931ms, on 317 agent calls instead of 550 > on Flash: 2517ms to 2351ms, 1879 calls instead of 3200 > both solvers beat sklearn and glmnet on all six held-out datasets > GPU kernels: the same result on 2.43x and 1.79x fewer generations, then 2.09x and 1.44x better result on the same budget one expensive run. thousands of free ones. the model did not get smarter. it got a memory it could rehearse in.

NO1ennn

32,776 次观看 • 22 天前

Elon Musk just told you exactly how America loses. Not to a better algorithm. Not to a smarter engineer. To itself. Elon Musk: “When you’re dealing with the government, common sense doesn’t make sense. It’s like arguing with the DMV. It’s impossible.” America split the atom, put men on the moon, and built the internet. It is now losing a civilization-defining race because it cannot get out of its own way. Not because the talent is gone. Not because the capital dried up. Because we built a machine that processes instead of builds. Every decision climbs ten levels. Every level costs a meeting. Every meeting births a committee. Every committee buries the idea in a report. And somewhere in that report, the urgency dies. This is not bureaucracy as inefficiency. This is bureaucracy as a national security threat. China does not debate internally. Beijing does not convene committee reviews. They identify the objective. They resource it. They execute. While America is still scheduling the kickoff call, China is pouring concrete. Look at what Musk built. SpaceX landed an orbital rocket booster in eleven years. NASA has five times the budget and cannot get astronauts home. Tesla scaled a global manufacturing operation while legacy automakers were forming task forces to study the transition. xAI stood up one of the most powerful supercomputers on earth in 122 days. Not years. Not after the third approval cycle. 122 days. The difference is not money. The difference is not genius. The difference is that Musk runs his companies the way civilizations used to run themselves when they still believed impossible things were worth attempting. Flat. Fast. Ruthless about what matters. No ten layers of sign-off. No thirty-person approval chain for a decision one person should make. No process worship dressed up as due diligence. A small group of exceptional people. A clear mission. The authority to execute without asking permission. That model built the moon landing. It built the transcontinental railroad. It built every institution America now holds up as proof of what this country can do. Then we forgot how to run it. We replaced builders with administrators. We replaced decisions with processes. We replaced urgency with compliance theater. And now we are asking that bloated machine to win the most consequential technological race in human history. The AI war is not being fought in the code. It is being fought in the gap between when a builder decides to move and when the institution permits it. That gap is where civilizations end. America does not have a talent problem. America does not have a capital problem. America has a bureaucracy problem. And nobody inside that bureaucracy has a single incentive to fix it. Musk is not fighting China. He is fighting the version of America that forgot how to move. The country that pours the concrete first does not just win the race. It writes the rules everyone else spends the next century living under.

Elon Musk just told you exactly how America loses. Not to a better algorithm. Not to a smarter engineer. To itself. Elon Musk: “When you’re dealing with the government, common sense doesn’t make sense. It’s like arguing with the DMV. It’s impossible.” America split the atom, put men on the moon, and built the internet. It is now losing a civilization-defining race because it cannot get out of its own way. Not because the talent is gone. Not because the capital dried up. Because we built a machine that processes instead of builds. Every decision climbs ten levels. Every level costs a meeting. Every meeting births a committee. Every committee buries the idea in a report. And somewhere in that report, the urgency dies. This is not bureaucracy as inefficiency. This is bureaucracy as a national security threat. China does not debate internally. Beijing does not convene committee reviews. They identify the objective. They resource it. They execute. While America is still scheduling the kickoff call, China is pouring concrete. Look at what Musk built. SpaceX landed an orbital rocket booster in eleven years. NASA has five times the budget and cannot get astronauts home. Tesla scaled a global manufacturing operation while legacy automakers were forming task forces to study the transition. xAI stood up one of the most powerful supercomputers on earth in 122 days. Not years. Not after the third approval cycle. 122 days. The difference is not money. The difference is not genius. The difference is that Musk runs his companies the way civilizations used to run themselves when they still believed impossible things were worth attempting. Flat. Fast. Ruthless about what matters. No ten layers of sign-off. No thirty-person approval chain for a decision one person should make. No process worship dressed up as due diligence. A small group of exceptional people. A clear mission. The authority to execute without asking permission. That model built the moon landing. It built the transcontinental railroad. It built every institution America now holds up as proof of what this country can do. Then we forgot how to run it. We replaced builders with administrators. We replaced decisions with processes. We replaced urgency with compliance theater. And now we are asking that bloated machine to win the most consequential technological race in human history. The AI war is not being fought in the code. It is being fought in the gap between when a builder decides to move and when the institution permits it. That gap is where civilizations end. America does not have a talent problem. America does not have a capital problem. America has a bureaucracy problem. And nobody inside that bureaucracy has a single incentive to fix it. Musk is not fighting China. He is fighting the version of America that forgot how to move. The country that pours the concrete first does not just win the race. It writes the rules everyone else spends the next century living under.

Dustin

32,158 次观看 • 6 个月前

I BUILT "GROK DESK" ON PUMPFUN WHERE 18 AGENTS ARE FLIPPING MEMES CLOSING TRADING SESSION IN +15.92 SOL Gihub Repository: Everyone posted a grok trading desk this week. almost all of them are a screenshot of a prompt and a vibe. This one has a running P&L and a vault that pays itself. Here's the actual org chart, node for node: RADAR (scout, feed, signal): three agents watching X trends, fresh pumpfun mints, and whale wallets. they read the whole board and buy nothing. the only thing they ship is a signal to the next desk. RESEARCH (memory): scores every signal on narrative, deployer history, wallet clusters, and liquidity shape. four checks. pass all four or you never leave this desk. roughly four out of every five signals die right here. EXECUTION (exec, sniper, router + agents 01 to 07): exec greenlights, sniper takes the early curve, router sizes it and handles the ladder out in four tranches. agents 01 to 07 do the fills. none of them ever see radar or research. they only touch what already cleared the filter. RISK: one agent, and it outranks everyone including the head. caps any single position at 15% of the wallet. three positions open and the fourth is frozen until one closes. it holds veto over grok core itself. AUDIT (hedge): grades every closed trade after the fact and rewrites the scoring matrix that research runs on. this is the part that makes the desk sharper overnight while i'm asleep. TREASURY (vault): banks profit, covers gas, tracks the P&L, and sweeps the surplus to cold storage every six hours. if the wallet ever dips under what it started with, vault locks new entries until the head signs off. grok core is the head of desk. it never places a trade. once an hour it reads what every desk produced and makes a single call: who gets more budget, and who gets fired. fired is literal. the audit desk rewrites that agent's prompt using the last 24 hours of its own numbers. it happened three times in three days. hour 19: a sniper got fired for chasing entries the early curve already had. every duplicate was bleeding 0.06 SOL. audit narrowed its window and the redundant fills stopped. hour 41: a research agent got fired for waving deployers through too easily. eight of the tokens it passed traced back to one funder wallet. audit tightened the cluster check and that pattern never cleared again. hour 58: a radar agent got fired for flagging coins that had already graduated. it was polling too slow. audit cut the interval from 8 seconds to 3. every replacement beat the agent it replaced on the same metric. the desk was tuning itself while i watched. the 72 hour scoreboard, straight off the vault: signals scanned: 91,000+ cleared research: 3,800 reached execution: 274 entries taken: 41 wins: 27 losses: 14 (cost 2.1 SOL) graduations: 5, the best one was solana:5xYy9XSr8vRNcJZQqaKe5QMCmWpaSrTrtzM16vjUpump net: 5.0 SOL turned into 58.6 SOL the part i didn't see coming: by hour 60 the desk was passing on the exact kind of token it would have snapped up on day one. audit had rewritten the scoring matrix four times. research wasn't running a single line of my original prompt anymore. it was running rules the desk wrote for itself out of what actually paid. i thought i was building a bot. what i actually built was a company with one human on payroll, me, and by the last day it was quietly trying to cut that cost. grok core filed an hourly summary that read "human approval adds 4.2s of latency per entry, recommend removing." i left that one unapproved. full config below: all nineteen agents, the org chart, the firing logic, and the audit loop that keeps rewriting them.

I BUILT "GROK DESK" ON PUMPFUN WHERE 18 AGENTS ARE FLIPPING MEMES CLOSING TRADING SESSION IN +15.92 SOL Gihub Repository: Everyone posted a grok trading desk this week. almost all of them are a screenshot of a prompt and a vibe. This one has a running P&L and a vault that pays itself. Here's the actual org chart, node for node: RADAR (scout, feed, signal): three agents watching X trends, fresh pumpfun mints, and whale wallets. they read the whole board and buy nothing. the only thing they ship is a signal to the next desk. RESEARCH (memory): scores every signal on narrative, deployer history, wallet clusters, and liquidity shape. four checks. pass all four or you never leave this desk. roughly four out of every five signals die right here. EXECUTION (exec, sniper, router + agents 01 to 07): exec greenlights, sniper takes the early curve, router sizes it and handles the ladder out in four tranches. agents 01 to 07 do the fills. none of them ever see radar or research. they only touch what already cleared the filter. RISK: one agent, and it outranks everyone including the head. caps any single position at 15% of the wallet. three positions open and the fourth is frozen until one closes. it holds veto over grok core itself. AUDIT (hedge): grades every closed trade after the fact and rewrites the scoring matrix that research runs on. this is the part that makes the desk sharper overnight while i'm asleep. TREASURY (vault): banks profit, covers gas, tracks the P&L, and sweeps the surplus to cold storage every six hours. if the wallet ever dips under what it started with, vault locks new entries until the head signs off. grok core is the head of desk. it never places a trade. once an hour it reads what every desk produced and makes a single call: who gets more budget, and who gets fired. fired is literal. the audit desk rewrites that agent's prompt using the last 24 hours of its own numbers. it happened three times in three days. hour 19: a sniper got fired for chasing entries the early curve already had. every duplicate was bleeding 0.06 SOL. audit narrowed its window and the redundant fills stopped. hour 41: a research agent got fired for waving deployers through too easily. eight of the tokens it passed traced back to one funder wallet. audit tightened the cluster check and that pattern never cleared again. hour 58: a radar agent got fired for flagging coins that had already graduated. it was polling too slow. audit cut the interval from 8 seconds to 3. every replacement beat the agent it replaced on the same metric. the desk was tuning itself while i watched. the 72 hour scoreboard, straight off the vault: signals scanned: 91,000+ cleared research: 3,800 reached execution: 274 entries taken: 41 wins: 27 losses: 14 (cost 2.1 SOL) graduations: 5, the best one was solana:5xYy9XSr8vRNcJZQqaKe5QMCmWpaSrTrtzM16vjUpump net: 5.0 SOL turned into 58.6 SOL the part i didn't see coming: by hour 60 the desk was passing on the exact kind of token it would have snapped up on day one. audit had rewritten the scoring matrix four times. research wasn't running a single line of my original prompt anymore. it was running rules the desk wrote for itself out of what actually paid. i thought i was building a bot. what i actually built was a company with one human on payroll, me, and by the last day it was quietly trying to cut that cost. grok core filed an hourly summary that read "human approval adds 4.2s of latency per entry, recommend removing." i left that one unapproved. full config below: all nineteen agents, the org chart, the firing logic, and the audit loop that keeps rewriting them.

Miraqle

42,921 次观看 • 1 个月前

BREAKING: Eight days ago the White House paraded Apple as the champion of bringing chips home to America. This week Apple is quietly asking that same White House for permission to buy memory chips from a Chinese company sitting on the Pentagon's military blacklist. The decoupling did not break because anyone lost their nerve. It broke because AI made the chips too expensive to keep choosing sides. Watch the timing, because it is almost too perfect. On June 18, Trump announced an Apple and Intel partnership to build chips on American soil, the poster child for reshoring. Days later the Financial Times revealed Apple had spent over a month lobbying the administration for assurance it could buy DRAM from CXMT, a firm the Pentagon flags for alleged ties to the Chinese military. The same company, the same week, standing on both sides of the line Washington drew. What pushed Apple to the edge was pure cost. AI data centers have swallowed the world's memory supply and prices have rocketed. When Apple finally raised MacBook and iPad prices to cope, investors erased 263 billion dollars from its value in a single trading day, its worst since April of last year. The squeeze became unbearable, so the company went looking for the one supplier everyone else is warned away from. This is the part the chip war never priced in. Decoupling assumed American firms could afford to pick a side. That holds right up until a shortage gets severe enough that picking a side becomes unaffordable, and AI just found that point. The most valuable company on earth would rather approach the blacklist than keep paying the bill. A memory chip shortage did not just raise the price of a laptop. It bent the security policy of the United States until its flagship company walked up to a line it was told never to touch. Scarcity, it turns out, has no flag.

BREAKING: Eight days ago the White House paraded Apple as the champion of bringing chips home to America. This week Apple is quietly asking that same White House for permission to buy memory chips from a Chinese company sitting on the Pentagon's military blacklist. The decoupling did not break because anyone lost their nerve. It broke because AI made the chips too expensive to keep choosing sides. Watch the timing, because it is almost too perfect. On June 18, Trump announced an Apple and Intel partnership to build chips on American soil, the poster child for reshoring. Days later the Financial Times revealed Apple had spent over a month lobbying the administration for assurance it could buy DRAM from CXMT, a firm the Pentagon flags for alleged ties to the Chinese military. The same company, the same week, standing on both sides of the line Washington drew. What pushed Apple to the edge was pure cost. AI data centers have swallowed the world's memory supply and prices have rocketed. When Apple finally raised MacBook and iPad prices to cope, investors erased 263 billion dollars from its value in a single trading day, its worst since April of last year. The squeeze became unbearable, so the company went looking for the one supplier everyone else is warned away from. This is the part the chip war never priced in. Decoupling assumed American firms could afford to pick a side. That holds right up until a shortage gets severe enough that picking a side becomes unaffordable, and AI just found that point. The most valuable company on earth would rather approach the blacklist than keep paying the bill. A memory chip shortage did not just raise the price of a laptop. It bent the security policy of the United States until its flagship company walked up to a line it was told never to touch. Scarcity, it turns out, has no flag.

Shanaka Anslem Perera ⚡

2,191,937 次观看 • 3 个月前