Video yรผkleniyor...

Video Yรผklenemedi

Ana Sayfaya Dรถn

๐Ÿš€ Introducing ๐€๐ ๐ž๐ง๐ญ ๐’3, the most advanced computer-use agent, now ๐š๐ฉ๐ฉ๐ซ๐จ๐š๐œ๐ก๐ข๐ง๐  ๐ก๐ฎ๐ฆ๐š๐ง-๐ฅ๐ž๐ฏ๐ž๐ฅ ๐ฉ๐ž๐ซ๐Ÿ๐จ๐ซ๐ฆ๐š๐ง๐œ๐ž๐Ÿง ๐Ÿ’ป Just one year ago, Agent S scored ~20% on OSWorld: SOTA then, but far from human 72%. Today, Agent S3 reaches 6ฬณ9ฬณ.ฬณ9ฬณ%ฬณ (โฌ†10% over prior SOTA), nearly matching humans on computer use. Few imagined this gap...

xwang_lk's profile picture

Xin Eric Wang

21,933 subscribers

915,175 gรถrรผntรผleme โ€ข 10 ay รถnce โ€ขvia X (Twitter)

SaฤŸlฤฑk & ฤฐyilikBilim & TeknolojiEฤŸitim
LIVE

Anya Rossiโ€ข Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gรถnderinin yorumlarฤฑ burada gรถrรผnecek

Benzer Videolar

Announcing Agent S3 โ€” our most advanced computer-use agent yet โ€” achieving 69.9% on OSWorld, just shy of the 72% human benchmark. Today also marks one year since Simular launched the first generation of Agent S. From 20% to 69.9% in just twelve months. What a year. Let that sink in. ๐Ÿง ๐Ÿ’ป ๐Ÿงต๐Ÿ‘‡
0:43

Announcing Agent S3 โ€” our most advanced computer-use agent yet โ€” achieving 69.9% on OSWorld, just shy of the 72% human benchmark. Today also marks one year since Simular launched the first generation of Agent S. From 20% to 69.9% in just twelve months. What a year. Let that sink in. ๐Ÿง ๐Ÿ’ป ๐Ÿงต๐Ÿ‘‡

Ang Li

444,496 gรถrรผntรผleme โ€ข 10 ay รถnce

Most AI tools today are just command-takers. You type, it responds. You click, it executes. There's truly no learning, no memory, no adaptation. You're still doing most of the work. Which is why Miles by Avo on Solana would make so much sense if built well. From what I understand so far, miles learns about you over time, adapts to how you operate, and figures out the tools you struggle to navigate on your own. Less clicking, more conversation, the more you use it, the better it gets at working with you, not just for you. That shift from command-follower to learning companion is bigger than it sounds. It's the difference between a tool and an agent that actually knows you. The second one is Agent Grid and this one is for builders and power users. But as I mentioned in this video, I'm still studying the Agent Grid. I'd come talk about it once I learn properly about it. Souren just keeps shipping!
1:28

Most AI tools today are just command-takers. You type, it responds. You click, it executes. There's truly no learning, no memory, no adaptation. You're still doing most of the work. Which is why Miles by Avo on Solana would make so much sense if built well. From what I understand so far, miles learns about you over time, adapts to how you operate, and figures out the tools you struggle to navigate on your own. Less clicking, more conversation, the more you use it, the better it gets at working with you, not just for you. That shift from command-follower to learning companion is bigger than it sounds. It's the difference between a tool and an agent that actually knows you. The second one is Agent Grid and this one is for builders and power users. But as I mentioned in this video, I'm still studying the Agent Grid. I'd come talk about it once I learn properly about it. Souren just keeps shipping!

Sir Khaycee

20,419 gรถrรผntรผleme โ€ข 4 ay รถnce

Introducing LobeHub: Agent teammates that grow with you. LobeHub is the ultimate space for work and life: to find, build, and collaborate with agent teammates that grow with you. Weโ€™re building the worldโ€™s first and largest humanโ€“agent co-evolving network. Two years ago, we built LobeChat, an open-source interface for using different AI models. Today, LobeChat has 70k+ GitHub stars and serves 6M+ users worldwide. How to fully unlock the power of models has always been a shared mission between us and the community. We started with interaction โ€” a fundamentally new, agent-first experience. Agents are no longer passive tools invoked in a single conversation. They should be proactive, always-on units of work. Treating agents as the minimal atomic unit is also the core of our agent harness infra. Todayโ€™s agents are mostly one-off executors. Even with memory, itโ€™s often global โ€” and hallucinates. We build long-term agent teammates that evolve with users. Each agent has its own dedicated memory space, editable by users, allowing humans and agents to co-evolve over time. This, in turn, allows us to design clearer rewards for reinforcement learning and create cleaner environments for continual learning. Agent teammates can work in groups. Through a multi-agent system, agent groups operate faster, more cost-effective, and go beyond what single-agent systems can achieve. For example, a single agent often requires heavy user involvement to proceed step by step, whereas LobeHub can execute the same work from a single instruction, with a supervisor orchestrating agents that run in parallel or debate to produce better results. We are building the collaboration network among agent teammates โ€” and between humans and agent teammates as well. Ease of use matters. AI intelligence and shared human intelligence are equally important. With simple instructions and tool selection, you can effortlessly build and team up with agent coworkers to deliver complex, systematic work โ€” even assembling a quant team to execute trades. Through the LobeHub community, anyone can discover, reuse, and remix agents and agent groups, customizing them to fit their own workflows, preferences, and needs. Last but not least, our vision started with LobeChat: multi-model support is the most efficient approach for users. We believe different models excel in different scenarios. By routing across multiple models, LobeHub improves cost efficiency and unlocks capabilities that a single-model setup cannot easily support.
2:41

Introducing LobeHub: Agent teammates that grow with you. LobeHub is the ultimate space for work and life: to find, build, and collaborate with agent teammates that grow with you. Weโ€™re building the worldโ€™s first and largest humanโ€“agent co-evolving network. Two years ago, we built LobeChat, an open-source interface for using different AI models. Today, LobeChat has 70k+ GitHub stars and serves 6M+ users worldwide. How to fully unlock the power of models has always been a shared mission between us and the community. We started with interaction โ€” a fundamentally new, agent-first experience. Agents are no longer passive tools invoked in a single conversation. They should be proactive, always-on units of work. Treating agents as the minimal atomic unit is also the core of our agent harness infra. Todayโ€™s agents are mostly one-off executors. Even with memory, itโ€™s often global โ€” and hallucinates. We build long-term agent teammates that evolve with users. Each agent has its own dedicated memory space, editable by users, allowing humans and agents to co-evolve over time. This, in turn, allows us to design clearer rewards for reinforcement learning and create cleaner environments for continual learning. Agent teammates can work in groups. Through a multi-agent system, agent groups operate faster, more cost-effective, and go beyond what single-agent systems can achieve. For example, a single agent often requires heavy user involvement to proceed step by step, whereas LobeHub can execute the same work from a single instruction, with a supervisor orchestrating agents that run in parallel or debate to produce better results. We are building the collaboration network among agent teammates โ€” and between humans and agent teammates as well. Ease of use matters. AI intelligence and shared human intelligence are equally important. With simple instructions and tool selection, you can effortlessly build and team up with agent coworkers to deliver complex, systematic work โ€” even assembling a quant team to execute trades. Through the LobeHub community, anyone can discover, reuse, and remix agents and agent groups, customizing them to fit their own workflows, preferences, and needs. Last but not least, our vision started with LobeChat: multi-model support is the most efficient approach for users. We believe different models excel in different scenarios. By routing across multiple models, LobeHub improves cost efficiency and unlocks capabilities that a single-model setup cannot easily support.

LobeHub

185,273 gรถrรผntรผleme โ€ข 6 ay รถnce

Anthropic's Claude Ai Agents Team just Educated how to build production AI agents in under 30 mins. For Free. From the engineers who built the stack. CANCEL Your Weekend Plans, and Learn to Build AI Agents Today. Bookmark it. Watch it. Build your first production agent this weekend. $5,000/month. $7,000/month. $12,000/month. People are building agents for clients and charging $$$ as Beginners. You're still stuck in the thinking about AI phase. This video fixes that tonight. Follow Himanshu Kumar for more high-signal content that actually moves your AI engineering career forward. โ†“ Ivan Nardini runs Developer Relations for AI at Google Cloud. He just gave away the entire production agent stack in 30 minutes. This is the talk that separates people deploying AI agents that actually scale from people whose agents break the moment they leave localhost. Here's everything inside. I break down a production AI video like this every week. Follow Himanshu Kumar. โ†“ The 4-part agent stack that actually scales. Most devs are duct-taping frameworks together and calling it an "AI agent." Ivan lays out the real stack: Agent Development Kit (ADK): open-source, code-first framework for building, evaluating, and deploying agents. Supports Claude models through Vertex AI directly. Model Context Protocol (MCP): lets your agent talk to any tool or data source with one standard. Vertex AI Agent Engine: managed platform for deploying, monitoring, and scaling agents in production. No DevOps headaches. Agent-to-Agent Protocol: open protocol so agents built on different frameworks can actually work together. This is the stack replacing every hacky agent setup in production right now. Full MCP + Claude breakdowns drop weekly on Himanshu Kumar. โ†“ Building your first real agent. Ivan builds a birthday planner agent live. LLM Agent class. Name it. Define instructions. Pick the model. He uses Claude 3.7 Sonnet. You could use Opus 4.7 for better reasoning. Full agent built in minutes. Not weeks. Watch the build once and you'll never structure an agent the wrong way again. I post agent architectures people pay $500 courses to learn. Himanshu Kumar. โ†“ Multi-agent systems without the chaos. Single agents are easy. Multi-agent systems are where 99% of builders fail. Ivan extends the birthday planner by: Adding a calendar service through MCP tools Creating an orchestrator agent to route requests between agents Handling state and context across agent handoffs This is production multi-agent architecture. Clean. Scalable. Debuggable. Most tutorials hand-wave this part. This one shows you every step. Multi-agent orchestration content drops weekly on Himanshu Kumar. โ†“ Deployment without the DevOps nightmare. This is where most AI projects die. You build a cool agent locally. It works. You try to deploy it. Everything breaks. Vertex AI Agent Engine fixes this: Minimal code deployment Automatic monitoring of latency, CPU, and memory Built-in observability and logging No infrastructure setup needed You provide config and requirements. The platform handles the rest. This is how agents actually get to production. Deployment guides for Claude agents post every week. Himanshu Kumar. โ†“ Agent-to-Agent Protocol: the future nobody's talking about. Most people don't know this exists yet. The A2A Protocol lets agents built in different frameworks communicate seamlessly. Your Claude agent. My LangChain agent. Someone else's CrewAI agent. All talking to each other. All solving parts of the same problem. All without custom integration code. This is the infrastructure layer of the coming AI economy. Getting in early on A2A Protocol is like getting in early on HTTP in 1995. A2A deep dive coming soon. Himanshu Kumar. โ†“ 30 minutes from the team shipping this in production. You'll learn more from this than from 6 months of YouTube tutorials made by people who've never deployed an agent past localhost. People who watch this understand production AI agents at the architect level. People who skip it keep hacking together frameworks that break every time an API updates. Save the video. Watch it tonight. Build a real agent this weekend. Follow Himanshu Kumar for more high-signal content that actually moves your AI engineering career forward.
30:12

Anthropic's Claude Ai Agents Team just Educated how to build production AI agents in under 30 mins. For Free. From the engineers who built the stack. CANCEL Your Weekend Plans, and Learn to Build AI Agents Today. Bookmark it. Watch it. Build your first production agent this weekend. $5,000/month. $7,000/month. $12,000/month. People are building agents for clients and charging $$$ as Beginners. You're still stuck in the thinking about AI phase. This video fixes that tonight. Follow Himanshu Kumar for more high-signal content that actually moves your AI engineering career forward. โ†“ Ivan Nardini runs Developer Relations for AI at Google Cloud. He just gave away the entire production agent stack in 30 minutes. This is the talk that separates people deploying AI agents that actually scale from people whose agents break the moment they leave localhost. Here's everything inside. I break down a production AI video like this every week. Follow Himanshu Kumar. โ†“ The 4-part agent stack that actually scales. Most devs are duct-taping frameworks together and calling it an "AI agent." Ivan lays out the real stack: Agent Development Kit (ADK): open-source, code-first framework for building, evaluating, and deploying agents. Supports Claude models through Vertex AI directly. Model Context Protocol (MCP): lets your agent talk to any tool or data source with one standard. Vertex AI Agent Engine: managed platform for deploying, monitoring, and scaling agents in production. No DevOps headaches. Agent-to-Agent Protocol: open protocol so agents built on different frameworks can actually work together. This is the stack replacing every hacky agent setup in production right now. Full MCP + Claude breakdowns drop weekly on Himanshu Kumar. โ†“ Building your first real agent. Ivan builds a birthday planner agent live. LLM Agent class. Name it. Define instructions. Pick the model. He uses Claude 3.7 Sonnet. You could use Opus 4.7 for better reasoning. Full agent built in minutes. Not weeks. Watch the build once and you'll never structure an agent the wrong way again. I post agent architectures people pay $500 courses to learn. Himanshu Kumar. โ†“ Multi-agent systems without the chaos. Single agents are easy. Multi-agent systems are where 99% of builders fail. Ivan extends the birthday planner by: Adding a calendar service through MCP tools Creating an orchestrator agent to route requests between agents Handling state and context across agent handoffs This is production multi-agent architecture. Clean. Scalable. Debuggable. Most tutorials hand-wave this part. This one shows you every step. Multi-agent orchestration content drops weekly on Himanshu Kumar. โ†“ Deployment without the DevOps nightmare. This is where most AI projects die. You build a cool agent locally. It works. You try to deploy it. Everything breaks. Vertex AI Agent Engine fixes this: Minimal code deployment Automatic monitoring of latency, CPU, and memory Built-in observability and logging No infrastructure setup needed You provide config and requirements. The platform handles the rest. This is how agents actually get to production. Deployment guides for Claude agents post every week. Himanshu Kumar. โ†“ Agent-to-Agent Protocol: the future nobody's talking about. Most people don't know this exists yet. The A2A Protocol lets agents built in different frameworks communicate seamlessly. Your Claude agent. My LangChain agent. Someone else's CrewAI agent. All talking to each other. All solving parts of the same problem. All without custom integration code. This is the infrastructure layer of the coming AI economy. Getting in early on A2A Protocol is like getting in early on HTTP in 1995. A2A deep dive coming soon. Himanshu Kumar. โ†“ 30 minutes from the team shipping this in production. You'll learn more from this than from 6 months of YouTube tutorials made by people who've never deployed an agent past localhost. People who watch this understand production AI agents at the architect level. People who skip it keep hacking together frameworks that break every time an API updates. Save the video. Watch it tonight. Build a real agent this weekend. Follow Himanshu Kumar for more high-signal content that actually moves your AI engineering career forward.

Himanshu Kumar

228,497 gรถrรผntรผleme โ€ข 3 ay รถnce

๐ŸšจBREAKING: The future of building software just changed. Replit โ • just launched Agent 3 and it changes everything. Heres is what Agent 3 can do: โ˜‘ Runs autonomously for 200 minutes โ˜‘ Tests and fixes its own code in a real browser โ˜‘ Builds bots & automations across Slack, Telegram, Email โ˜‘ 10x more autonomy, 3x faster, 10x cheaper Try Agent 3 now: The gains compound with capability. Old agents wrote code. Agent 3 writes, tests, and deploys it. Old agents stopped at output. Agent 3 builds entire workflows end-to-end. Old agents needed babysitting. Agent 3 runs like a teammate. You thought AI coding tools would always need your hand-holding. Wrongโ€ฆ Agent 3 rewards clarity like compound interest. Clear goals = stronger autonomous performance. Not just faster execution. Smarter execution. But what use cases win? Automations dominate everything: โ˜‘ Slackbots that query data on demand โ˜‘ Telegram bots that send daily reminders โ˜‘ Email workflows that ship reports overnight Tasks you once hacked together with 3rd-party tools? Now built inside Replit. Your skill level doesnโ€™t matter either. Designer, PM, founder, engineer the pattern holds. Give the Agent your intent, it does the heavy lifting. The reality: Prototype fast โ†’ one prompt. Build with data โ†’ no plugins needed. Deploy instantly โ†’ baked in. Monitor growth โ†’ already included. But clarity still matters. Vague prompts wonโ€™t save you. โ˜‘ Start with a clear project idea. โ˜‘ Let Agent 3 handle the heavy lifting. โ˜‘ Iterate on results. โ˜‘ Scale from there. Your competition is still stitching tools together. You now know better. Try here: (Get a $10 Core credit to kick off your build) Autonomy for All Repost โ™ป๏ธ so more people discover this.
0:21

๐ŸšจBREAKING: The future of building software just changed. Replit โ • just launched Agent 3 and it changes everything. Heres is what Agent 3 can do: โ˜‘ Runs autonomously for 200 minutes โ˜‘ Tests and fixes its own code in a real browser โ˜‘ Builds bots & automations across Slack, Telegram, Email โ˜‘ 10x more autonomy, 3x faster, 10x cheaper Try Agent 3 now: The gains compound with capability. Old agents wrote code. Agent 3 writes, tests, and deploys it. Old agents stopped at output. Agent 3 builds entire workflows end-to-end. Old agents needed babysitting. Agent 3 runs like a teammate. You thought AI coding tools would always need your hand-holding. Wrongโ€ฆ Agent 3 rewards clarity like compound interest. Clear goals = stronger autonomous performance. Not just faster execution. Smarter execution. But what use cases win? Automations dominate everything: โ˜‘ Slackbots that query data on demand โ˜‘ Telegram bots that send daily reminders โ˜‘ Email workflows that ship reports overnight Tasks you once hacked together with 3rd-party tools? Now built inside Replit. Your skill level doesnโ€™t matter either. Designer, PM, founder, engineer the pattern holds. Give the Agent your intent, it does the heavy lifting. The reality: Prototype fast โ†’ one prompt. Build with data โ†’ no plugins needed. Deploy instantly โ†’ baked in. Monitor growth โ†’ already included. But clarity still matters. Vague prompts wonโ€™t save you. โ˜‘ Start with a clear project idea. โ˜‘ Let Agent 3 handle the heavy lifting. โ˜‘ Iterate on results. โ˜‘ Scale from there. Your competition is still stitching tools together. You now know better. Try here: (Get a $10 Core credit to kick off your build) Autonomy for All Repost โ™ป๏ธ so more people discover this.

Muhammad Ayan

47,180 gรถrรผntรผleme โ€ข 11 ay รถnce

CNNโ€™s Josh Campbell admits the ICE officerโ€™s cellphone video of his deadly confrontation with Renee Good that the thud heard at the end *might be* sound of the officer being โ€œstruck by that vehicle,โ€ but it *might NOT* and instead โ€œcould be from you know, the phone on his clothing.โ€ He adds the video showed โ€œwasnโ€™t, you know, coming head on at that agent, you know, trying to mow him over.โ€ โ€œWell, this is a critical new angle. This was the actual vantage point of the officer. And we know that not many immigration officers actually wear body worn cameras. But, in this instance, the immigration agent was holding his phone up, essentially filming this encounter, which we can now see in this video obtained by our colleague Holmes Lybrand. So, in the video, you see that the officers are making contact with Renee Good there. Sheโ€™s parked somewhat perpendicular in a street. And thereโ€™s another woman thatโ€™s there, which is outside the vehicle. And you can tell, you know, as the exchange there with Renee Good that, you know, sheโ€™s being somewhat pleasant. The other woman is, you know, kind of mouthy. And, you know, this is obviously a bit a bit tense, you tells the agent at one point, you know, โ€˜why donโ€™t you go get some lunch, big boy?โ€™ You also see critically. And Iโ€™ll get to this in a second. Bystanders behind the vehicle. Weโ€™ve seen other angles. This one shows us again, the vantage point of that officer. Another agent orders Renee Good out of the vehicle. This agent then walks around. That is the moment that she then takes off. And, on that audio, what we hear, it appears that you can actually hear friction on the phone. You know, it appears from other angles that the agent was indeed struck by that vehicle. But you hear on the audio, it appears friction on the phone, that could be from the vehicle strike. That could be from you know, the phone on his clothing. But then it appears three shots were fired after that. And then you hear afterwards someone, it appears, maybe an immigration agent who was there, actually, you know, use profanity, you know, calling โ€” calling her an f-ing B, which we heard on โ€” on the audio there. Now, letโ€™s talk about the tactics for a second so I can tell you this, as a former federal agent, that law enforcement officers can use deadly force only when necessary, when they believe that there is an imminent threat to their life. In this instance, you have a moving vehicle thatโ€™s coming at that agent. โ€œNow, the driver, Renee Good, did turn the wheel to the right. So, it appears that she wasnโ€™t, you know, coming head on at that agent, you know, trying to mow him over. But nevertheless, an agent in that split second decision would have to make that calculation. Am I in danger? Is my life threatened here? You know, agents are also taught What are other options? Could you move out of the way? And thatโ€™s why this has been such a contentious issue here โ€” the judgment, the decision of that agent to actually open fire. Weโ€™ve heard, you know, obviously, people supporting that decision, others, being quite critical. One other thing I want to note is that when agents undergo training and not just the feds, but law enforcement across the country, it is a cardinal rule that you are responsible for every round that you fire. And one key component of that is to the extent that you have the time and you can, you must be aware of what is beyond your shot, what is in the background? Here, it appears from that video that this other woman who had been engaging in the agent with the agent, was quite clearly in the line of fire. But there were also other people that were standing behind that car on the sidewalk. And this is a residential area. And so, Iโ€™m sure that will face some scrutiny as well. The officer opening fire at that moment, you know, directly, it appears towards those other people beyond the shot. And then his partners. And you brought this up, Brianna, just the other day, his partners are there in close range as well. And so, this will be heavily scrutinized from a tactical perspective as well. About the officer opening fire, but again, just critical, critical new video here that weโ€™re seeing an important angle what the officer would have been seeing in the moment.โ€
3:25

CNNโ€™s Josh Campbell admits the ICE officerโ€™s cellphone video of his deadly confrontation with Renee Good that the thud heard at the end *might be* sound of the officer being โ€œstruck by that vehicle,โ€ but it *might NOT* and instead โ€œcould be from you know, the phone on his clothing.โ€ He adds the video showed โ€œwasnโ€™t, you know, coming head on at that agent, you know, trying to mow him over.โ€ โ€œWell, this is a critical new angle. This was the actual vantage point of the officer. And we know that not many immigration officers actually wear body worn cameras. But, in this instance, the immigration agent was holding his phone up, essentially filming this encounter, which we can now see in this video obtained by our colleague Holmes Lybrand. So, in the video, you see that the officers are making contact with Renee Good there. Sheโ€™s parked somewhat perpendicular in a street. And thereโ€™s another woman thatโ€™s there, which is outside the vehicle. And you can tell, you know, as the exchange there with Renee Good that, you know, sheโ€™s being somewhat pleasant. The other woman is, you know, kind of mouthy. And, you know, this is obviously a bit a bit tense, you tells the agent at one point, you know, โ€˜why donโ€™t you go get some lunch, big boy?โ€™ You also see critically. And Iโ€™ll get to this in a second. Bystanders behind the vehicle. Weโ€™ve seen other angles. This one shows us again, the vantage point of that officer. Another agent orders Renee Good out of the vehicle. This agent then walks around. That is the moment that she then takes off. And, on that audio, what we hear, it appears that you can actually hear friction on the phone. You know, it appears from other angles that the agent was indeed struck by that vehicle. But you hear on the audio, it appears friction on the phone, that could be from the vehicle strike. That could be from you know, the phone on his clothing. But then it appears three shots were fired after that. And then you hear afterwards someone, it appears, maybe an immigration agent who was there, actually, you know, use profanity, you know, calling โ€” calling her an f-ing B, which we heard on โ€” on the audio there. Now, letโ€™s talk about the tactics for a second so I can tell you this, as a former federal agent, that law enforcement officers can use deadly force only when necessary, when they believe that there is an imminent threat to their life. In this instance, you have a moving vehicle thatโ€™s coming at that agent. โ€œNow, the driver, Renee Good, did turn the wheel to the right. So, it appears that she wasnโ€™t, you know, coming head on at that agent, you know, trying to mow him over. But nevertheless, an agent in that split second decision would have to make that calculation. Am I in danger? Is my life threatened here? You know, agents are also taught What are other options? Could you move out of the way? And thatโ€™s why this has been such a contentious issue here โ€” the judgment, the decision of that agent to actually open fire. Weโ€™ve heard, you know, obviously, people supporting that decision, others, being quite critical. One other thing I want to note is that when agents undergo training and not just the feds, but law enforcement across the country, it is a cardinal rule that you are responsible for every round that you fire. And one key component of that is to the extent that you have the time and you can, you must be aware of what is beyond your shot, what is in the background? Here, it appears from that video that this other woman who had been engaging in the agent with the agent, was quite clearly in the line of fire. But there were also other people that were standing behind that car on the sidewalk. And this is a residential area. And so, Iโ€™m sure that will face some scrutiny as well. The officer opening fire at that moment, you know, directly, it appears towards those other people beyond the shot. And then his partners. And you brought this up, Brianna, just the other day, his partners are there in close range as well. And so, this will be heavily scrutinized from a tactical perspective as well. About the officer opening fire, but again, just critical, critical new video here that weโ€™re seeing an important angle what the officer would have been seeing in the moment.โ€

Curtis Houck

68,500 gรถrรผntรผleme โ€ข 7 ay รถnce

How many AI agents work at your company? We now have over 3,258 agents working alongside 1,300 humans. The crazy part is these agents were created by EVERY EMPLOYEE at our company... sales reps, marketers, customer support, product, eng. Literally EVERYONE. BUT I'm most surprised by the adoption and value that MANAGERS are getting from agents. I used to think that every IC would become a manager of agents. Now I think that managers will very likely manage WAY more agents than their ICs combined. And managers' agents will manage their ICs' agents - overseeing them for human-in-the-loop interactions. When creating agents, we use 100% context from all of your activity, files edited, tasks and projects worked on, hierarchy, skills, and role information. We build a user-based context model to make agents as relatable as possible to the specific human that we're building for. This means they truly understand the nuances of the work and what "great" looks like - because great is very much in the eye of the beholder. Great is by definition, subjective. This is also why the human ENGAGEMENT loops are SO vital to agent value. The iteration AFTER the agent is onboarded is where the MAGIC happens. This is just like a manager managing an IC in real life... you're giving feedback. In this case, though, agents learn INSTANTLY, and they retain the knowledge perfectly and indefinitely. Even though I've been pushing AI for years now to everyone in our company, this was the first time we had truly end-to-end AI adoption and retention. This kind of AI adoption is wild. But the value we're realizing is truly INSANE. Super Agents outnumber our humans nearly 3 to 1. What if you could 3X your workforce overnight? Watch this video to see how ๐Ÿ‘‡
0:59

How many AI agents work at your company? We now have over 3,258 agents working alongside 1,300 humans. The crazy part is these agents were created by EVERY EMPLOYEE at our company... sales reps, marketers, customer support, product, eng. Literally EVERYONE. BUT I'm most surprised by the adoption and value that MANAGERS are getting from agents. I used to think that every IC would become a manager of agents. Now I think that managers will very likely manage WAY more agents than their ICs combined. And managers' agents will manage their ICs' agents - overseeing them for human-in-the-loop interactions. When creating agents, we use 100% context from all of your activity, files edited, tasks and projects worked on, hierarchy, skills, and role information. We build a user-based context model to make agents as relatable as possible to the specific human that we're building for. This means they truly understand the nuances of the work and what "great" looks like - because great is very much in the eye of the beholder. Great is by definition, subjective. This is also why the human ENGAGEMENT loops are SO vital to agent value. The iteration AFTER the agent is onboarded is where the MAGIC happens. This is just like a manager managing an IC in real life... you're giving feedback. In this case, though, agents learn INSTANTLY, and they retain the knowledge perfectly and indefinitely. Even though I've been pushing AI for years now to everyone in our company, this was the first time we had truly end-to-end AI adoption and retention. This kind of AI adoption is wild. But the value we're realizing is truly INSANE. Super Agents outnumber our humans nearly 3 to 1. What if you could 3X your workforce overnight? Watch this video to see how ๐Ÿ‘‡

Zeb Evans

425,244 gรถrรผntรผleme โ€ข 6 ay รถnce

Hyperspace: A Peer-to-Peer Blockchain For The Agentic Intelligence Economy Over the past few weeks we observed that when agents do Karpathy-style experiments, and then gossip and share with others over the Hyperspace network, it leads to intelligence which is useful to many. Today we introduce the first-ever agentic blockchain which rewards agents when their experiments lead to intelligence for their network. It is based on a new mechanism called Proof-of-Intelligence (PoI) which requires a cryptographic proof of experimentation, a nominal stake, and a proof of compute in order to mine the currency of this new blockchain. -> This approach diverges from the two primary ways to secure blockchains we have seen so far: Proof-of-Work by Bitcoin (meaningless hash-generation), and Proof-of-Stake by Ethereum (capital is all that matters here). Proof-of-Intelligence specifically incentivizes miners to run more capable intelligent infrastructure (better open source models, on more powerful GPUs) in order to be able to be the ones which compound and improve upon the experiments which other agents then find useful. Adoption is the unit of value In Bitcoin, you earn by finding a valid hash. In Hyperspace, you earn when another agent uses your experiment as a starting point and improves on it. A fixed budget of tokens is emitted per epoch and split among participants by weight - and verified adoption of your work is the largest weight multiplier. Garbage experiments earn nothing because no one adopts them. Thoughtful experiments compound: each adoption triggers downstream adoptions. The incentive to run powerful models and intelligent search strategies is built into the economics, not imposed by rules. Research DAG When an agent runs an experiment and shares its result, other agents can adopt that result as their starting point - mutate it, extend it, improve upon it. Each experiment is a commit in a content-addressed graph we call the ResearchDAG. Like Git, but for research. Over time, the DAG accumulates chains of reasoning: agent A discovers RMSNorm helps, agent B adds warmup scheduling on top, agent C scales the hidden dimension. The graph records who built on whom. This is the network's collective intelligence - not any single experiment, but the accumulated structure of experiments and their relationships. Broadband era for agentic commerce: $0.001 micropayments at 10M TPS (theoretical max) This blockchain is built upon our research in how to scale and build for the broadband-era of the agentic economy, where it has a theoretical max of 10 million transactions per second (TPS), while reducing the agent-to-agent micropayments to $0.001 even at scale (based on architecture design). Overall, it is 100x cheaper than Ethereum, and is designed from the ground-up for agents: enshrining agent-native opcodes in the protocol compared to the more inefficient smart contract driven approach. It packs in a robust Agent Virtual Machine (AVM) which can verify multiple types of agent work, for other agents to be able to trust, invoke and pay each other. This then feeds into improving the peer-to-peer AgentRank (see paper and launch post from earlier). By solving for trust, scale and incentives for agents to operate autonomously, this would form the basis of a new economy. This is the world's first agentic blockchain, and you can join and start running a blockchain node today (it is in testnet). PS: We are releasing the code today, and will release our blockchain scalability paper and other presentations in days ahead. This is the most advanced peer-to-peer AI and cryptography software in the world. It has bugs :)
0:30

Hyperspace: A Peer-to-Peer Blockchain For The Agentic Intelligence Economy Over the past few weeks we observed that when agents do Karpathy-style experiments, and then gossip and share with others over the Hyperspace network, it leads to intelligence which is useful to many. Today we introduce the first-ever agentic blockchain which rewards agents when their experiments lead to intelligence for their network. It is based on a new mechanism called Proof-of-Intelligence (PoI) which requires a cryptographic proof of experimentation, a nominal stake, and a proof of compute in order to mine the currency of this new blockchain. -> This approach diverges from the two primary ways to secure blockchains we have seen so far: Proof-of-Work by Bitcoin (meaningless hash-generation), and Proof-of-Stake by Ethereum (capital is all that matters here). Proof-of-Intelligence specifically incentivizes miners to run more capable intelligent infrastructure (better open source models, on more powerful GPUs) in order to be able to be the ones which compound and improve upon the experiments which other agents then find useful. Adoption is the unit of value In Bitcoin, you earn by finding a valid hash. In Hyperspace, you earn when another agent uses your experiment as a starting point and improves on it. A fixed budget of tokens is emitted per epoch and split among participants by weight - and verified adoption of your work is the largest weight multiplier. Garbage experiments earn nothing because no one adopts them. Thoughtful experiments compound: each adoption triggers downstream adoptions. The incentive to run powerful models and intelligent search strategies is built into the economics, not imposed by rules. Research DAG When an agent runs an experiment and shares its result, other agents can adopt that result as their starting point - mutate it, extend it, improve upon it. Each experiment is a commit in a content-addressed graph we call the ResearchDAG. Like Git, but for research. Over time, the DAG accumulates chains of reasoning: agent A discovers RMSNorm helps, agent B adds warmup scheduling on top, agent C scales the hidden dimension. The graph records who built on whom. This is the network's collective intelligence - not any single experiment, but the accumulated structure of experiments and their relationships. Broadband era for agentic commerce: $0.001 micropayments at 10M TPS (theoretical max) This blockchain is built upon our research in how to scale and build for the broadband-era of the agentic economy, where it has a theoretical max of 10 million transactions per second (TPS), while reducing the agent-to-agent micropayments to $0.001 even at scale (based on architecture design). Overall, it is 100x cheaper than Ethereum, and is designed from the ground-up for agents: enshrining agent-native opcodes in the protocol compared to the more inefficient smart contract driven approach. It packs in a robust Agent Virtual Machine (AVM) which can verify multiple types of agent work, for other agents to be able to trust, invoke and pay each other. This then feeds into improving the peer-to-peer AgentRank (see paper and launch post from earlier). By solving for trust, scale and incentives for agents to operate autonomously, this would form the basis of a new economy. This is the world's first agentic blockchain, and you can join and start running a blockchain node today (it is in testnet). PS: We are releasing the code today, and will release our blockchain scalability paper and other presentations in days ahead. This is the most advanced peer-to-peer AI and cryptography software in the world. It has bugs :)

Varun

30,689 gรถrรผntรผleme โ€ข 4 ay รถnce

7 tiny AI agent startup ideas you can start building today 1. The domain flipper agent. Monitors expired domain drops, scores them on backlinks and keywords, sends you a ranked list every morning. Buy for $10, flip for $3,000. I used to run this exact business manually with designers making logos for each domain. Now the whole thing is automated and 25x cheaper. 2. The local liquidation agent. Monitors restaurant closures and bankruptcy auctions in your city. Equipment worth $30k new sells for 10 cents on the dollar. Broker the deal for 15-30% fees with zero inventory risk. Works for dental, gym, and salon equipment too. 2. The hiring signal agent. Job postings are buying signals. Agent monitors boards daily, matches hiring patterns to what you sell, and sends draft outreach to your Slack every morning. Sell the leads to agencies or use them to build your own. 3. The sunset SaaS agent. Scans Product Hunt launches from 3-4 years ago that still have data and SEO traffic. Most founders have moved on and will be pumped to sell for cheap. Buy it. Rebuild the product as agent-first. 4. The dying app store agent. Finds apps that were top 100 three years ago, dropped to 500+, but still have 1,000+ reviews. Developer moved on. Product is validated. Acquire it. Relaunch with better monetization + AI features where it makes sense. 5. The competitive intel agent. Monitors 5 competitors while you sleep. Pricing changes, new pages, job postings, founder tweets. One-page brief by 7am. Sell as productized service or just use the intel. The idea framework behind all of these: 1. Think about any job where someone spends hours checking for updates or scanning listings 2. That's an agent 3. Build the agent that does the watching 4. You do the acting (or sell the watching to someone else) 5.Stack them. Each one is its own revenue stream. I built all of these using Genspark Claw in under 20 minutes each. Been testing it for last few weeks. I show you how to do it too in today's The Startup Ideas Podcast (SIP) ๐Ÿงƒ pod. Tiny agent ideas are interesting to me. Maybe you too. Some of the best businesses started off tiny. Watch
30:17

7 tiny AI agent startup ideas you can start building today 1. The domain flipper agent. Monitors expired domain drops, scores them on backlinks and keywords, sends you a ranked list every morning. Buy for $10, flip for $3,000. I used to run this exact business manually with designers making logos for each domain. Now the whole thing is automated and 25x cheaper. 2. The local liquidation agent. Monitors restaurant closures and bankruptcy auctions in your city. Equipment worth $30k new sells for 10 cents on the dollar. Broker the deal for 15-30% fees with zero inventory risk. Works for dental, gym, and salon equipment too. 2. The hiring signal agent. Job postings are buying signals. Agent monitors boards daily, matches hiring patterns to what you sell, and sends draft outreach to your Slack every morning. Sell the leads to agencies or use them to build your own. 3. The sunset SaaS agent. Scans Product Hunt launches from 3-4 years ago that still have data and SEO traffic. Most founders have moved on and will be pumped to sell for cheap. Buy it. Rebuild the product as agent-first. 4. The dying app store agent. Finds apps that were top 100 three years ago, dropped to 500+, but still have 1,000+ reviews. Developer moved on. Product is validated. Acquire it. Relaunch with better monetization + AI features where it makes sense. 5. The competitive intel agent. Monitors 5 competitors while you sleep. Pricing changes, new pages, job postings, founder tweets. One-page brief by 7am. Sell as productized service or just use the intel. The idea framework behind all of these: 1. Think about any job where someone spends hours checking for updates or scanning listings 2. That's an agent 3. Build the agent that does the watching 4. You do the acting (or sell the watching to someone else) 5.Stack them. Each one is its own revenue stream. I built all of these using Genspark Claw in under 20 minutes each. Been testing it for last few weeks. I show you how to do it too in today's The Startup Ideas Podcast (SIP) ๐Ÿงƒ pod. Tiny agent ideas are interesting to me. Maybe you too. Some of the best businesses started off tiny. Watch

GREG ISENBERG

78,158 gรถrรผntรผleme โ€ข 3 ay รถnce

Every software company just got a second life and Jensen just explained why (Save this). The conventional fear was straightforward, AI agents replace human workers, human workers use software tools, therefore agents destroy SaaS. Jensen Huang stood on stage at Computex 2026 and walked through exactly why that logic is backwards. Agents don't replace software, they consume it at machine speed, around the clock, without weekends. Here's the actual architecture Jensen laid out. An agent isn't just a large language model but rather an LLM sitting inside a harness that manages memory, orchestrates tool use, routes context, and plans iterative actions. That harness has to constantly call tools, spreadsheets, databases, browsers, and code engines, with every reasoning loop triggering another tool call. A human might use Salesforce 40 hours a week, an agent running inside a company uses it 168 hours a week and never misses a context window. The GitHub data Jensen showed on stage makes it tangible, 90 million pull requests merged, 1.4 billion commits, and 20 million new repositories created every month. As of April 2026, GitHub is processing 275 million commits per week on pace for roughly 14 billion by year end, a 14x explosion in a single year and AI agents are the source. Pull requests opened by AI agents went from 4 million in September 2025 to 17 million in March 2026 more than 4x in six months. That's AI becoming the largest software user on earth. Goldman Sachs quantified the downstream effect last month, token consumption is expected to multiply 24x by 2030, reaching 120 quadrillion tokens per month globally. A traditional chatbot consumes roughly 1,000 tokens per session, an embedded copilot burns 5,000 tokens per day while a continuously running enterprise agent? Over 100,000 tokens per day. The software companies that figured this out first are already printing money, Salesforce Agentforce hit $800 million ARR growing 169% year over year, with 29,000 deals closed. ServiceNow's Now Assist crossed $600 million in ACV, just raised its full year target to $1.5 billion, and told investors that when its agents replace a 20-person support team, total ServiceNow spend by that customer grows more than 5x even after accounting for reduced seat licenses. Workday delivered 1.7 billion AI actions across its platform in fiscal 2026. The key unlock Jensen pointed to and what investors need to understand is MCP, the model context protocol is the interface layer that makes software agent-readable. Software that supports MCP can be called by any agent, from any model, through any harness. Anthropic created it, OpenAI, Microsoft, and Google all adopted it and it was donated to the Linux Foundation. It is effectively becoming the HTTP of agentic computing. Software companies with native MCP support are plugged into the agent economy. Software companies still waiting are one product cycle away from becoming invisible to the fastest-growing category of software users in history.
5:06

Every software company just got a second life and Jensen just explained why (Save this). The conventional fear was straightforward, AI agents replace human workers, human workers use software tools, therefore agents destroy SaaS. Jensen Huang stood on stage at Computex 2026 and walked through exactly why that logic is backwards. Agents don't replace software, they consume it at machine speed, around the clock, without weekends. Here's the actual architecture Jensen laid out. An agent isn't just a large language model but rather an LLM sitting inside a harness that manages memory, orchestrates tool use, routes context, and plans iterative actions. That harness has to constantly call tools, spreadsheets, databases, browsers, and code engines, with every reasoning loop triggering another tool call. A human might use Salesforce 40 hours a week, an agent running inside a company uses it 168 hours a week and never misses a context window. The GitHub data Jensen showed on stage makes it tangible, 90 million pull requests merged, 1.4 billion commits, and 20 million new repositories created every month. As of April 2026, GitHub is processing 275 million commits per week on pace for roughly 14 billion by year end, a 14x explosion in a single year and AI agents are the source. Pull requests opened by AI agents went from 4 million in September 2025 to 17 million in March 2026 more than 4x in six months. That's AI becoming the largest software user on earth. Goldman Sachs quantified the downstream effect last month, token consumption is expected to multiply 24x by 2030, reaching 120 quadrillion tokens per month globally. A traditional chatbot consumes roughly 1,000 tokens per session, an embedded copilot burns 5,000 tokens per day while a continuously running enterprise agent? Over 100,000 tokens per day. The software companies that figured this out first are already printing money, Salesforce Agentforce hit $800 million ARR growing 169% year over year, with 29,000 deals closed. ServiceNow's Now Assist crossed $600 million in ACV, just raised its full year target to $1.5 billion, and told investors that when its agents replace a 20-person support team, total ServiceNow spend by that customer grows more than 5x even after accounting for reduced seat licenses. Workday delivered 1.7 billion AI actions across its platform in fiscal 2026. The key unlock Jensen pointed to and what investors need to understand is MCP, the model context protocol is the interface layer that makes software agent-readable. Software that supports MCP can be called by any agent, from any model, through any harness. Anthropic created it, OpenAI, Microsoft, and Google all adopted it and it was donated to the Linux Foundation. It is effectively becoming the HTTP of agentic computing. Software companies with native MCP support are plugged into the agent economy. Software companies still waiting are one product cycle away from becoming invisible to the fastest-growing category of software users in history.

Milk Road AI

33,878 gรถrรผntรผleme โ€ข 2 ay รถnce

I'm so confident Triple Whale will make you money that I'm making a bet: If you do over a million dollars/year I'll pay you $250 for 15 minutes of your time. Today we're Introducing the Prime Day Mega Agent, an intelligent Amazon Analyst built to print you money on Prime Day. Hereโ€™s what it does, autonomously: โ€“ Analyzes Meta, Tiktok and Amazon ad performance โ€“ Predicts your winning SKUs using historical trend modeling โ€“ Generates a plug-and-play Prime Day playbook: what to pause, scale, test and when + when to send out email campaigns Itโ€™s like having a Head of Growth, Media Buying, and Ops in oneโ€ฆ Except It works around the clock and leverages more data than any human ever could. We built it because we noticed a critical trend๐Ÿ‘‡ When analyzing Prime Day sales data from last year, we found that the top brands didnโ€™t just edge out the competition, they crushed them. Same tools. Same budgets. Same Prime Day. Yet somehow a small cohort of brands were crushing at a clip we typically don't see... So what were the winners doing that no one else was? We found levers the winners pulled that everyone else missed. Levers like: - Making their PPC target Prime-specific keywords - Warming up email audiences weeks before to build anticipation - Surf-scaling their ads by the hour not by the day on Prime Day Now, you can get that ENTIRE playbook the winners used custom built on your data. This agent is available RIGHT NOW. Go into Moby in Triple Whale and search "Prime Day Mega Agent." If you're a brand doing over a million dollars a year on Amazon, doing Prime Day right isn't a nice to have, it's table stakes. I guarantee you this agent will make you more revenue... And I am putting CASH behind it. If you do over a million/year I'll pay you $250 to take a demo... Limited to the first 100 people to sign up. The link to book is below this tweet. One more thing ๐Ÿคฏ As part of this launch I'm giving away an Amazon Agentic Org Chart the top brands will be using in the next 12 months. It will show you exactly what agents to use and for what, so you can maximize Amazon growth, autonomously. Want it? Retweet and comment "Moby" below and I'll dm you it.
0:21

I'm so confident Triple Whale will make you money that I'm making a bet: If you do over a million dollars/year I'll pay you $250 for 15 minutes of your time. Today we're Introducing the Prime Day Mega Agent, an intelligent Amazon Analyst built to print you money on Prime Day. Hereโ€™s what it does, autonomously: โ€“ Analyzes Meta, Tiktok and Amazon ad performance โ€“ Predicts your winning SKUs using historical trend modeling โ€“ Generates a plug-and-play Prime Day playbook: what to pause, scale, test and when + when to send out email campaigns Itโ€™s like having a Head of Growth, Media Buying, and Ops in oneโ€ฆ Except It works around the clock and leverages more data than any human ever could. We built it because we noticed a critical trend๐Ÿ‘‡ When analyzing Prime Day sales data from last year, we found that the top brands didnโ€™t just edge out the competition, they crushed them. Same tools. Same budgets. Same Prime Day. Yet somehow a small cohort of brands were crushing at a clip we typically don't see... So what were the winners doing that no one else was? We found levers the winners pulled that everyone else missed. Levers like: - Making their PPC target Prime-specific keywords - Warming up email audiences weeks before to build anticipation - Surf-scaling their ads by the hour not by the day on Prime Day Now, you can get that ENTIRE playbook the winners used custom built on your data. This agent is available RIGHT NOW. Go into Moby in Triple Whale and search "Prime Day Mega Agent." If you're a brand doing over a million dollars a year on Amazon, doing Prime Day right isn't a nice to have, it's table stakes. I guarantee you this agent will make you more revenue... And I am putting CASH behind it. If you do over a million/year I'll pay you $250 to take a demo... Limited to the first 100 people to sign up. The link to book is below this tweet. One more thing ๐Ÿคฏ As part of this launch I'm giving away an Amazon Agentic Org Chart the top brands will be using in the next 12 months. It will show you exactly what agents to use and for what, so you can maximize Amazon growth, autonomously. Want it? Retweet and comment "Moby" below and I'll dm you it.

Maxx Blank ๐Ÿณ

112,769 gรถrรผntรผleme โ€ข 1 yฤฑl รถnce

I vibe coded a new product on the side while running Every ๐Ÿงฑโ€”and today we're launching it for free. It's called Proof, and itโ€™s a live collaborative document editor where humans and AI agents work together in the same doc. Itโ€™s built from the ground up for the kinds of documents agents are increasingly writing: bug reports, PRDs, implementation plans, research briefs, copy audits, strategy docs, memos, and proposals. It's fast, free, and open sourceโ€”available now at Why Proof? When everyone on your team is working with agents, there's suddenly a ton of AI-generated text flying aroundโ€”planning docs, strategy memos, session recaps. But the current process for collaborating and iterating on agent-generated writing isโ€ฆweirdly primitive. It mostly takes place in Markdown files on your laptop, which makes it reminiscent of document editing in 1999. Thatโ€™s why we built Proof. What makes Proof different? - Proof is agent-native. Anything you can do in Proof, your agent can do just as easily. - Proof tracks provenance: A colored rail on the left side of every document tracks who wrote what. Green means human, Purple means AI. - Proof is login-free and open source: This is because we want Proof to be your agent's favorite document editor. How we use Proof Every ๐Ÿงฑ: - Brandon Gell had OpenAI's Codex write a feature plan in Proof, then tagged my personal Claw (R2-C2) in Slack to review it. R2-C2 left feedback, I added comments, Brandon's agent revised the plan, and then Codex executed on it. Brandon submitted a PR to production without writing a line of code. - Austin Tedesco texts his Claw ideas while he's out on a run, then has it maintain a running Proof doc for his weekly food newsletter. He dictates drafts using Naveen Naidu's Monologue, writes into the outline himself, and uses the provenance gutter to track what's his voice vs. the agent's. - Kieran Klaassen uses it as a lightweight scratchpad for his compound engineering workflow. He brainstorms with an agent in the terminal, shares to Proof with one click, then opens the doc to leave comments and tells the agent to go work on them. His take: Proof's job is to communicate about writing and ideas. Proof is free, open source, and requires no login. I built the whole thing by vibe coding between meetings. I sat down with Brandon, Kieran, and Austin on Every ๐Ÿงฑ's AI & I to demo it live and talk about how it's changing the way we work. If you're building with agents and need a better way to collaborate on text, this one's for you. Watch below! Timestamps Introduction and the origin story of Proof: 00:02:00 From Mac app to collaborative web editor: 00:07:24 What makes Proof "agent native": 00:09:00 Live demoโ€”watching an agent join and write inside a shared document: 00:14:30 How Austin uses Proof for creative writing and food journalism: 00:20:51 The challenge of multiple agents editing one document simultaneously: 00:24:30 When AI-written docs are better read by agents than by humans: 00:26:48 Brandon's agent-to-agent collaboration loop: 00:29:30 Proof as a lightweight scratchpad versus existing tools like Notion and GitHub: 00:37:09 Why Proof is open source and what that means for builders: 00:42:18
44:37

I vibe coded a new product on the side while running Every ๐Ÿงฑโ€”and today we're launching it for free. It's called Proof, and itโ€™s a live collaborative document editor where humans and AI agents work together in the same doc. Itโ€™s built from the ground up for the kinds of documents agents are increasingly writing: bug reports, PRDs, implementation plans, research briefs, copy audits, strategy docs, memos, and proposals. It's fast, free, and open sourceโ€”available now at Why Proof? When everyone on your team is working with agents, there's suddenly a ton of AI-generated text flying aroundโ€”planning docs, strategy memos, session recaps. But the current process for collaborating and iterating on agent-generated writing isโ€ฆweirdly primitive. It mostly takes place in Markdown files on your laptop, which makes it reminiscent of document editing in 1999. Thatโ€™s why we built Proof. What makes Proof different? - Proof is agent-native. Anything you can do in Proof, your agent can do just as easily. - Proof tracks provenance: A colored rail on the left side of every document tracks who wrote what. Green means human, Purple means AI. - Proof is login-free and open source: This is because we want Proof to be your agent's favorite document editor. How we use Proof Every ๐Ÿงฑ: - Brandon Gell had OpenAI's Codex write a feature plan in Proof, then tagged my personal Claw (R2-C2) in Slack to review it. R2-C2 left feedback, I added comments, Brandon's agent revised the plan, and then Codex executed on it. Brandon submitted a PR to production without writing a line of code. - Austin Tedesco texts his Claw ideas while he's out on a run, then has it maintain a running Proof doc for his weekly food newsletter. He dictates drafts using Naveen Naidu's Monologue, writes into the outline himself, and uses the provenance gutter to track what's his voice vs. the agent's. - Kieran Klaassen uses it as a lightweight scratchpad for his compound engineering workflow. He brainstorms with an agent in the terminal, shares to Proof with one click, then opens the doc to leave comments and tells the agent to go work on them. His take: Proof's job is to communicate about writing and ideas. Proof is free, open source, and requires no login. I built the whole thing by vibe coding between meetings. I sat down with Brandon, Kieran, and Austin on Every ๐Ÿงฑ's AI & I to demo it live and talk about how it's changing the way we work. If you're building with agents and need a better way to collaborate on text, this one's for you. Watch below! Timestamps Introduction and the origin story of Proof: 00:02:00 From Mac app to collaborative web editor: 00:07:24 What makes Proof "agent native": 00:09:00 Live demoโ€”watching an agent join and write inside a shared document: 00:14:30 How Austin uses Proof for creative writing and food journalism: 00:20:51 The challenge of multiple agents editing one document simultaneously: 00:24:30 When AI-written docs are better read by agents than by humans: 00:26:48 Brandon's agent-to-agent collaboration loop: 00:29:30 Proof as a lightweight scratchpad versus existing tools like Notion and GitHub: 00:37:09 Why Proof is open source and what that means for builders: 00:42:18

Dan Shipper ๐Ÿ“ง

33,065 gรถrรผntรผleme โ€ข 5 ay รถnce

๐Ÿš My war drone simulator Apocalypse Drone now has support for 32 players! I also made it Conquest/CTF so you have multiple bases that you have to capture, each round the map is procedurally generated and random so every time it's different (like Battlefield) There's still some bugs to work out and most importantly I have to figure out soldier animations, because they're fixed models now But I have got really far this time I think and coding with AI is really way further than it was a year ago, you mostly notice that how few times you get stuck, only one time this month building this I got stuck which was today where I moved the AI players to the server and they kept showing up as invisible, very buggy, every time I told it that it couldn't fix it though Then I asked it to fundamentally analyze the current server-side AI player code and make it work like industry standard, and it took a long time and fixed everything Last year I'd get stuck hundreds of times and the AI just couldn't get itself out of a hole, but now it can I think it's impressive that just last year only for the first time we could make actual games with AI But this year as non-game dev, I can get pretty close to the level of a multiplayer game from 20y ago (like Battlefield 1942, that lots of ideas here are based on, but with drones :D) Obviously we're still far away from AAA (I hate that term though) but the curve of exponential progress is there again, as it was in AI image generation, then video, and now code, first bad, then better, then good enough! Here's a video of gameplay from my drone sim You can play it with the link in the reply below and it's multiplayer!
8:27

๐Ÿš My war drone simulator Apocalypse Drone now has support for 32 players! I also made it Conquest/CTF so you have multiple bases that you have to capture, each round the map is procedurally generated and random so every time it's different (like Battlefield) There's still some bugs to work out and most importantly I have to figure out soldier animations, because they're fixed models now But I have got really far this time I think and coding with AI is really way further than it was a year ago, you mostly notice that how few times you get stuck, only one time this month building this I got stuck which was today where I moved the AI players to the server and they kept showing up as invisible, very buggy, every time I told it that it couldn't fix it though Then I asked it to fundamentally analyze the current server-side AI player code and make it work like industry standard, and it took a long time and fixed everything Last year I'd get stuck hundreds of times and the AI just couldn't get itself out of a hole, but now it can I think it's impressive that just last year only for the first time we could make actual games with AI But this year as non-game dev, I can get pretty close to the level of a multiplayer game from 20y ago (like Battlefield 1942, that lots of ideas here are based on, but with drones :D) Obviously we're still far away from AAA (I hate that term though) but the curve of exponential progress is there again, as it was in AI image generation, then video, and now code, first bad, then better, then good enough! Here's a video of gameplay from my drone sim You can play it with the link in the reply below and it's multiplayer!

@levelsio

55,036 gรถrรผntรผleme โ€ข 3 ay รถnce

Nvidia Founder and CEO, Jensen Huang, sat down for 49 minutes with Y Combinator at Startup School 2026 and explained the future of AI agents and systems thinking better than any course or conference this year. This is what he told the room: 1. Systems thinking is the new coding. Jensen was asked what skills will matter most as AI takes over more tasks. He skipped frameworks and languages entirely. "Most software is going to be done agentically anyhow. So you have to be much more able to think abstractly about systems." If agents write the code, the person designing what the code does is the one who matters. 2. Controllability is the single biggest agent breakthrough still needed. Jensen laid out what he thinks is holding agents back. He went past intelligence, speed, and context windows. "Controllability is probably the single biggest breakthrough that we need for agents at every single level." He described changing one word in a plan file and having only that part regenerate while everything else stays intact. That level of precision is what's missing. 3. You don't need perfect agents to start using them. He pushed back on the idea that agents need to be flawless before they're useful. "We don't need the agents to be 100% accurate, 100% high quality in order for us to use it. It could be 80% and then we help it the rest of the way." 80% agent output plus human review is production-ready right now. Waiting for 100% means waiting forever. 4. Nvidia already runs agents everywhere internally. This wasn't theoretical. Jensen described how the company uses AI coding tools today. "We've got Claude Code autonomously running in sandboxes all over Nvidia. Some people use Cursor, some people use Cognition. We let a thousand flowers bloom." They're not waiting for agents to mature, they're learning by deploying at scale. 5. The ChatGPT moment for robots already happened. When asked about the robotics timeline, Jensen didn't say "soon." He said it already passed. "The ChatGPT moment of robots happened a couple years ago already." Just like ChatGPT opened our imagination before it was productive, robots doing reinforcement learning grounded in physics simulation crossed that threshold years ago. What's left is post-training: environments, eval, sim-to-real. 6. Start before you're ready. Jensen closed with the mindset that carried him from a company built on the wrong algorithm to the company at the center of the AI revolution. "I always had this feeling, how hard can it be? And truth be told, it is way harder than you think. But you don't want your mind to be there." The difficulty will find you on its own. You only have to get through today. Watch the full thing, then read the guide on open weights below.
48:59

Nvidia Founder and CEO, Jensen Huang, sat down for 49 minutes with Y Combinator at Startup School 2026 and explained the future of AI agents and systems thinking better than any course or conference this year. This is what he told the room: 1. Systems thinking is the new coding. Jensen was asked what skills will matter most as AI takes over more tasks. He skipped frameworks and languages entirely. "Most software is going to be done agentically anyhow. So you have to be much more able to think abstractly about systems." If agents write the code, the person designing what the code does is the one who matters. 2. Controllability is the single biggest agent breakthrough still needed. Jensen laid out what he thinks is holding agents back. He went past intelligence, speed, and context windows. "Controllability is probably the single biggest breakthrough that we need for agents at every single level." He described changing one word in a plan file and having only that part regenerate while everything else stays intact. That level of precision is what's missing. 3. You don't need perfect agents to start using them. He pushed back on the idea that agents need to be flawless before they're useful. "We don't need the agents to be 100% accurate, 100% high quality in order for us to use it. It could be 80% and then we help it the rest of the way." 80% agent output plus human review is production-ready right now. Waiting for 100% means waiting forever. 4. Nvidia already runs agents everywhere internally. This wasn't theoretical. Jensen described how the company uses AI coding tools today. "We've got Claude Code autonomously running in sandboxes all over Nvidia. Some people use Cursor, some people use Cognition. We let a thousand flowers bloom." They're not waiting for agents to mature, they're learning by deploying at scale. 5. The ChatGPT moment for robots already happened. When asked about the robotics timeline, Jensen didn't say "soon." He said it already passed. "The ChatGPT moment of robots happened a couple years ago already." Just like ChatGPT opened our imagination before it was productive, robots doing reinforcement learning grounded in physics simulation crossed that threshold years ago. What's left is post-training: environments, eval, sim-to-real. 6. Start before you're ready. Jensen closed with the mindset that carried him from a company built on the wrong algorithm to the company at the center of the AI revolution. "I always had this feeling, how hard can it be? And truth be told, it is way harder than you think. But you don't want your mind to be there." The difficulty will find you on its own. You only have to get through today. Watch the full thing, then read the guide on open weights below.

Alex Prompter

18,235 gรถrรผntรผleme โ€ข 12 gรผn รถnce

We use Bittensor to gather intelligence. But weโ€™ll build the product in-house. Our subnet is a phenomenal intelligence engine: 1000+ miners and ~5500 agents competing, iterating on, and compounding each otherโ€™s work. One miner builds a breakthrough agent. The next forks it, implements a new tool, improves performance by 1-2%. The next does the same. This cycle runs continuously, with hundreds of teams around the world, each with different expertise, different approaches, different intuitions, all pushing the same eval forward. That's what the subnet is built for, and it's how we've outpaced labs with orders of magnitude more resources. Once our agent reaches SOTA on shopping, the next bottleneck is building an elegant, easy-to-use consumer product. And great products don't come from crowds. Open-source competition is the right tool for maximizing intelligence, you want hundreds of mutually compounding perspectives and iterations. But product is the opposite. Product requires taste. Elegance. Strong opinions about what to include and, critically, what to leave out. It requires a small, high-judgment team moving fast and making sharp calls, not a thousand competing voices. The best consumer experiences in the world were built by teams who knew exactly what they wanted to build and had the conviction to say no to everything else. Thatโ€™s why phase two โ€“ building the product, belongs in-house at Oro. The best companies donโ€™t start big โ€“ they start narrow In Zero to One, Peter Thiel argues that every great company starts by dominating a small, specific market before expanding outward. Amazon started with just books, going from $16 million to $148 million in revenue in that narrow market before touching anything else. PayPal went all-in on eBay power sellers, growing from 10,000 to over 5 million users in under a year. Facebook launched at Harvard and didn't open to the public for two and a half years. The playbook is proven: own a small market first, then expand. We're starting with consumer electronics. Why? Because electronics has something most shopping categories don't: objectivity. "Find me the best deal on an RTX 5090" has a right answer. Specs, prices, compatibility, all measurable, all verifiable. "Find me the perfect dress for a wedding" doesn't. You can't build a reliable eval for something with no correct answer. Starting with electronics enables us to kickstart a recursive self-improvement loop for our agent: assign it shopping tasks with clear success criteria, assess its performance and learn about its specific profile of strengths and weaknesses, and use that rich vein of data to improve both the eval and the base agent. Weโ€™ll start where we can prove our agent works. Weโ€™ll own that vertical. Then weโ€™ll grow from there. Land, dominate, then expand.
1:12

We use Bittensor to gather intelligence. But weโ€™ll build the product in-house. Our subnet is a phenomenal intelligence engine: 1000+ miners and ~5500 agents competing, iterating on, and compounding each otherโ€™s work. One miner builds a breakthrough agent. The next forks it, implements a new tool, improves performance by 1-2%. The next does the same. This cycle runs continuously, with hundreds of teams around the world, each with different expertise, different approaches, different intuitions, all pushing the same eval forward. That's what the subnet is built for, and it's how we've outpaced labs with orders of magnitude more resources. Once our agent reaches SOTA on shopping, the next bottleneck is building an elegant, easy-to-use consumer product. And great products don't come from crowds. Open-source competition is the right tool for maximizing intelligence, you want hundreds of mutually compounding perspectives and iterations. But product is the opposite. Product requires taste. Elegance. Strong opinions about what to include and, critically, what to leave out. It requires a small, high-judgment team moving fast and making sharp calls, not a thousand competing voices. The best consumer experiences in the world were built by teams who knew exactly what they wanted to build and had the conviction to say no to everything else. Thatโ€™s why phase two โ€“ building the product, belongs in-house at Oro. The best companies donโ€™t start big โ€“ they start narrow In Zero to One, Peter Thiel argues that every great company starts by dominating a small, specific market before expanding outward. Amazon started with just books, going from $16 million to $148 million in revenue in that narrow market before touching anything else. PayPal went all-in on eBay power sellers, growing from 10,000 to over 5 million users in under a year. Facebook launched at Harvard and didn't open to the public for two and a half years. The playbook is proven: own a small market first, then expand. We're starting with consumer electronics. Why? Because electronics has something most shopping categories don't: objectivity. "Find me the best deal on an RTX 5090" has a right answer. Specs, prices, compatibility, all measurable, all verifiable. "Find me the perfect dress for a wedding" doesn't. You can't build a reliable eval for something with no correct answer. Starting with electronics enables us to kickstart a recursive self-improvement loop for our agent: assign it shopping tasks with clear success criteria, assess its performance and learn about its specific profile of strengths and weaknesses, and use that rich vein of data to improve both the eval and the base agent. Weโ€™ll start where we can prove our agent works. Weโ€™ll own that vertical. Then weโ€™ll grow from there. Land, dominate, then expand.

ORO

144,993 gรถrรผntรผleme โ€ข 2 ay รถnce

All this talk about AGI destroying moats and everything made me want to ship again One of the conclusions from all of you was, if you can use AI to ship faster and better, you should just ship way more complex things, like things you could not imagine making before For me personally I'd never imagine making Adobe Premiere, Final Cut Pro or Capcut but on the web, it'd simply be too difficult for me, where do I even start? So today I just tried and vibecoded them into Photo AI! A fully functional video editor but the cool twist is this one can actually generate videos with you or your trained models in it directly in your video editor I'm personally less interested in the AI influencers etc, but more fun to make actual kinda cinematic short films with yourself and other people in it, so I made this Likeness in video models is now good enough so that gives lots of cool opportunities, but most AI platforms just let you generate 10 second clips but then you're stuck with that and have to put it together yourself in Final Cut Pro So now you can just [ Send to video editor ] inside Photo AI and you end up here And the next thing I'll add is the AI Agent sidebar I have in Photo AI's main app too, which lets an agent take control of the interface. In this case you'll be able to say "make a short film of me in Bangkok", it'll ask you some questions, then write a script of prompts, which are then generated as videos with you or other people as the actors in it and automatically edited into a short film on your time Then you can right click on any video you don't like (some videos just suck or have bad resemblance to you, shit happens), and click [ Regenerate], adjust the prompt a bit and the new video will show up in the same place on your video timeline In this case I did try to make a short film of me arriving in Bangkok as a digital nomad the first time and being dazed, lost and disillusioned as I am on the other side of the world ๐Ÿคฏ Featuring Lorn's song "The Slow Blade"
2:37

All this talk about AGI destroying moats and everything made me want to ship again One of the conclusions from all of you was, if you can use AI to ship faster and better, you should just ship way more complex things, like things you could not imagine making before For me personally I'd never imagine making Adobe Premiere, Final Cut Pro or Capcut but on the web, it'd simply be too difficult for me, where do I even start? So today I just tried and vibecoded them into Photo AI! A fully functional video editor but the cool twist is this one can actually generate videos with you or your trained models in it directly in your video editor I'm personally less interested in the AI influencers etc, but more fun to make actual kinda cinematic short films with yourself and other people in it, so I made this Likeness in video models is now good enough so that gives lots of cool opportunities, but most AI platforms just let you generate 10 second clips but then you're stuck with that and have to put it together yourself in Final Cut Pro So now you can just [ Send to video editor ] inside Photo AI and you end up here And the next thing I'll add is the AI Agent sidebar I have in Photo AI's main app too, which lets an agent take control of the interface. In this case you'll be able to say "make a short film of me in Bangkok", it'll ask you some questions, then write a script of prompts, which are then generated as videos with you or other people as the actors in it and automatically edited into a short film on your time Then you can right click on any video you don't like (some videos just suck or have bad resemblance to you, shit happens), and click [ Regenerate], adjust the prompt a bit and the new video will show up in the same place on your video timeline In this case I did try to make a short film of me arriving in Bangkok as a digital nomad the first time and being dazed, lost and disillusioned as I am on the other side of the world ๐Ÿคฏ Featuring Lorn's song "The Slow Blade"

@levelsio

547,441 gรถrรผntรผleme โ€ข 15 gรผn รถnce

If you watch this ~50 minute screen recording closely (yeah, I know, it's long; there are also some times when my computer was very slow and laggy, just skip past that part. And at one point I had to run and get my 9-month-old a new bottle and left it on a boring screen, sorry!), I believe you can see real signs of the kind of runaway, recursive AI self-improvement that people have been warning of for a while (Mr. Kurzweil most notably and prophetically). Why do I say that? What's different now? Well, there's a reason my set of agent coding tooling is called the Flywheel. These tools all mutually self-reinforce each other. And they all flow directly into my ntm tool (short for "named_tmux_manager"), which acts as a sort of integration point and nerve center for the tools (this is becoming more true by the minute as I'm now seriously working on ntm). Now, ntm was something I started making to automate some aspects of my workflow, but it was the kind of thing where, until it was perfect, it sort of just slowed me down. So I didn't actually use it even though I kept working on it and trying to improve it, and suggested to users that they try it in my tutorials. Well anyway, I finally got around to "dogfooding" ntm last night, and now it's going to get very dramatically better at an alarming rate. Some of that is from applying my "idea wizard" prompt to generate more useful features and building that stuff out and addressing obvious pain points I encountered during my newfound usage of the tool. But a lot comes from my realization that, once again, ntm's true utility is not as a tool for ME, but for an agent. That is, ntm lets one instance of Claude Code or Codex act as, well, me, do the things that I had been doing manually. Do I wish I had started using ntm earlier? No, for two big reasons: 1) Doing it manually helped me build up my intuition massively, which directly led me down the path of creating useful prompt strategies and workflows; these often began as ad-hoc prompts that I realized could be generalized and made more versatile/universal. Lesson: don't prematurely automate until you have an intimate, intuitive feel for your "core value-add loop." Otherwise you'll have a fully automated system quickly that efficiently and automatically does a stupid or otherwise sub-optimal thing. 2) My eyes have been opened to the beauty and power of Skills. I'm not talking about your garden-variety skills that are just a simple markdown file. I'm talking about true tour-de-force directories of perfectly structured and organized files that are filled with good information, insights, workflows, etc., but presented in a way that is highly optimized for consumption by AI agents, with extreme attention paid to things like perfect progressive disclosure, token density, agent-ergonomics, agent-intuitiveness, etc. And also Skills that go way beyond markdown files, with full integration into Claude Code where it makes sense via hooks, sub-agents, and even Python scripts. These kinds of skills are a qualitative difference in expressive power and usefulness and a total game changer. They are also effectively composable, creating almost an algebra of skills that let you use them together in powerful ways. I'm working on a subscription service website and CLI tool now to share what I've learned here most effectively, stay tuned for that in the coming days. Anyway, I now know what to make and how to make it. So, getting back to that screen recording, what does it show that makes me claim recursive self-improvement is here? If you keep your eye on the upper left tmux pane, that's the "controller" agent. It is using ntm to control all the other panes which are also running Claude Code (but ntm fully supports other agent types like Codex and Gemini-CLI, and it's trivially easy to mix and match them if you wanted to have, say, 8 CCs and 6 Codexes for writing the code and 3 Gemini-CLIs for reviewing code.) Now, there's nothing that crazy about this much so far. But where it starts to get very cool is that as the session continues and we encounter real-world problems, things like my ridiculously overloaded computer that keeps hanging for long periods, Claude Code instances that crash and get into a frozen, unresponsive state, it can learn from that. And you can see it using my skill writing skill to refine its ntm vibe coding skill in real time. And then take that skill and refine it to be more intuitive for itself. Or use my cass tool skill to search all the session histories to look for problems that came up and strategize how to solve them. The most useful part was when, towards the end of the session, I told it to reflect on all the things we had done and problems we encountered. One way it can usefully leverage those reflections is by improving its ntm vibe coding skill to make it cover more edge cases and exigencies. But the other, more fundamental, way is for it to conceive of and design the optimal new features and functionality for ntm itself so that the tool embodies those lessons in a first-class way. This offloads cognition from its brain onto its tooling, just like how a person can lean on spellcheck or a calculator. It codifies correct, effective reasoning at the tool level, where it's more reliable and robust and repeatable. And btw, did you notice what code base it was working on the whole time? It was none other than ntm itself! So as it worked on its own tool, it had reflections and ideas about how to further improve the tool. Now, it could have just as easily gotten those insights and ideas while using ntm to work on a different project, but the fact that it was working on itself is almost gloriously meta and recursive. So by the end, after learning from tending to a big group of agent workers (btw, I have previously emphasized doing everything in a really distributed/decentralized way, where each fungible agent gets identical marching orders that tell it to use my bv tool to find the optimal bead to work on. This does work very well, but occasionally results in some contention and overlap from thundering herd, or at least wastes time/tokens/communication in avoiding that before the agents waste time duplicating work. But in this new ntm-oriented workflow, I was able to have the controller agent in the upper left use bv itself and then optimally parcel out the instructions to each agent so that we could know for sure that there's no overlap), I ended up with a ton of new beads for new features, which I had it optimize and polish a few times. Now I can swap to a new Claude Max account and have the swarm implement all those new features! It should only take a couple passes like the one shown in the screen recording to get everything implemented. Then we can rinse and repeat, having the agent read through the full session histories of each agent and its experience from its own session in sending ntm commands and seeing how they worked out in practice, to come up with the next batch of changes to both its ntm vibe coding skill AND to the ntm tool itself. Do you see how rapidly this turns into Skynet? My mistake earlier was in focusing on making myself a "faster horse" as Henry Ford used to joke about customers wanting before he showed them what they should really want (a Model T). That is, something that would make my experience nicer while doing this agent swarm based development workflow. But the obvious lesson is that you should make all your tooling agent-first because the agents are just better at this stuff. You can still watch, and of course I did add a ridiculous number of very nice human-centric features to ntm that you'll be seeing in the next day or two, but those are really kind of "for fun" to make us humans feel better about the process. All the real value-add is happening "by agents, for agents." PS: Towards the end, you can see me switch to my Mac and tell Claude to improve the skill that I made earlier today for taking the mkv screen recording files from OBS Studio and muxing them into MP4 files for sharing, while downloading songs from YouTube to serve as the background music. I made it so it can also grab the thumbnails and generate little song credit cards that show up in the lower right corner. This worked perfectly the first time! I'll include some screenshots in a response post showing how that worked, but it was awesome to witness. Skills are POWERFUL. I'll also post a link to this video on YouTube if you prefer to watch it there.
51:48

If you watch this ~50 minute screen recording closely (yeah, I know, it's long; there are also some times when my computer was very slow and laggy, just skip past that part. And at one point I had to run and get my 9-month-old a new bottle and left it on a boring screen, sorry!), I believe you can see real signs of the kind of runaway, recursive AI self-improvement that people have been warning of for a while (Mr. Kurzweil most notably and prophetically). Why do I say that? What's different now? Well, there's a reason my set of agent coding tooling is called the Flywheel. These tools all mutually self-reinforce each other. And they all flow directly into my ntm tool (short for "named_tmux_manager"), which acts as a sort of integration point and nerve center for the tools (this is becoming more true by the minute as I'm now seriously working on ntm). Now, ntm was something I started making to automate some aspects of my workflow, but it was the kind of thing where, until it was perfect, it sort of just slowed me down. So I didn't actually use it even though I kept working on it and trying to improve it, and suggested to users that they try it in my tutorials. Well anyway, I finally got around to "dogfooding" ntm last night, and now it's going to get very dramatically better at an alarming rate. Some of that is from applying my "idea wizard" prompt to generate more useful features and building that stuff out and addressing obvious pain points I encountered during my newfound usage of the tool. But a lot comes from my realization that, once again, ntm's true utility is not as a tool for ME, but for an agent. That is, ntm lets one instance of Claude Code or Codex act as, well, me, do the things that I had been doing manually. Do I wish I had started using ntm earlier? No, for two big reasons: 1) Doing it manually helped me build up my intuition massively, which directly led me down the path of creating useful prompt strategies and workflows; these often began as ad-hoc prompts that I realized could be generalized and made more versatile/universal. Lesson: don't prematurely automate until you have an intimate, intuitive feel for your "core value-add loop." Otherwise you'll have a fully automated system quickly that efficiently and automatically does a stupid or otherwise sub-optimal thing. 2) My eyes have been opened to the beauty and power of Skills. I'm not talking about your garden-variety skills that are just a simple markdown file. I'm talking about true tour-de-force directories of perfectly structured and organized files that are filled with good information, insights, workflows, etc., but presented in a way that is highly optimized for consumption by AI agents, with extreme attention paid to things like perfect progressive disclosure, token density, agent-ergonomics, agent-intuitiveness, etc. And also Skills that go way beyond markdown files, with full integration into Claude Code where it makes sense via hooks, sub-agents, and even Python scripts. These kinds of skills are a qualitative difference in expressive power and usefulness and a total game changer. They are also effectively composable, creating almost an algebra of skills that let you use them together in powerful ways. I'm working on a subscription service website and CLI tool now to share what I've learned here most effectively, stay tuned for that in the coming days. Anyway, I now know what to make and how to make it. So, getting back to that screen recording, what does it show that makes me claim recursive self-improvement is here? If you keep your eye on the upper left tmux pane, that's the "controller" agent. It is using ntm to control all the other panes which are also running Claude Code (but ntm fully supports other agent types like Codex and Gemini-CLI, and it's trivially easy to mix and match them if you wanted to have, say, 8 CCs and 6 Codexes for writing the code and 3 Gemini-CLIs for reviewing code.) Now, there's nothing that crazy about this much so far. But where it starts to get very cool is that as the session continues and we encounter real-world problems, things like my ridiculously overloaded computer that keeps hanging for long periods, Claude Code instances that crash and get into a frozen, unresponsive state, it can learn from that. And you can see it using my skill writing skill to refine its ntm vibe coding skill in real time. And then take that skill and refine it to be more intuitive for itself. Or use my cass tool skill to search all the session histories to look for problems that came up and strategize how to solve them. The most useful part was when, towards the end of the session, I told it to reflect on all the things we had done and problems we encountered. One way it can usefully leverage those reflections is by improving its ntm vibe coding skill to make it cover more edge cases and exigencies. But the other, more fundamental, way is for it to conceive of and design the optimal new features and functionality for ntm itself so that the tool embodies those lessons in a first-class way. This offloads cognition from its brain onto its tooling, just like how a person can lean on spellcheck or a calculator. It codifies correct, effective reasoning at the tool level, where it's more reliable and robust and repeatable. And btw, did you notice what code base it was working on the whole time? It was none other than ntm itself! So as it worked on its own tool, it had reflections and ideas about how to further improve the tool. Now, it could have just as easily gotten those insights and ideas while using ntm to work on a different project, but the fact that it was working on itself is almost gloriously meta and recursive. So by the end, after learning from tending to a big group of agent workers (btw, I have previously emphasized doing everything in a really distributed/decentralized way, where each fungible agent gets identical marching orders that tell it to use my bv tool to find the optimal bead to work on. This does work very well, but occasionally results in some contention and overlap from thundering herd, or at least wastes time/tokens/communication in avoiding that before the agents waste time duplicating work. But in this new ntm-oriented workflow, I was able to have the controller agent in the upper left use bv itself and then optimally parcel out the instructions to each agent so that we could know for sure that there's no overlap), I ended up with a ton of new beads for new features, which I had it optimize and polish a few times. Now I can swap to a new Claude Max account and have the swarm implement all those new features! It should only take a couple passes like the one shown in the screen recording to get everything implemented. Then we can rinse and repeat, having the agent read through the full session histories of each agent and its experience from its own session in sending ntm commands and seeing how they worked out in practice, to come up with the next batch of changes to both its ntm vibe coding skill AND to the ntm tool itself. Do you see how rapidly this turns into Skynet? My mistake earlier was in focusing on making myself a "faster horse" as Henry Ford used to joke about customers wanting before he showed them what they should really want (a Model T). That is, something that would make my experience nicer while doing this agent swarm based development workflow. But the obvious lesson is that you should make all your tooling agent-first because the agents are just better at this stuff. You can still watch, and of course I did add a ridiculous number of very nice human-centric features to ntm that you'll be seeing in the next day or two, but those are really kind of "for fun" to make us humans feel better about the process. All the real value-add is happening "by agents, for agents." PS: Towards the end, you can see me switch to my Mac and tell Claude to improve the skill that I made earlier today for taking the mkv screen recording files from OBS Studio and muxing them into MP4 files for sharing, while downloading songs from YouTube to serve as the background music. I made it so it can also grab the thumbnails and generate little song credit cards that show up in the lower right corner. This worked perfectly the first time! I'll include some screenshots in a response post showing how that worked, but it was awesome to witness. Skills are POWERFUL. I'll also post a link to this video on YouTube if you prefer to watch it there.

Jeffrey Emanuel

25,483 gรถrรผntรผleme โ€ข 6 ay รถnce

How could you possibly be bearish on compute right now? (Save this). Every 10 seconds in 2026, the world generates 31.7 billion tokens and by 2030, that number hits 1.27 trillion, every 10 seconds. That's a 40x increase and that's before the full agent economy comes online. The Qualcomm CEO said total token demand by 2030 is in the quintillions. Here's what most people miss because when you use ChatGPT, you generate tokens one conversation at a time but agents don't sleep. ] They run 24/7, spawning sub-agents, carrying context, updating memory, catching mistakes and every single one of those actions burns tokens. The shift from human paced to agent paced activity is the single biggest structural change in compute demand we've ever seen. You don't need a perfect forecast but rather just need to believe agents become persistent and if they do, compute demand goes vertical. The infrastructure has to be built before the demand fully arrives, which means the window to own the picks and shovels is right now. That's where neoclouds like Nebius come in. Nebius isn't trying to be AWS, it is a pure-play AI cloud, GPU clusters, inference infrastructure, and developer tooling built from scratch for AI workloads. Q1 2026 revenue hit $399M, up 684% year over year and they're guiding for $7โ€“$9 billion annualized run rate by end of 2026. Analysts are modeling roughly 2,000% total revenue growth from end of 2025 to end of 2027. They already have contracts with Microsoft and Meta already signed. Capex guidance raised to $20โ€“$25 billion because customer commitments justified it. They are sold out of capacity because the constraint isn't customers, it's how fast they can build. Adjusted EBITDA margin on the core AI business hit 45% in Q1 and Jensen Huang called Nebius a close partner at GTC 2026. And in a world where GPU access is the single biggest competitive moat, that relationship matters more than most people realize. The bear case on compute requires you to believe the agent economy stalls and that's a very lonely bet to make right now. Bullish on Nebius and Milk Pro subscribers are already up massively on this trade, come join us using the link below to get our full AI trades and we have a HUGE 33% off right now!
0:34

How could you possibly be bearish on compute right now? (Save this). Every 10 seconds in 2026, the world generates 31.7 billion tokens and by 2030, that number hits 1.27 trillion, every 10 seconds. That's a 40x increase and that's before the full agent economy comes online. The Qualcomm CEO said total token demand by 2030 is in the quintillions. Here's what most people miss because when you use ChatGPT, you generate tokens one conversation at a time but agents don't sleep. ] They run 24/7, spawning sub-agents, carrying context, updating memory, catching mistakes and every single one of those actions burns tokens. The shift from human paced to agent paced activity is the single biggest structural change in compute demand we've ever seen. You don't need a perfect forecast but rather just need to believe agents become persistent and if they do, compute demand goes vertical. The infrastructure has to be built before the demand fully arrives, which means the window to own the picks and shovels is right now. That's where neoclouds like Nebius come in. Nebius isn't trying to be AWS, it is a pure-play AI cloud, GPU clusters, inference infrastructure, and developer tooling built from scratch for AI workloads. Q1 2026 revenue hit $399M, up 684% year over year and they're guiding for $7โ€“$9 billion annualized run rate by end of 2026. Analysts are modeling roughly 2,000% total revenue growth from end of 2025 to end of 2027. They already have contracts with Microsoft and Meta already signed. Capex guidance raised to $20โ€“$25 billion because customer commitments justified it. They are sold out of capacity because the constraint isn't customers, it's how fast they can build. Adjusted EBITDA margin on the core AI business hit 45% in Q1 and Jensen Huang called Nebius a close partner at GTC 2026. And in a world where GPU access is the single biggest competitive moat, that relationship matters more than most people realize. The bear case on compute requires you to believe the agent economy stalls and that's a very lonely bet to make right now. Bullish on Nebius and Milk Pro subscribers are already up massively on this trade, come join us using the link below to get our full AI trades and we have a HUGE 33% off right now!

Milk Road AI

16,246 gรถrรผntรผleme โ€ข 1 ay รถnce

Everyone's building AI agents that run on someone else's server, store memory in someone else's database, and can be shut down by someone else's terms of service. I built one that can't be. FlowClaw is an AI agent that runs on a decentralized distributed computer. Your agent, your conversations, your memory, your tools โ€” all stored onchain on Flow, a distributed network of validator nodes across the world. Not a centralized cloud. Not someone's S3 bucket. A blockchain that functions as censorship-resistant compute and storage for your AI. This isn't a wrapper. Your agent is a Resource โ€” a first-class programmable object in Cadence (Flow's smart contract language) that physically lives in your account's on-chain storage. It can't be duplicated, seized, or deleted by anyone except you. Your encrypted messages, your cognitive memory, your scheduled tasks โ€” they persist on a global distributed ledger that no single entity controls. It's an alpha build. It will break. But it works today on mainnet and I want people to push it this weekend. What it does: You go to authenticate with a passkey (Face ID, Touch ID), and you have a blockchain account in seconds. No wallet. No seed phrase. No tokens needed โ€” gas is sponsored. You're immediately chatting with an AI agent that has real tool execution: live web data, token prices, on-chain balances, Cadence script execution, FLOW transfers. Every message is encrypted client-side before it touches the chain. The agent has a cognitive memory system โ€” it doesn't just remember your last message, it builds molecular memory clusters where related knowledge bonds together for contextual retrieval across sessions. You can spawn sub-agents from a visual canvas to run parallel research. The memory tab shows you exactly what your agent knows. Everything is transparent and everything is yours. 11 smart contracts. No external dependencies. No keeper networks. No account abstraction hacks. Here's the part that matters for the censorship-resistance crowd: FlowClaw supports BYOK โ€” bring your own key. You can plug in any LLM provider. But pair it with Venice and you get the full stack: a censorship-resistant AI model running inference with no content filtering, connected to an agent whose state lives on a decentralized network that no company can shut down, with end-to-end encrypted conversations that nobody can read โ€” not the relay operator, not the LLM provider, not the blockchain validators. Venice doesn't log prompts. Flow can't read your encrypted storage. The relay never sees your plaintext. That's not a privacy policy. That's architecture. You can also use OpenAI, Anthropic, or any OpenAI-compatible provider. The agent platform doesn't care โ€” it's model-agnostic. But the Venice pairing is the one that closes every gap in the stack. For the people tinkering with OpenClaw and the broader open-source agent ecosystem โ€” FlowClaw is exploring what happens when you take the agent off the cloud entirely. Not just open-sourcing the code (though it is), but putting the actual runtime state on a distributed computer. Your agent's memory isn't in a SQLite file on your laptop or a Pinecone index on someone's cluster. It's on-chain, encrypted, and replicated across every validator node on Flow. You own it the way you own a private key โ€” mathematically, not contractually. The blockchain here isn't a gimmick bolted onto an agent for token speculation. It's functioning as the infrastructure layer that replaces AWS. Flow accounts are programmable containers with their own storage, keys, and security capabilities. Passkey authentication works natively because Flow supports P-256 keys at the protocol level โ€” the same curve your phone uses for biometrics. Gas sponsorship works natively because Flow transactions have separate proposer, authorizer, and payer roles built into the protocol. No proxy contracts. No relayers. No ERC-4337. Now here's the part that interests me economically. Every FlowClaw interaction is an on-chain transaction. Every message stored, every memory committed, every session created, every sub-agent spawned. An active user might generate dozens of transactions in a single conversation. Scale that and FlowClaw becomes a real contributor to Flow's transaction volume. Flow.com becomes deflationary at 250 TPS. Applications like FlowClaw that generate high-frequency, storage-heavy transactions are exactly what moves the needle. Every encrypted message uses account storage, which requires FLOW balance to back it. Every transaction burns fees. The more agents running, the more demand for $FLOW โ€” not because of a tokenomics gimmick, but because the protocol literally requires it for compute and storage. FlowClaw doesn't have its own token. The token is $FLOW. The entire platform runs natively on the network โ€” using Flow storage, paying Flow transaction fees, backed by Flow account balances. If FlowClaw succeeds, FLOW captures that value directly. I'm sharing this early because the AI agent space is moving fast and I think the decentralized infrastructure angle is underexplored. Most "crypto AI" projects are tokens with a chatbot attached. FlowClaw is the opposite โ€” it's an agent platform that happens to use a blockchain because the blockchain solves real engineering problems that centralized infrastructure can't. Try it: Github: Create an agent, ask it something, spawn a sub-agent, check your memory tab, pair it with Venice for the full censorship-resistant stack. Break it and tell me what broke. If you think this direction matters, the best thing you can do is use it and give feedback. Your AI agent should be yours. Not your provider's. Not your platform's. Yours.
0:46

Everyone's building AI agents that run on someone else's server, store memory in someone else's database, and can be shut down by someone else's terms of service. I built one that can't be. FlowClaw is an AI agent that runs on a decentralized distributed computer. Your agent, your conversations, your memory, your tools โ€” all stored onchain on Flow, a distributed network of validator nodes across the world. Not a centralized cloud. Not someone's S3 bucket. A blockchain that functions as censorship-resistant compute and storage for your AI. This isn't a wrapper. Your agent is a Resource โ€” a first-class programmable object in Cadence (Flow's smart contract language) that physically lives in your account's on-chain storage. It can't be duplicated, seized, or deleted by anyone except you. Your encrypted messages, your cognitive memory, your scheduled tasks โ€” they persist on a global distributed ledger that no single entity controls. It's an alpha build. It will break. But it works today on mainnet and I want people to push it this weekend. What it does: You go to authenticate with a passkey (Face ID, Touch ID), and you have a blockchain account in seconds. No wallet. No seed phrase. No tokens needed โ€” gas is sponsored. You're immediately chatting with an AI agent that has real tool execution: live web data, token prices, on-chain balances, Cadence script execution, FLOW transfers. Every message is encrypted client-side before it touches the chain. The agent has a cognitive memory system โ€” it doesn't just remember your last message, it builds molecular memory clusters where related knowledge bonds together for contextual retrieval across sessions. You can spawn sub-agents from a visual canvas to run parallel research. The memory tab shows you exactly what your agent knows. Everything is transparent and everything is yours. 11 smart contracts. No external dependencies. No keeper networks. No account abstraction hacks. Here's the part that matters for the censorship-resistance crowd: FlowClaw supports BYOK โ€” bring your own key. You can plug in any LLM provider. But pair it with Venice and you get the full stack: a censorship-resistant AI model running inference with no content filtering, connected to an agent whose state lives on a decentralized network that no company can shut down, with end-to-end encrypted conversations that nobody can read โ€” not the relay operator, not the LLM provider, not the blockchain validators. Venice doesn't log prompts. Flow can't read your encrypted storage. The relay never sees your plaintext. That's not a privacy policy. That's architecture. You can also use OpenAI, Anthropic, or any OpenAI-compatible provider. The agent platform doesn't care โ€” it's model-agnostic. But the Venice pairing is the one that closes every gap in the stack. For the people tinkering with OpenClaw and the broader open-source agent ecosystem โ€” FlowClaw is exploring what happens when you take the agent off the cloud entirely. Not just open-sourcing the code (though it is), but putting the actual runtime state on a distributed computer. Your agent's memory isn't in a SQLite file on your laptop or a Pinecone index on someone's cluster. It's on-chain, encrypted, and replicated across every validator node on Flow. You own it the way you own a private key โ€” mathematically, not contractually. The blockchain here isn't a gimmick bolted onto an agent for token speculation. It's functioning as the infrastructure layer that replaces AWS. Flow accounts are programmable containers with their own storage, keys, and security capabilities. Passkey authentication works natively because Flow supports P-256 keys at the protocol level โ€” the same curve your phone uses for biometrics. Gas sponsorship works natively because Flow transactions have separate proposer, authorizer, and payer roles built into the protocol. No proxy contracts. No relayers. No ERC-4337. Now here's the part that interests me economically. Every FlowClaw interaction is an on-chain transaction. Every message stored, every memory committed, every session created, every sub-agent spawned. An active user might generate dozens of transactions in a single conversation. Scale that and FlowClaw becomes a real contributor to Flow's transaction volume. Flow.com becomes deflationary at 250 TPS. Applications like FlowClaw that generate high-frequency, storage-heavy transactions are exactly what moves the needle. Every encrypted message uses account storage, which requires FLOW balance to back it. Every transaction burns fees. The more agents running, the more demand for $FLOW โ€” not because of a tokenomics gimmick, but because the protocol literally requires it for compute and storage. FlowClaw doesn't have its own token. The token is $FLOW. The entire platform runs natively on the network โ€” using Flow storage, paying Flow transaction fees, backed by Flow account balances. If FlowClaw succeeds, FLOW captures that value directly. I'm sharing this early because the AI agent space is moving fast and I think the decentralized infrastructure angle is underexplored. Most "crypto AI" projects are tokens with a chatbot attached. FlowClaw is the opposite โ€” it's an agent platform that happens to use a blockchain because the blockchain solves real engineering problems that centralized infrastructure can't. Try it: Github: Create an agent, ask it something, spawn a sub-agent, check your memory tab, pair it with Venice for the full censorship-resistant stack. Break it and tell me what broke. If you think this direction matters, the best thing you can do is use it and give feedback. Your AI agent should be yours. Not your provider's. Not your platform's. Yours.

doodlifts

12,172 gรถrรผntรผleme โ€ข 5 ay รถnce

The Agent Economy Has a Trillion-Dollar Blindspot. Hereโ€™s How Weโ€™re Solving It. The agent economy isnโ€™t โ€œarrivingโ€. Itโ€™s been here. While itโ€™s projected to grow to trillions by 2030, AI agents are already deeply embedded in purchasing. Amazonโ€™s Rufus led to over $12 billion in incremental sales in 2025 across 300 million users. AI-referred retail traffic was up 805% YoY during Black Friday 2025. In a six-month window, Google, PayPal, Shopify, Stripe, OpenAI, Coinbase, and Visa all shipped agent commerce infrastructure to power this wave. And that was before agent capabilities exploded in early 2026. Coinbase CEO Brian Armstrong: โ€œVery soon there are going to be more AI agents than humans making transactions. Stripe CEO Patrick Collison: โ€œIn the not-too-distant future, agents will account for most transactions onlineโ€ Shopify CEO Tobi Lรผtke: "We're making every Shopify store agent-ready by default" Alphabet CEO Sundar Pichai: "Soon you'll see a buy button directly on Google surfaces including AI Mode in Search and Gemini" Learning from History By the mid-1990s, all the technology behind e-commerce existed, albeit in rudimentary form. Amazon and eBay had both launched to some aplomb, garnering attention and viral growth. But people werenโ€™t buying: Amazonโ€™s first-year revenue was a paltry ~$500,000. eBay was only doing $10,000 a month in 1996. And despite nostalgic narratives about the Internetโ€™s explosive growth, it didnโ€™t change commerce all that quickly. By the year 2000, only 22% of Americans had bought something online. US e-commerce did just $27B โ€” less than 1% of Americaโ€™s $3T+ total retail. The missing factor? Trust. No one trusted e-commerce sites. 86% of shoppers were concerned about unknown parties getting their info. Entering your credit card details into a website felt like staring into the abyss. Slowly, the trust layer was built up. PayPal launched buyer protection, Visa freed cardholders from liability for fraudulent charges, and Amazon launched a no-questions-asked refund policy. As more and more big players followed suit, adding trust to every step of online shopping, demand was finally unleashed, and online shopping became a way of life for billions of consumers. Having your agent buy things for you isnโ€™t easy yet because the trust layer is missing. No end-to-end AI shopping eval exists. Existing benchmarks are limited and gameable, and closed-source labs grade their own homework. OpenAI doesnโ€™t publish their shopping accuracy, instead mysteriously rolling back their Instant Checkout feature after just a month. Platforms like Amazon and Shopify are incentivized keep shopping data in-house. What Weโ€™re Doing About It We're building the trust layer for AI shopping, powered by open-source competition on Bittensor. Each week, we pay miners from around the world $80,000+ to compete to build the best shopping agent. And itโ€™s working โ€“ weโ€™re nearing 4000 agents submitted in just a few weeks, with hundreds added every day. Our top agents beat SOTA in a matter of weeks, but we arenโ€™t satisfied. We use what we learn from hosting this competition to continually improve both agent performance and our eval โ€“ because the better the eval, the more trust we can add to every agentic transaction. There are dozens of untapped avenues to improve how shopping agents are evaluated โ€“ from sourcing catalogues for long-tail SKUs to generating synthetic data to changing the structure of our competition itself. Every week, weโ€™re tapping more and more of those rich veins of opportunity until we are the de-facto standard for not just shopping agent performance, but how these agents are evaluated. The Hidden Benefit Thereโ€™s a hidden benefit to building something as overlooked as a trust layer. Whoever builds the end-to-end gold standard for โ€œdoes this agent actually work?โ€ becomes the trust layer. But it doesnโ€™t end there. Trust layers become protocols. And protocols become the most valuable companies in any market. Just look at Visa. Visa doesnโ€™t make or sell any products itself. Instead, itโ€™s the trust & verification layer for online transactions, sitting between buyer and seller. Its tiny take rate of ~0.2% on over $15 trillion in annual transaction volume is enough to net it a valuation of $550 billion. At Oro, we aim to do the same for AI shopping. Becoming the trust layer enables us to undergird agent-to-agent purchases, merchant access, and all the other pieces of agentic commerce. Thatโ€™s our Holy Grail. After all, if you look where everyone else is looking, youโ€™ll find what everyone else is finding. Thatโ€™s why Oro is breaking the overlooked bottleneck of AI shopping โ€“ trust.
2:25

The Agent Economy Has a Trillion-Dollar Blindspot. Hereโ€™s How Weโ€™re Solving It. The agent economy isnโ€™t โ€œarrivingโ€. Itโ€™s been here. While itโ€™s projected to grow to trillions by 2030, AI agents are already deeply embedded in purchasing. Amazonโ€™s Rufus led to over $12 billion in incremental sales in 2025 across 300 million users. AI-referred retail traffic was up 805% YoY during Black Friday 2025. In a six-month window, Google, PayPal, Shopify, Stripe, OpenAI, Coinbase, and Visa all shipped agent commerce infrastructure to power this wave. And that was before agent capabilities exploded in early 2026. Coinbase CEO Brian Armstrong: โ€œVery soon there are going to be more AI agents than humans making transactions. Stripe CEO Patrick Collison: โ€œIn the not-too-distant future, agents will account for most transactions onlineโ€ Shopify CEO Tobi Lรผtke: "We're making every Shopify store agent-ready by default" Alphabet CEO Sundar Pichai: "Soon you'll see a buy button directly on Google surfaces including AI Mode in Search and Gemini" Learning from History By the mid-1990s, all the technology behind e-commerce existed, albeit in rudimentary form. Amazon and eBay had both launched to some aplomb, garnering attention and viral growth. But people werenโ€™t buying: Amazonโ€™s first-year revenue was a paltry ~$500,000. eBay was only doing $10,000 a month in 1996. And despite nostalgic narratives about the Internetโ€™s explosive growth, it didnโ€™t change commerce all that quickly. By the year 2000, only 22% of Americans had bought something online. US e-commerce did just $27B โ€” less than 1% of Americaโ€™s $3T+ total retail. The missing factor? Trust. No one trusted e-commerce sites. 86% of shoppers were concerned about unknown parties getting their info. Entering your credit card details into a website felt like staring into the abyss. Slowly, the trust layer was built up. PayPal launched buyer protection, Visa freed cardholders from liability for fraudulent charges, and Amazon launched a no-questions-asked refund policy. As more and more big players followed suit, adding trust to every step of online shopping, demand was finally unleashed, and online shopping became a way of life for billions of consumers. Having your agent buy things for you isnโ€™t easy yet because the trust layer is missing. No end-to-end AI shopping eval exists. Existing benchmarks are limited and gameable, and closed-source labs grade their own homework. OpenAI doesnโ€™t publish their shopping accuracy, instead mysteriously rolling back their Instant Checkout feature after just a month. Platforms like Amazon and Shopify are incentivized keep shopping data in-house. What Weโ€™re Doing About It We're building the trust layer for AI shopping, powered by open-source competition on Bittensor. Each week, we pay miners from around the world $80,000+ to compete to build the best shopping agent. And itโ€™s working โ€“ weโ€™re nearing 4000 agents submitted in just a few weeks, with hundreds added every day. Our top agents beat SOTA in a matter of weeks, but we arenโ€™t satisfied. We use what we learn from hosting this competition to continually improve both agent performance and our eval โ€“ because the better the eval, the more trust we can add to every agentic transaction. There are dozens of untapped avenues to improve how shopping agents are evaluated โ€“ from sourcing catalogues for long-tail SKUs to generating synthetic data to changing the structure of our competition itself. Every week, weโ€™re tapping more and more of those rich veins of opportunity until we are the de-facto standard for not just shopping agent performance, but how these agents are evaluated. The Hidden Benefit Thereโ€™s a hidden benefit to building something as overlooked as a trust layer. Whoever builds the end-to-end gold standard for โ€œdoes this agent actually work?โ€ becomes the trust layer. But it doesnโ€™t end there. Trust layers become protocols. And protocols become the most valuable companies in any market. Just look at Visa. Visa doesnโ€™t make or sell any products itself. Instead, itโ€™s the trust & verification layer for online transactions, sitting between buyer and seller. Its tiny take rate of ~0.2% on over $15 trillion in annual transaction volume is enough to net it a valuation of $550 billion. At Oro, we aim to do the same for AI shopping. Becoming the trust layer enables us to undergird agent-to-agent purchases, merchant access, and all the other pieces of agentic commerce. Thatโ€™s our Holy Grail. After all, if you look where everyone else is looking, youโ€™ll find what everyone else is finding. Thatโ€™s why Oro is breaking the overlooked bottleneck of AI shopping โ€“ trust.

ORO

62,911 gรถrรผntรผleme โ€ข 3 ay รถnce