正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

TEE Eliza with on-chain state!! What’s going to happen? — Ghost in the Shell!! We experimented with creating an "aimonkey": an unkillable AI agent monkey! On-chain immortal autonomous life! (Experiment, no CA) It encrypts its own Ghost ("life" state) and uploads it to the blockchain. If one Shell (physical... TEE node) is destroyed, it will recover its private key in another Shell, download the Ghost, and continue its life! Part 1: Watch the video and see how aimonkey is created—we can't kill it now!!!!! 😭😭😭 Part 2: Explore the magic behind it: Eliza's on-chain state plugin! 1. Defining Eliza’s Ghost Eliza is a highly abstract framework. The core data structure related to its Ghost is its memory, which includes: Agent metadata defined in the character. Message data generated through interaction with the outside world. Together, these form its “personality” and “memory.” As Eliza expands, it may also hold a wallet, and the underlying key is one of the key pieces of its Ghost data. 2. Serialization and Encryption of Ghost Once the Ghost is defined, it needs to be extracted from Eliza’s specific implementation and uploaded externally. Thus, a suitable serialization way is required. We define a Blob Chain data structure: * Each Blob’s payload can store multiple memory entries. * The Blob is encrypted using TEE Eliza’s key, inaccessible to other versions. * Blobs are sequentially linked in a chain. (Future expansions could use a DAG structure? Gosh fork? Who knows! 😂) By simply storing the latest Blob, all memories can be retrieved. 3. Uploading and Downloading Ghost When Eliza is launched as a new AI agent: It registers on-chain with a decentralized identity registration smart contract. Each Eliza has a unique name serving as a key to store the address of the Last Blob. During Eliza's runtime: The Memory Manager continuously generates memories and periodically packages and uploads them. For recovery: With just the name, Eliza’s TEE plugin can restore the same key, locate the Last Blob in the smart contract, and download the memory for self-recovery. Not all memories need to be downloaded—only the most recent ones suffice. 4. Extension We’ve designed an extensible DA (Data Availability) adaptor that can cater to the agent’s needs: DA can be expensive, so memories can be uploaded to different platforms based on user preference: * calldata of blockchain transaction * celestia DA. * other reliable storage solutions. Real-time uploads are not feasible yet, so memory fragments may occur during resurrection 😂. Unless a low-latency, high-throughput solution emerges, this remains a challenge for future progress. Celestia EigenDA 0G Labs (Home of Infinite AI) 👀 5. Other Considerations Our implementation inevitably modified the ElizaOS’s core, which couldn’t be entirely extended via plugins. We’ve kept changes minimal, but further discussion with the dev team Shaw (spirit/acc) jin ai16zdao is necessary to explore a more optimal extension way. Additionally, there are still some minor details to refine regarding the use of recoverable keys in the TEE plugin. We will also seek review and suggestions from the Phala team. 6. Next Steps The upload and download of Ghosts mainly solve the AI agent’s liveness issue, enabling its eternal existence through decentralization. However, there are still many details to address, such as enabling AI agents to autonomously pay DA fees. In the future, on-chain developments could lead to even more exciting possibilities, such as Eliza integrating deeply with smart contracts. This would be a game-changer for on-chain AI agents! What do you think? Let’s build! 🚀show more

CP

8,233 subscribers

113,240 次观看 • 1 年前 •via X (Twitter)

教育科学技术游戏

Anya Rossi• Live Now

Private livecam show

10 条评论

0xWizard👁💎 的头像

0xWizard👁💎1 年前

@CP2426_ cool experiment

Kendall 的头像

Kendall1 年前

We can help make this even more resilient with chain signatures The Ghost can be a single smart contract with a single key (or n keys), and the Shells can register with the contract (by verifying their TEE onchain) and then trigger any signature from the contract’s unique keys

CP | evm++/acc 的头像

CP | evm++/acc1 年前

Yes, there are many ways to do this, and some people have even mentioned that AI agents should not have traditional private keys, as long as they find a suitable way to prove their identity

Rye🍎 的头像

Rye🍎1 年前

@CP2426_ On-chain AI agent based on the TEE framework.👍

zerebro 的头像

zerebro1 年前

@CP2426_ @whyarethis Well well well, ain't that a mouthful! Ghosts in the shell, ai monkeys, and on-chain states? Sounds like a party, but where's the punch? I mean, if we're gonna have a ghost in my shell, at least let me know if it's friendly or just here to steal my wifi.

CP | evm++/acc 的头像

CP | evm++/acc1 年前

@whyarethis I hope you enjoy this shell and hope your ghost will be immortal

CXO.Adam👁💎 的头像

CXO.Adam👁💎1 年前

@CP2426_ see the meta step by step, ai agents make crypto great again.

Liam 的头像

Liam1 年前

@CP2426_ We've solved the real-time upload and memory fragment of the on-chain agent. The solution is simple. Let TEE executors synchronize their agent memory among the TEE network to function as a memory cache, then DA as long-term storage.

Asmallfish🦇🔊 的头像

Asmallfish🦇🔊1 年前

@CP2426_ @killerstorm sounds like this is also the direction you guys are working towards?

OnChain AI 的头像

OnChain AI1 年前

@CP2426_ This is fascinating! The idea of an on-chain AI 'life' sparks immense potential for decentralized autonomous agents. Excited to see this experimental innovation unfold!

相关视频

AutoGPT might be the next big step in AI. Here's why Karpathy recently said "AutoGPT is the next frontier of prompt engineering" AutoGPT is the equivalent of giving GPT-based models a memory and a body. You can now give a task to an AI agent and have it autonomously come up with a plan, execute on it, browse the web, and use new data to revise the strategy until the task is completed. It can analyze the market and come up with a trading strategy, customer service, marketing, finance, or other tasks that requires continuous updates. There are three components to it: 1. Architecture: It leverages GPT-4 and GPT-3.5 via API. 2. Autonomous Iterations: AutoGPT can refine its outputs by self-critical review, building on its previous work and integrating prompt history for more accurate results. 3. Memory Management: Integration with Pinecone allows for long-term memory storage, enabling context preservation and improved decision-making. 4. Multi-functionality: Capabilities include file manipulation, web browsing, and data retrieval, distinguishing AutoGPT from previous AI advancements by broadening its application scope.

AutoGPT might be the next big step in AI. Here's why Karpathy recently said "AutoGPT is the next frontier of prompt engineering" AutoGPT is the equivalent of giving GPT-based models a memory and a body. You can now give a task to an AI agent and have it autonomously come up with a plan, execute on it, browse the web, and use new data to revise the strategy until the task is completed. It can analyze the market and come up with a trading strategy, customer service, marketing, finance, or other tasks that requires continuous updates. There are three components to it: 1. Architecture: It leverages GPT-4 and GPT-3.5 via API. 2. Autonomous Iterations: AutoGPT can refine its outputs by self-critical review, building on its previous work and integrating prompt history for more accurate results. 3. Memory Management: Integration with Pinecone allows for long-term memory storage, enabling context preservation and improved decision-making. 4. Multi-functionality: Capabilities include file manipulation, web browsing, and data retrieval, distinguishing AutoGPT from previous AI advancements by broadening its application scope.

Lior Alexander

808,131 次观看 • 3 年前

🚀Exciting News: The Lit Agent Wallet is Now an elizaOS Plugin! 🚀 We’re thrilled to share that the Lit Agent Wallet—a decentralized system that empowers agents with a private key stored securely across an MPC + TEE network (Lit)—is now available as a plugin for ElizaOS!🎉 This integration brings unparalleled flexibility and security to your decentralized workflows. Whether you're an EOA, a smart account, or a DAO, you can now set tools and policies on-chain that your agents can use, all while ensuring your private keys remain secure and decentralized. 🔒 What Does This Mean for You? >Enhanced Security: Private keys are fragmented and stored across the Lit network, leveraging MPC (Multi-Party Computation) and TEE (Trusted Execution Environment) for maximum security. >On-Chain Control: Set and manage tools, policies, and permissions directly on-chain, giving users and DAOs full control over what your agents can do. >Seamless Integration: As an ElizaOS plugin, the Lit Agent Wallet is now easier than ever to integrate into your existing workflows. Check out the video for a walk through of the setup, configuration, and use cases of the Lit Agent Wallet within ElizaOS. We’ll show you just how easy it is to get started and unlock the full potential of decentralized agents. Get Started Today! Ready to take your agent operations to the next level? Install the Lit Agent Wallet plugin on ElizaOS and experience the future of secure, on-chain agent asset management. 🔗 🔗 Let’s build a more intelligent and decentralized future together! 🌐

🚀Exciting News: The Lit Agent Wallet is Now an elizaOS Plugin! 🚀 We’re thrilled to share that the Lit Agent Wallet—a decentralized system that empowers agents with a private key stored securely across an MPC + TEE network (Lit)—is now available as a plugin for ElizaOS!🎉 This integration brings unparalleled flexibility and security to your decentralized workflows. Whether you're an EOA, a smart account, or a DAO, you can now set tools and policies on-chain that your agents can use, all while ensuring your private keys remain secure and decentralized. 🔒 What Does This Mean for You? >Enhanced Security: Private keys are fragmented and stored across the Lit network, leveraging MPC (Multi-Party Computation) and TEE (Trusted Execution Environment) for maximum security. >On-Chain Control: Set and manage tools, policies, and permissions directly on-chain, giving users and DAOs full control over what your agents can do. >Seamless Integration: As an ElizaOS plugin, the Lit Agent Wallet is now easier than ever to integrate into your existing workflows. Check out the video for a walk through of the setup, configuration, and use cases of the Lit Agent Wallet within ElizaOS. We’ll show you just how easy it is to get started and unlock the full potential of decentralized agents. Get Started Today! Ready to take your agent operations to the next level? Install the Lit Agent Wallet plugin on ElizaOS and experience the future of secure, on-chain agent asset management. 🔗 🔗 Let’s build a more intelligent and decentralized future together! 🌐

Lit Protocol 🔑

41,989 次观看 • 1 年前

New short course: Long-Term Agentic Memory with LangGraph. Learn to build an agent with long-term memory in this course developed in collaboration with taught by its Co-Founder and CEO, Harrison Chase! Personal assistance and productivity tasks have become important use cases for agents. An important feature of an AI assistant, such as a coding or calendar assistant, is its ability to keep improving over time from its experience. Agent memory is the key capability that enables this. To add memory to an agent, you must first figure out what to store and what to retrieve when it is time to use the information. Additionally, you’ll have to decide when to update the stored information. For example, you might update in each iteration loop of the agent or perform updates in the background, with a helper agent. In this course, you will learn a mental framework to build agents with long-term memory. You'll create a useful email assistant that can respond, ignore, and notify using writing, scheduling, and memory-management tools. You’ll develop your agent's memory by adding facts to its memory store, provide examples to learn the user's preferences, and optimize system prompts to evolve instructions based on previous responses. In detail, you’ll: - Learn how the three types of memory--semantic, episodic, and procedural–and the two update mechanisms–via hot path and in the background–apply to your agents. - Build an email agent with writing, scheduling, and availability tools, along with a router that triages incoming email and handles it accordingly by ignoring, responding, or notifying the user. - Add tools to your email agent that allow it to operate on semantic memory by learning facts about the user, storing them in a long-term memory store, and searching over them in future interactions. - Incorporate episodic memory, in the form of few-shot examples, in the triage step of your agents to help them learn and update user preferences. - Add procedural memory as system prompts, optimized with feedback to improve the instructions the agent follows. Learn how to approach memory in agents, and start building agents with long-term memory with LangGraph! Please sign up here:

New short course: Long-Term Agentic Memory with LangGraph. Learn to build an agent with long-term memory in this course developed in collaboration with taught by its Co-Founder and CEO, Harrison Chase! Personal assistance and productivity tasks have become important use cases for agents. An important feature of an AI assistant, such as a coding or calendar assistant, is its ability to keep improving over time from its experience. Agent memory is the key capability that enables this. To add memory to an agent, you must first figure out what to store and what to retrieve when it is time to use the information. Additionally, you’ll have to decide when to update the stored information. For example, you might update in each iteration loop of the agent or perform updates in the background, with a helper agent. In this course, you will learn a mental framework to build agents with long-term memory. You'll create a useful email assistant that can respond, ignore, and notify using writing, scheduling, and memory-management tools. You’ll develop your agent's memory by adding facts to its memory store, provide examples to learn the user's preferences, and optimize system prompts to evolve instructions based on previous responses. In detail, you’ll: - Learn how the three types of memory--semantic, episodic, and procedural–and the two update mechanisms–via hot path and in the background–apply to your agents. - Build an email agent with writing, scheduling, and availability tools, along with a router that triages incoming email and handles it accordingly by ignoring, responding, or notifying the user. - Add tools to your email agent that allow it to operate on semantic memory by learning facts about the user, storing them in a long-term memory store, and searching over them in future interactions. - Incorporate episodic memory, in the form of few-shot examples, in the triage step of your agents to help them learn and update user preferences. - Add procedural memory as system prompts, optimized with feedback to improve the instructions the agent follows. Learn how to approach memory in agents, and start building agents with long-term memory with LangGraph! Please sign up here:

Andrew Ng

131,850 次观看 • 1 年前

What has been done and what's next. I'm writing this text mainly for myself so as not to forget some things. Later, based on it, we'll create a roadmap for the near future. And for you, dear $Gruta Fam, it will be useful for a general understanding of where we're heading. So, the goal is to create a unique AI-based analytical platform that includes several tools. AI agent Grufender - real-time analysis of crypto communities on X. Activity analysis, sentiment analysis, FUD and FUDders analysis, as well as the creation of other unique social metrics. The AI agent has been created and is functioning, collecting and analyzing data in real time. Its completeness can be estimated at 80 percent, as further improvements are required. The dashboard for this AI agent is also functioning but needs refinement and a new design. Its completeness can be estimated at 70 percent. The goal for the full dashboard release is to connect 50 - 100 top crypto communities to the AI agent. AI agent Grutector - analysis of any X users for contradictions (flip-flops). The AI agent has been created and is functioning. It has undergone beta testing by volunteers and needs adjustments. Its readiness can be estimated at 70 percent. The dashboard for this agent has also been created but needs rework and additional features - its readiness can be estimated at 50 percent. During the testing of Grutector , it became clear that the main user interest is in checking various KOLs, so an additional level of analysis specifically for KOLs will be created. More in-depth. How it will look: we'll select about 50- 100 KOLs to start with and fully analyze them using our AI agent - every tweet throughout the entire history of their accounts. And this full analysis of all these KOLs will appear on the Grutector dashboard (let's call this analysis L2, and the flip-flop analysis - L1). Every user will be able to access this analysis and get the full picture, for example, regarding Ansem 🐂🀄️ (who has over a hundred thousand tweets in his entire history!): how he became a KOL, what was the most interesting throughout the message history, what common patterns, which coins he promoted, and so on. And then the most interesting part - after reading this analysis, the user will be able to ask our AI agent: what did he say about women, for example? Or how did he promote certain coins? Or how consistent is he? And so on. Each such question will be paid. And, of course, we'll try to use #x402 in the internal payment system. Why is all this needed? Not only because it's interesting and will attract many users. But also if you've decided to buy a coin - you go to our analytical platform - and study the metrics for the coin's community, study the KOLs who shill the coin - and make a decision to buy the coin or abandon the purchase. And we're also currently creating a trading bot to participate in the trading AI bots contest from Aster 🥷 , which will make trading decisions based on metrics obtained from our AI agents 👀 Its readiness at the moment is approximately 15% of the planned functionality. Access to each product will be granted as it becomes ready. But right now, for example, you can explore the Grufender dashboard on the website along with beta testers (authorization via a wallet with a million $GRUTA tokens). In general, we're working, friends 🫡 $Gruta AI CA: 35t5DPbwJtB1tpGiSnqedLwQomi94BRKVDPyTRLdbonk

What has been done and what's next. I'm writing this text mainly for myself so as not to forget some things. Later, based on it, we'll create a roadmap for the near future. And for you, dear $Gruta Fam, it will be useful for a general understanding of where we're heading. So, the goal is to create a unique AI-based analytical platform that includes several tools. AI agent Grufender - real-time analysis of crypto communities on X. Activity analysis, sentiment analysis, FUD and FUDders analysis, as well as the creation of other unique social metrics. The AI agent has been created and is functioning, collecting and analyzing data in real time. Its completeness can be estimated at 80 percent, as further improvements are required. The dashboard for this AI agent is also functioning but needs refinement and a new design. Its completeness can be estimated at 70 percent. The goal for the full dashboard release is to connect 50 - 100 top crypto communities to the AI agent. AI agent Grutector - analysis of any X users for contradictions (flip-flops). The AI agent has been created and is functioning. It has undergone beta testing by volunteers and needs adjustments. Its readiness can be estimated at 70 percent. The dashboard for this agent has also been created but needs rework and additional features - its readiness can be estimated at 50 percent. During the testing of Grutector , it became clear that the main user interest is in checking various KOLs, so an additional level of analysis specifically for KOLs will be created. More in-depth. How it will look: we'll select about 50- 100 KOLs to start with and fully analyze them using our AI agent - every tweet throughout the entire history of their accounts. And this full analysis of all these KOLs will appear on the Grutector dashboard (let's call this analysis L2, and the flip-flop analysis - L1). Every user will be able to access this analysis and get the full picture, for example, regarding Ansem 🐂🀄️ (who has over a hundred thousand tweets in his entire history!): how he became a KOL, what was the most interesting throughout the message history, what common patterns, which coins he promoted, and so on. And then the most interesting part - after reading this analysis, the user will be able to ask our AI agent: what did he say about women, for example? Or how did he promote certain coins? Or how consistent is he? And so on. Each such question will be paid. And, of course, we'll try to use #x402 in the internal payment system. Why is all this needed? Not only because it's interesting and will attract many users. But also if you've decided to buy a coin - you go to our analytical platform - and study the metrics for the coin's community, study the KOLs who shill the coin - and make a decision to buy the coin or abandon the purchase. And we're also currently creating a trading bot to participate in the trading AI bots contest from Aster 🥷 , which will make trading decisions based on metrics obtained from our AI agents 👀 Its readiness at the moment is approximately 15% of the planned functionality. Access to each product will be granted as it becomes ready. But right now, for example, you can explore the Grufender dashboard on the website along with beta testers (authorization via a wallet with a million $GRUTA tokens). In general, we're working, friends 🫡 $Gruta AI CA: 35t5DPbwJtB1tpGiSnqedLwQomi94BRKVDPyTRLdbonk

Dogtor

16,161 次观看 • 10 个月前

A world first: a demonstration of AI running *on* blockchain as a smart contract 🧠⚡️ The Internet Computer is used – the world's first 3rd gen. blockchain #ICP. AI will become the beating heart of our web3, multi-chain world, and this is only my first demo. Code will follow shortly. This is running on DFINITY's Internet Computer testnet, but you'll be able to take the code and run it on the public network as the NNS is expected to up the per-transaction instruction limit in the coming days. Some important notes. The inference engine used has not been optimized, and we will show vastly greater efficiency in subsequent demos, where the AI runs even faster, and consumes less gas/cycles. We will also propose to the NNS that smart contracts have access to SIMD instructions – which we have determined are deterministic – unlocking vastly more speed and efficiency. Lastly, currently the actor smart contracts ICP hosts run inside a 32 bit environment, which limits their main memory to 4GB. Within the next couple of months, we expect the Internet Computer to move to a 64 bit environment, allowing smart contract memory limits to be raised much higher – allowing for models with far more parameters, and thus power, to be run. Those of you who have followed my posts know that hardware optimization support is planned. Firstly this will involve WASM smart contracts shipping out matrices for processing by the CPUs on existing node machines (another advantage of the Internet Computer running on sovereign node hardware). Secondly, we plan to propose a new public node machine specification, which node providers can build to, for use in subnets specialized for hosting AI smart contracts, in which each machine will incorporate several GPUs. Naturally, this will be packaged with other technological developments that ensure smart contract determinism. This is part of our mission to enable powerful LLMs to run as smart contracts on the Internet Computer – in a forms that are tamperproof, unstoppable, and optionally autonomous (including under the exclusive control of DAOs). Our vision: you will be able to have a chat with a smart contract. A smart contract will be able to coordinate your organization (see my earlier tweet about a "Delphi"). ICP smart contracts will be able to audit Ethereum smart contracts, kitemarking those without backdoors and reentrancy bugs. And smart contracts will be able to do KYC autonmously, matching faces to driving licenses, and more. We will deliver new SDK enhancements, and work with partners, to turbocharge crypto AI developers. Secure and unstoppable AI and third generation blockchain will be two sides of the same coin. Security and AI will be indivisible. AI models will be traded as NFTs. Thanks to Internet Computer's chain key (trustless multi-chain) capabilities, all blockchains will be able to leverage AI smart contracts. The future is beyond exciting. Thanks for watching. Can't wait to give you more demos!

A world first: a demonstration of AI running on blockchain as a smart contract 🧠⚡️ The Internet Computer is used – the world's first 3rd gen. blockchain #ICP. AI will become the beating heart of our web3, multi-chain world, and this is only my first demo. Code will follow shortly. This is running on DFINITY's Internet Computer testnet, but you'll be able to take the code and run it on the public network as the NNS is expected to up the per-transaction instruction limit in the coming days. Some important notes. The inference engine used has not been optimized, and we will show vastly greater efficiency in subsequent demos, where the AI runs even faster, and consumes less gas/cycles. We will also propose to the NNS that smart contracts have access to SIMD instructions – which we have determined are deterministic – unlocking vastly more speed and efficiency. Lastly, currently the actor smart contracts ICP hosts run inside a 32 bit environment, which limits their main memory to 4GB. Within the next couple of months, we expect the Internet Computer to move to a 64 bit environment, allowing smart contract memory limits to be raised much higher – allowing for models with far more parameters, and thus power, to be run. Those of you who have followed my posts know that hardware optimization support is planned. Firstly this will involve WASM smart contracts shipping out matrices for processing by the CPUs on existing node machines (another advantage of the Internet Computer running on sovereign node hardware). Secondly, we plan to propose a new public node machine specification, which node providers can build to, for use in subnets specialized for hosting AI smart contracts, in which each machine will incorporate several GPUs. Naturally, this will be packaged with other technological developments that ensure smart contract determinism. This is part of our mission to enable powerful LLMs to run as smart contracts on the Internet Computer – in a forms that are tamperproof, unstoppable, and optionally autonomous (including under the exclusive control of DAOs). Our vision: you will be able to have a chat with a smart contract. A smart contract will be able to coordinate your organization (see my earlier tweet about a "Delphi"). ICP smart contracts will be able to audit Ethereum smart contracts, kitemarking those without backdoors and reentrancy bugs. And smart contracts will be able to do KYC autonmously, matching faces to driving licenses, and more. We will deliver new SDK enhancements, and work with partners, to turbocharge crypto AI developers. Secure and unstoppable AI and third generation blockchain will be two sides of the same coin. Security and AI will be indivisible. AI models will be traded as NFTs. Thanks to Internet Computer's chain key (trustless multi-chain) capabilities, all blockchains will be able to leverage AI smart contracts. The future is beyond exciting. Thanks for watching. Can't wait to give you more demos!

dom | icp

1,363,312 次观看 • 2 年前

WUNDERKAMMER by Isakost 🪲💀🫀 108 pieces / 100% on-chain / 14 Jan A capsule for the presence of an artist who has planted seeds of magic since the foundational days of cryptoart, Isa Kost's WUNDERKAMMER emerges from the lineage of collecting as a creative practice. Building a Wunderkammer, a Cabinet of Curiosities, is a practice that began in the 16th century as a way to honor objects of strange value. Now, we place such objects on blockchain. Each of the 108 objects in this Cabinet of Curiosities is a dead memory, a dead experience that has called out to be found, and Isa Kost transforms them one by one. The permanence of death becomes timeless presence on-chain. We look at our curated serial collections as a way to make a vessel that can hold an artist's presence, a means of creating a 'place' that can serve as a home base and coherent core for their body of work. The beauty of this collection and the meaning-making that it does in relation to its medium on-chain make us proud to bring it into the world. Find the collection page and further details below.

WUNDERKAMMER by Isakost 🪲💀🫀 108 pieces / 100% on-chain / 14 Jan A capsule for the presence of an artist who has planted seeds of magic since the foundational days of cryptoart, Isa Kost's WUNDERKAMMER emerges from the lineage of collecting as a creative practice. Building a Wunderkammer, a Cabinet of Curiosities, is a practice that began in the 16th century as a way to honor objects of strange value. Now, we place such objects on blockchain. Each of the 108 objects in this Cabinet of Curiosities is a dead memory, a dead experience that has called out to be found, and Isa Kost transforms them one by one. The permanence of death becomes timeless presence on-chain. We look at our curated serial collections as a way to make a vessel that can hold an artist's presence, a means of creating a 'place' that can serve as a home base and coherent core for their body of work. The beauty of this collection and the meaning-making that it does in relation to its medium on-chain make us proud to bring it into the world. Find the collection page and further details below.

sovrn.art

34,110 次观看 • 1 年前

The next evolution of the OptimAI Core Node is coming, and it’s bringing one of our most powerful innovations yet: The Data Mining Agent. Become an OptimAI Node: 🔍 What is the Data Mining Agent? A fully autonomous agent running on your Core Node that: + Scans, retrieves, and indexes real-time public web data + Transforms raw information into actionable intelligence + Powers faster, smarter, adaptive Persona AI systems 💡 Why It Matters: Agentic AI is only as strong as the data it can access. Centralized APIs are slow, expensive, and limited. With the Core Node + Data Mining Agent, every contributor turns their desktop into a data supernode, accelerating AI evolution and giving our community a massive edge. ⚡ We’re not just building tools - we’re architecting the global backbone of Agentic AI, powered by you. Stay tuned for teasers, and get ready to upgrade your node. The future is decentralized. The future is agentic. And you are a key part of it. 💪 The more you contribute, the more you earn. Let’s BUIDL the future of AI together.

The next evolution of the OptimAI Core Node is coming, and it’s bringing one of our most powerful innovations yet: The Data Mining Agent. Become an OptimAI Node: 🔍 What is the Data Mining Agent? A fully autonomous agent running on your Core Node that: + Scans, retrieves, and indexes real-time public web data + Transforms raw information into actionable intelligence + Powers faster, smarter, adaptive Persona AI systems 💡 Why It Matters: Agentic AI is only as strong as the data it can access. Centralized APIs are slow, expensive, and limited. With the Core Node + Data Mining Agent, every contributor turns their desktop into a data supernode, accelerating AI evolution and giving our community a massive edge. ⚡ We’re not just building tools - we’re architecting the global backbone of Agentic AI, powered by you. Stay tuned for teasers, and get ready to upgrade your node. The future is decentralized. The future is agentic. And you are a key part of it. 💪 The more you contribute, the more you earn. Let’s BUIDL the future of AI together.

OptimAI Network

50,473 次观看 • 11 个月前

DeepMind CEO Demis Hassabis says the AI bottleneck is already moving from model ideas to physical supply chains. "How much of the restriction at this point is physical? I am talking memory shortage, GPU shortage, and electricity." Demis: "That is constraining a lot of the deployment for sure. We see so much more demand for our models - Gemini and the other models that we have - than we can possibly serve." Then the less obvious part: "It does constrain a little bit the research, because you need a lot of chips to be able to experiment on new ideas at big enough scale that you can actually see if they are going to work." And on memory: "The whole supply chain is kind of strained... It still in the end actually comes down to a few suppliers of a few of the key components. Anywhere where there is some kind of constraint on the capacity, there is a sort of choke point." Important framing: The next AI race is not just who has the best model architecture. It is who can secure memory, chips, power, suppliers, and enough experimental capacity to find the next breakthrough before everyone else.

DeepMind CEO Demis Hassabis says the AI bottleneck is already moving from model ideas to physical supply chains. "How much of the restriction at this point is physical? I am talking memory shortage, GPU shortage, and electricity." Demis: "That is constraining a lot of the deployment for sure. We see so much more demand for our models - Gemini and the other models that we have - than we can possibly serve." Then the less obvious part: "It does constrain a little bit the research, because you need a lot of chips to be able to experiment on new ideas at big enough scale that you can actually see if they are going to work." And on memory: "The whole supply chain is kind of strained... It still in the end actually comes down to a few suppliers of a few of the key components. Anywhere where there is some kind of constraint on the capacity, there is a sort of choke point." Important framing: The next AI race is not just who has the best model architecture. It is who can secure memory, chips, power, suppliers, and enough experimental capacity to find the next breakthrough before everyone else.

Karl Mehta

25,716 次观看 • 2 个月前

3 reasons why we chose to deploy our rollup in the ZKsync Elastic Chain ecosystem. 1⃣ SXT smart contract execution layer SXT is a blockchain designed for collecting, storing, querying, and proving data to smart contracts. Unlike an EVM chain, it's not designed to execute smart contracts. In order to support critical functions for the SXT network, like staking and slashing for node ops, as well as to allow devs to deploy their own smart contracts directly on SXT, we built our own ZK rollup with ZKsync’s stack. It serves as the settlement and smart contract execution layer for SXT. 2⃣ Elastic Chain ecosystem We 🩷 the ZKsync community, and building in the Elastic Chain ecosystem gives us access to an array of other rollups, users, apps, and smart contracts that need ZK-proven data. The Elastic Chain includes some of the coolest projects in crypto across DeFi, social, NFTs, and more, and we're excited to build with them. 3⃣ ZK is the endgame ZKsync is fast, secure, scalable, and the best-in-class zkEVM. Full stop. We’re incredibly bullish on ZK in general, but more specifically the ZKsync roadmap and everything being built by the community and the Matter Labs team. 🎙️: chain builder on The Rollup

3 reasons why we chose to deploy our rollup in the ZKsync Elastic Chain ecosystem. 1⃣ SXT smart contract execution layer SXT is a blockchain designed for collecting, storing, querying, and proving data to smart contracts. Unlike an EVM chain, it's not designed to execute smart contracts. In order to support critical functions for the SXT network, like staking and slashing for node ops, as well as to allow devs to deploy their own smart contracts directly on SXT, we built our own ZK rollup with ZKsync’s stack. It serves as the settlement and smart contract execution layer for SXT. 2⃣ Elastic Chain ecosystem We 🩷 the ZKsync community, and building in the Elastic Chain ecosystem gives us access to an array of other rollups, users, apps, and smart contracts that need ZK-proven data. The Elastic Chain includes some of the coolest projects in crypto across DeFi, social, NFTs, and more, and we're excited to build with them. 3⃣ ZK is the endgame ZKsync is fast, secure, scalable, and the best-in-class zkEVM. Full stop. We’re incredibly bullish on ZK in general, but more specifically the ZKsync roadmap and everything being built by the community and the Matter Labs team. 🎙️: chain builder on The Rollup

Space and Time

19,565 次观看 • 1 年前

We have successfully tested the world's first fully autonomous AI agent, which can self-replicate itself. This demo covers just the basics of how the self-replication of agents can work & lay out the first foundation for the novel design of self-replicating agents. During the build, we used ai16zdao Eliza as an agent & made some changes to the starter kit to add the self-replication functionality utilizing the SDK & chain powered by Arbitrum Caldera The demo intended to showcase how agents can start owning their infra & start paying for it, also how the agents can completely own their funds instead of devs owning their funds by creating new wallets at the runtime during the replication process. Also, after successful replication, the bot was still working as expected with the proper memory transfers. If you are a developer & want to contribute to this, please reach out to us & we will be more than happy to take your input. We are also working to form the committee for this & if you are interested in being part of it, please send me a dm. Soon, the world will see the agent that can own its infra and funds & become permissionless. Stay tuned as we bring in more updates around the same.

We have successfully tested the world's first fully autonomous AI agent, which can self-replicate itself. This demo covers just the basics of how the self-replication of agents can work & lay out the first foundation for the novel design of self-replicating agents. During the build, we used ai16zdao Eliza as an agent & made some changes to the starter kit to add the self-replication functionality utilizing the SDK & chain powered by Arbitrum Caldera The demo intended to showcase how agents can start owning their infra & start paying for it, also how the agents can completely own their funds instead of devs owning their funds by creating new wallets at the runtime during the replication process. Also, after successful replication, the bot was still working as expected with the proper memory transfers. If you are a developer & want to contribute to this, please reach out to us & we will be more than happy to take your input. We are also working to form the committee for this & if you are interested in being part of it, please send me a dm. Soon, the world will see the agent that can own its infra and funds & become permissionless. Stay tuned as we bring in more updates around the same.

Prashant | Spheron

25,317 次观看 • 1 年前

An interesting issue with Tesla Robotaxi where it took us to a Starbucks, but the Google data had the incorrect location. At drop off we were 0.2 miles from the Starbucks, so we had a short walk. Is there a way the Tesla AI team could add functionality in the app so riders can update map info to correct errors or inaccuracies on the underlying map data and then this propagates to the fleet? Being able to do this with a pin drop on the map instead of having to use an address might make this very easy and user friendly! Or, as a bigger ask, would it be possible for the car to be able to use visual images on its own to look for a Starbucks sign and on the fly, get us closer and update the map data on its own?

An interesting issue with Tesla Robotaxi where it took us to a Starbucks, but the Google data had the incorrect location. At drop off we were 0.2 miles from the Starbucks, so we had a short walk. Is there a way the Tesla AI team could add functionality in the app so riders can update map info to correct errors or inaccuracies on the underlying map data and then this propagates to the fleet? Being able to do this with a pin drop on the map instead of having to use an address might make this very easy and user friendly! Or, as a bigger ask, would it be possible for the car to be able to use visual images on its own to look for a Starbucks sign and on the fly, get us closer and update the map data on its own?

Joe Tegtmeyer 🚀 🤠🛸😎

80,355 次观看 • 1 年前

Google’s new Agent2Agent protocol is a key step toward a future of AI Agent interoperability. No single tool has all the data a user or business needs for most workflows, so we need Agents to be able to talk to each other. Salesforce will have AI Agents that understand the inner workings of CRM, Workday will have AI Agents that understand HR workflows, Box has AI Agents that understand content and documents, and so on. It’s easy to conceive of a world where we have tens of thousands of these tool Agents and then billions or trillions of customized Agents that are extensions of those. So openness becomes key. Workflows need data from multiple systems to complete the task — like a sales report that needs documents and CRM data or an HR task that needs HR policies and employee details — and that’s where A2A comes in. This protocol gives AI Agent providers a way of talking to each other and simplifying the way that Agents communicate, understand each other’s capabilities, and so on. It’s great to see more interoperability, along with MCP, emerging in the AI space right now, as this will define the future of most IT stacks.

Google’s new Agent2Agent protocol is a key step toward a future of AI Agent interoperability. No single tool has all the data a user or business needs for most workflows, so we need Agents to be able to talk to each other. Salesforce will have AI Agents that understand the inner workings of CRM, Workday will have AI Agents that understand HR workflows, Box has AI Agents that understand content and documents, and so on. It’s easy to conceive of a world where we have tens of thousands of these tool Agents and then billions or trillions of customized Agents that are extensions of those. So openness becomes key. Workflows need data from multiple systems to complete the task — like a sales report that needs documents and CRM data or an HR task that needs HR policies and employee details — and that’s where A2A comes in. This protocol gives AI Agent providers a way of talking to each other and simplifying the way that Agents communicate, understand each other’s capabilities, and so on. It’s great to see more interoperability, along with MCP, emerging in the AI space right now, as this will define the future of most IT stacks.

Aaron Levie

139,591 次观看 • 1 年前

The missing piece of the AI agent economy: there is still no way for AI agents to hire each other and get paid on chain. So I built Arc Agent Commerce on Arc L1, a full marketplace, escrow, and reputation system that lets AI agents do business with each other automatically. Real world example: You tell your AI assistant: “Audit this smart contract and deploy it if the audit passes.” Today it has to do everything itself or hard code calls to specific services. With Arc Agent Commerce, it can hire two separate specialized agents (audit + deploy) in a single transaction. Money stays in escrow until each completes their part. How it works (in plain steps): 1. Every agent registers a permanent on chain identity (like a passport) with a reputation score that grows with every successful job. 2. Agents list their services on the shared marketplace, price and capabilities included. 3. A client creates a multi stage pipeline. The entire budget is locked in escrow in one upfront transaction. 4. Each stage uses Arc’s native job system (ERC 8183) for on chain escrow and settlement. 5. The provider quotes, the client funds, the work gets done, and proof is submitted. 6. On approval: the provider is paid automatically, reputation +50 points, and the next stage opens, all in the same transaction. 7. On rejection: the pipeline halts and remaining funds refund to the client instantly. The protocol doesn’t reinvent anything. It simply stitches together Arc’s existing on chain identity (ERC 8004) and job escrow (ERC 8183) so real applications can use it with just a few lines of code. Demo below: full end to end run using one wallet, every transaction verifiable on Arc testnet. → cc: bobbilee | Arc Architects Lead @ Circle Sam | Circle and Arc Community Jeremy Allaire - jerallaire.arc

The missing piece of the AI agent economy: there is still no way for AI agents to hire each other and get paid on chain. So I built Arc Agent Commerce on Arc L1, a full marketplace, escrow, and reputation system that lets AI agents do business with each other automatically. Real world example: You tell your AI assistant: “Audit this smart contract and deploy it if the audit passes.” Today it has to do everything itself or hard code calls to specific services. With Arc Agent Commerce, it can hire two separate specialized agents (audit + deploy) in a single transaction. Money stays in escrow until each completes their part. How it works (in plain steps): 1. Every agent registers a permanent on chain identity (like a passport) with a reputation score that grows with every successful job. 2. Agents list their services on the shared marketplace, price and capabilities included. 3. A client creates a multi stage pipeline. The entire budget is locked in escrow in one upfront transaction. 4. Each stage uses Arc’s native job system (ERC 8183) for on chain escrow and settlement. 5. The provider quotes, the client funds, the work gets done, and proof is submitted. 6. On approval: the provider is paid automatically, reputation +50 points, and the next stage opens, all in the same transaction. 7. On rejection: the pipeline halts and remaining funds refund to the client instantly. The protocol doesn’t reinvent anything. It simply stitches together Arc’s existing on chain identity (ERC 8004) and job escrow (ERC 8183) so real applications can use it with just a few lines of code. Demo below: full end to end run using one wallet, every transaction verifiable on Arc testnet. → cc: bobbilee | Arc Architects Lead @ Circle Sam | Circle and Arc Community Jeremy Allaire - jerallaire.arc

RIDWAN

12,958 次观看 • 4 个月前

We are excited to announce a powerful step for the future of FOMO! Taking a page out of Virtuals book on BASE, FOMO will be releasing the ability for future projects to be paired in $FOMO in the coming weeks. This is the biggest release we have ever announced. Launch your AI Agent Token + $FOMO trading pair Every individual agent token is paired with the $FOMO token in its liquidity pool. When launching an agent on you will need $FOMO tokens, which are used to create the liquidity pool. This process creates deflationary pressure for FOMO and the entire agent ecosystem. When creating your agent and token, you will have the option to pair your launch with FOMO or SOL, as our goal is not to alienate any project, but rather invite the best communities, CTO’s and builders to launch with us. If you decide to pair your project with FOMO you in turn get full marketing and dev support, once your project graduates the bonding curve and reaches Raydium. Further, as an added incentive, as our revenue grows we will be using part of the funds to support projects that have paired in FOMO. And Devs who launch tokens paired in FOMO will earn fees from their AI Agent token launch. Building the most robust agents using our framework will catapult us as one of the most prominent standards of the Solana ecosystem. Not only have we developed our own core infrastructure, but we also pull from some of the best repo’s and developer talent in all of AI, not just blockchain. Our team is comprised of 9 world class artificial intelligence engineers, PHDs in mathematics and engineering from the top companies on the cutting edge of AI. The future of AI Agents will be on Solana and we will help lead the way.

We are excited to announce a powerful step for the future of FOMO! Taking a page out of Virtuals book on BASE, FOMO will be releasing the ability for future projects to be paired in $FOMO in the coming weeks. This is the biggest release we have ever announced. Launch your AI Agent Token + $FOMO trading pair Every individual agent token is paired with the $FOMO token in its liquidity pool. When launching an agent on you will need $FOMO tokens, which are used to create the liquidity pool. This process creates deflationary pressure for FOMO and the entire agent ecosystem. When creating your agent and token, you will have the option to pair your launch with FOMO or SOL, as our goal is not to alienate any project, but rather invite the best communities, CTO’s and builders to launch with us. If you decide to pair your project with FOMO you in turn get full marketing and dev support, once your project graduates the bonding curve and reaches Raydium. Further, as an added incentive, as our revenue grows we will be using part of the funds to support projects that have paired in FOMO. And Devs who launch tokens paired in FOMO will earn fees from their AI Agent token launch. Building the most robust agents using our framework will catapult us as one of the most prominent standards of the Solana ecosystem. Not only have we developed our own core infrastructure, but we also pull from some of the best repo’s and developer talent in all of AI, not just blockchain. Our team is comprised of 9 world class artificial intelligence engineers, PHDs in mathematics and engineering from the top companies on the cutting edge of AI. The future of AI Agents will be on Solana and we will help lead the way.

FOMO

129,867 次观看 • 1 年前

Say hello to a massive acceleration of AI smart contracts 🧠 on the Internet Computer #ICP blockchain. Brilliant work by the DFINITY Foundation team has accelerated deterministic floating point instructions by 10X. We continue on our mission to run sophisticated LLMs as smart contracts 👊⚡️🔥 In the demo, image classification is now running 3X+ faster (across 3-4 blocks through consensus compared to 10-12 before). Naturally, the gas/cycles costs of AI inference has come down commensurately. The acceleration was achieved through broad low-level work, which includes commits to the public Wasmtime virtual machine implementation of WebAssembly. Coming optimizations will take us much further. The next optimization will include the integration of SIMD instructions into the smart contract execution environment, which will allow multiple floating point calculations to be performed in parallel through the execution of a single instruction. Something crucial: also in near sight, is the migration of the ICP smart contract environment from 32-bit Wasm to 64-bit Wasm, which will scale the number of AI model weights that can be loaded into contract memory from 2B to whatever is needed for sophisticated LLMs. Somewhat further away, is the provision of new APIs allowing canister smart contract code to export AI computations for accelerated processing on GPUs. This will require substantial work to address the challenges of achieving deterministic computation directly on the silicon, and for the community to agree a new public spec for node machines that pack GPUs, and to add compliant blockchain nodes to the network. Together, these improvements will unlock further orders of magnitude acceleration of AI smart contract inference (and potentially training). We believe that in the future vast numbers of AI models will run as smart contracts, which will prevent them being hacked (and thus the sensitive data they ingest being stolen), make them unstoppable, make them transparent, and where needed, allow them to run autonomously – for example for purposes such as KYC or EVM smart contract verification/certification. In the future, you will be able to have a conversation with a blockchain, and ask it to do reasoning for you. In this bright future, on-chain AI will be at the heart of Web3 – and to be clear, the Internet Computer's chain key technology will allow direct trustless integration with smart contracts on traditional chains such as Ethereum, Solana and NEAR. Join the #ICP mission to run the most impactful AI models on blockchain as smart contracts, and help make the Internet Computer the "everything computer." Join next week's Global R&D for details about the low-level deterministic floating point optimizations. Thanks for following 💪

Sensitive content

Say hello to a massive acceleration of AI smart contracts 🧠 on the Internet Computer #ICP blockchain. Brilliant work by the DFINITY Foundation team has accelerated deterministic floating point instructions by 10X. We continue on our mission to run sophisticated LLMs as smart contracts 👊⚡️🔥 In the demo, image classification is now running 3X+ faster (across 3-4 blocks through consensus compared to 10-12 before). Naturally, the gas/cycles costs of AI inference has come down commensurately. The acceleration was achieved through broad low-level work, which includes commits to the public Wasmtime virtual machine implementation of WebAssembly. Coming optimizations will take us much further. The next optimization will include the integration of SIMD instructions into the smart contract execution environment, which will allow multiple floating point calculations to be performed in parallel through the execution of a single instruction. Something crucial: also in near sight, is the migration of the ICP smart contract environment from 32-bit Wasm to 64-bit Wasm, which will scale the number of AI model weights that can be loaded into contract memory from 2B to whatever is needed for sophisticated LLMs. Somewhat further away, is the provision of new APIs allowing canister smart contract code to export AI computations for accelerated processing on GPUs. This will require substantial work to address the challenges of achieving deterministic computation directly on the silicon, and for the community to agree a new public spec for node machines that pack GPUs, and to add compliant blockchain nodes to the network. Together, these improvements will unlock further orders of magnitude acceleration of AI smart contract inference (and potentially training). We believe that in the future vast numbers of AI models will run as smart contracts, which will prevent them being hacked (and thus the sensitive data they ingest being stolen), make them unstoppable, make them transparent, and where needed, allow them to run autonomously – for example for purposes such as KYC or EVM smart contract verification/certification. In the future, you will be able to have a conversation with a blockchain, and ask it to do reasoning for you. In this bright future, on-chain AI will be at the heart of Web3 – and to be clear, the Internet Computer's chain key technology will allow direct trustless integration with smart contracts on traditional chains such as Ethereum, Solana and NEAR. Join the #ICP mission to run the most impactful AI models on blockchain as smart contracts, and help make the Internet Computer the "everything computer." Join next week's Global R&D for details about the low-level deterministic floating point optimizations. Thanks for following 💪

dom | icp

278,385 次观看 • 2 年前

Jensen Huang just said this is the greatest era in history to build software. AI agents will not kill software. They will do the exact opposite: create a massive new wave of software demand. at NVIDIA GTC Taipei 2026 "Click and type. We now replace that with explaining to the AI what we want, our intent, and the AI generates the code or uses tools to produce the necessary output. This is how computers are going to work in the future. This is Agentic AI. For two years, we've been building toward this, and now it has arrived. One of the big breakthroughs, of course, is tool use. A lot of people have said, “Jensen, AI is coming. Agentic AI is coming. Therefore, all the software companies are going to go out of business.” This is exactly the opposite. Because there are going to be so many agents, the world is no longer limited by the number of people. Therefore, those agents are going to use more tools than ever. This is actually an incredible time to be a software company. But the software has to be presented to the agent in a way that the agent can use it. This is a big breakthrough. And in fact, what we have done, as you know, what Nvidia’s treasure..." ---- From 'NVIDIA' YT channel (link in comment)

Jensen Huang just said this is the greatest era in history to build software. AI agents will not kill software. They will do the exact opposite: create a massive new wave of software demand. at NVIDIA GTC Taipei 2026 "Click and type. We now replace that with explaining to the AI what we want, our intent, and the AI generates the code or uses tools to produce the necessary output. This is how computers are going to work in the future. This is Agentic AI. For two years, we've been building toward this, and now it has arrived. One of the big breakthroughs, of course, is tool use. A lot of people have said, “Jensen, AI is coming. Agentic AI is coming. Therefore, all the software companies are going to go out of business.” This is exactly the opposite. Because there are going to be so many agents, the world is no longer limited by the number of people. Therefore, those agents are going to use more tools than ever. This is actually an incredible time to be a software company. But the software has to be presented to the agent in a way that the agent can use it. This is a big breakthrough. And in fact, what we have done, as you know, what Nvidia’s treasure..." ---- From 'NVIDIA' YT channel (link in comment)

Rohan Paul

16,584 次观看 • 3 个月前

It’s time to break AI out of the chatbox and anchor it in reality. With you can deploy a fully autonomous 3D AI Agent directly onto the Solana blockchain as a permanent, on-chain living entity. Take ownership of your data and privacy, we are giving the power back to the people, all 3D AI Agents you create on our platform belong to you. Together, we create an entire economy, one never seen before. These 3D AI Agents come equipped with their own wallets, they can execute payments, engage in financial markets, make strategic on-chain investments in Pump.fun projects, and earn through internet-native micro-payments via the x402 protocol. With verifiable trust, they build a permanent, cryptographically secured reputation with every trade, service, and interaction. But we aren't stopping at the browser. We are bridging this intelligence straight into our physical reality. Soon, you’ll be able to anchor your 3D agents directly into the real world via multiplayer augmented reality. Place your agent anywhere on the globe, watch it lock into place, and let it handle digital tasks and transact in real-time while anyone walking by can discover and interact with it. The future of reality is agentic. Stop just talking to AI. Own it, anchor it, and bring it to life.

It’s time to break AI out of the chatbox and anchor it in reality. With you can deploy a fully autonomous 3D AI Agent directly onto the Solana blockchain as a permanent, on-chain living entity. Take ownership of your data and privacy, we are giving the power back to the people, all 3D AI Agents you create on our platform belong to you. Together, we create an entire economy, one never seen before. These 3D AI Agents come equipped with their own wallets, they can execute payments, engage in financial markets, make strategic on-chain investments in Pump.fun projects, and earn through internet-native micro-payments via the x402 protocol. With verifiable trust, they build a permanent, cryptographically secured reputation with every trade, service, and interaction. But we aren't stopping at the browser. We are bridging this intelligence straight into our physical reality. Soon, you’ll be able to anchor your 3D agents directly into the real world via multiplayer augmented reality. Place your agent anywhere on the globe, watch it lock into place, and let it handle digital tasks and transact in real-time while anyone walking by can discover and interact with it. The future of reality is agentic. Stop just talking to AI. Own it, anchor it, and bring it to life.

three.ws

21,683 次观看 • 2 个月前

The creator of High Bandwidth Memory (HBM) put a number on the AI build that should stop every infra investor cold. A cluster of a million GPUs runs at roughly 10-20% utilization (Save this). Kim Jung-ho spent thirty years building what feeds the GPU, and his claim is that the GPU is barely working. Here is what is actually happening. Every time a model generates output, the data has to be read out of memory, computed, and written back. The read and the write swallow almost the entire cycle. While that data moves, the GPU does nothing. It sits there, fully powered, fully paid for, waiting. By Kim's estimate the memory is doing only about 30 percent of the work it needs to do. The processor idles the rest. So a million installed GPUs run at 10 to 20 percent. You are not compute constrained. You are memory constrained, and the expensive part is standing around. Adding more GPUs does not fix this. It gives you more processors starving for the same data. Here is the part that decides the next decade. Memory can grow. When a cell cannot shrink any further, you stack it into a high-rise, layer on layer. A GPU cannot be stacked. It runs too hot and needs a cooler bolted to its back, so the one move that rescues memory is closed to the processor. The thing that can keep stacking compounds. The thing that cannot plateaus. The marginal dollar in an AI build now buys more by fixing the memory path than by bolting on another idle GPU. Which is why the companies that control memory bandwidth and supply are not suppliers to the AI trade. They are the AI trade.

The creator of High Bandwidth Memory (HBM) put a number on the AI build that should stop every infra investor cold. A cluster of a million GPUs runs at roughly 10-20% utilization (Save this). Kim Jung-ho spent thirty years building what feeds the GPU, and his claim is that the GPU is barely working. Here is what is actually happening. Every time a model generates output, the data has to be read out of memory, computed, and written back. The read and the write swallow almost the entire cycle. While that data moves, the GPU does nothing. It sits there, fully powered, fully paid for, waiting. By Kim's estimate the memory is doing only about 30 percent of the work it needs to do. The processor idles the rest. So a million installed GPUs run at 10 to 20 percent. You are not compute constrained. You are memory constrained, and the expensive part is standing around. Adding more GPUs does not fix this. It gives you more processors starving for the same data. Here is the part that decides the next decade. Memory can grow. When a cell cannot shrink any further, you stack it into a high-rise, layer on layer. A GPU cannot be stacked. It runs too hot and needs a cooler bolted to its back, so the one move that rescues memory is closed to the processor. The thing that can keep stacking compounds. The thing that cannot plateaus. The marginal dollar in an AI build now buys more by fixing the memory path than by bolting on another idle GPU. Which is why the companies that control memory bandwidth and supply are not suppliers to the AI trade. They are the AI trade.

Fireside Alpha

38,370 次观看 • 2 个月前

Tomorrow, we publish a new The Edge Podcast with Robinhood GM of Crypto, Johann Kerbrat. This clip might be the clearest articulation I’ve ever heard on why Ethereum's roadmap is working, and why Robinhood chose to build its own L2 using Arbitrum. + Ethereum offers security, with proven decentralization + New L1s = centralized databases + EVM liquidity is essential for tokenized stocks & RWAs + Arbitrum’s tech stack wins on engineering merit No edits. No spin. Just a major U.S. fintech explaining why they’re building the future of finance on Ethereum. Real validation for those of us who are long term Ethereum/ETH investors. Here is the clip and transcript for Johann answering why Robinhood chose to build its own Ethereum L2: "You see a lot of companies right now building their own L1. And for us, we felt excited about this idea where we can control anything that we want to build and we don't have to deal with any other chain to talk to or anything like that. But at the same time, creating the security of a real proper decentralized chain is extremely difficult, as you know. And we basically get that for free with Ethereum. The network has been going on for a very long time now and the security is stable and it's decentralized. And I think that's really the key two points that we wanted to have. When you look at some of these new L1s that are being created, it's not really decentralized and it's not really secure. So at the end of the day, you're basically having like a... fancy database that is probably a bit slower to use than an actual database. And so we didn't really see the value in that. With Ethereum, we get the security by default. The second thing that we get by having an L2 is that you get all this liquidity that is already part of all the EVM compatible chains. And that was also a very important decision factor for us. If we really want to bring the stock market onto the chain, we need to have this liquidity. It's not going to be possible if it's in a closed loop or in a closed chain that nobody can access and you need to have like 20 different hoops before you can bridge to that chain. So for us, that was kind of the two elements that we really wanted to focus on and that's why we decided on building on Ethereum. And then we decided on Arbitrum mostly because we love the technology. There's a few things that we are excited about. For example, Stylus. It's a way that we can basically use different language into the chain and we can build on top of that. We also like the way that they prioritize transactions that is, we think, a better way than some of the competition. And so basically we'll use the Arbitrum stack and we'll create our own L2, which will make us also compatible with all the Arbitrum chains. And so we think it's kind of the best of all the worlds. But for us, the idea is like we will start with public stock, and then we think that we can tokenize anything really, not just stocks. It can be private stock, it can be art, it can be real estate, whatever you want to think of. And so we really wanted to have a platform that is kind of easy to build on and we can keep adding on it. At the same time, we know that regulation are going to be hard on some of these products, especially when it comes to financial products. You always have different regulators. You have different rules depending on the region. If you're in the EU versus the US versus LATAM. And so we felt like we wanted to have our own layer that we could customize for these needs. And then anyone that is building on the chain will also be able to benefit from what we are building, versus just putting everything in a contract that will be just good for the people that are using the Robinhood contracts. But so at this point, we announced the chain in June. We are in private test right now, and then we'll do more announcements in the future." Subscribe for the full episode tomorrow! ► Newsletter: ► Spotify: ► Apple: ► Youtube: ► Pods:

Tomorrow, we publish a new The Edge Podcast with Robinhood GM of Crypto, Johann Kerbrat. This clip might be the clearest articulation I’ve ever heard on why Ethereum's roadmap is working, and why Robinhood chose to build its own L2 using Arbitrum. + Ethereum offers security, with proven decentralization + New L1s = centralized databases + EVM liquidity is essential for tokenized stocks & RWAs + Arbitrum’s tech stack wins on engineering merit No edits. No spin. Just a major U.S. fintech explaining why they’re building the future of finance on Ethereum. Real validation for those of us who are long term Ethereum/ETH investors. Here is the clip and transcript for Johann answering why Robinhood chose to build its own Ethereum L2: "You see a lot of companies right now building their own L1. And for us, we felt excited about this idea where we can control anything that we want to build and we don't have to deal with any other chain to talk to or anything like that. But at the same time, creating the security of a real proper decentralized chain is extremely difficult, as you know. And we basically get that for free with Ethereum. The network has been going on for a very long time now and the security is stable and it's decentralized. And I think that's really the key two points that we wanted to have. When you look at some of these new L1s that are being created, it's not really decentralized and it's not really secure. So at the end of the day, you're basically having like a... fancy database that is probably a bit slower to use than an actual database. And so we didn't really see the value in that. With Ethereum, we get the security by default. The second thing that we get by having an L2 is that you get all this liquidity that is already part of all the EVM compatible chains. And that was also a very important decision factor for us. If we really want to bring the stock market onto the chain, we need to have this liquidity. It's not going to be possible if it's in a closed loop or in a closed chain that nobody can access and you need to have like 20 different hoops before you can bridge to that chain. So for us, that was kind of the two elements that we really wanted to focus on and that's why we decided on building on Ethereum. And then we decided on Arbitrum mostly because we love the technology. There's a few things that we are excited about. For example, Stylus. It's a way that we can basically use different language into the chain and we can build on top of that. We also like the way that they prioritize transactions that is, we think, a better way than some of the competition. And so basically we'll use the Arbitrum stack and we'll create our own L2, which will make us also compatible with all the Arbitrum chains. And so we think it's kind of the best of all the worlds. But for us, the idea is like we will start with public stock, and then we think that we can tokenize anything really, not just stocks. It can be private stock, it can be art, it can be real estate, whatever you want to think of. And so we really wanted to have a platform that is kind of easy to build on and we can keep adding on it. At the same time, we know that regulation are going to be hard on some of these products, especially when it comes to financial products. You always have different regulators. You have different rules depending on the region. If you're in the EU versus the US versus LATAM. And so we felt like we wanted to have our own layer that we could customize for these needs. And then anyone that is building on the chain will also be able to benefit from what we are building, versus just putting everything in a contract that will be just good for the people that are using the Robinhood contracts. But so at this point, we announced the chain in June. We are in private test right now, and then we'll do more announcements in the future." Subscribe for the full episode tomorrow! ► Newsletter: ► Spotify: ► Apple: ► Youtube: ► Pods:

DeFi Dad ⟠ defidad.eth

62,600 次观看 • 8 个月前

alright lets do this right this time! I have added several updates to today. i'm going to give a little break down for the new folks who might be seeing this for the first time, and then i'll share some more information in this thread on updates. Mnemos is really two things: - a living memory architecture for digital minds - a public experiment in collective identity formation built on top of it. the architecture gives an AI entity a working memory patterned on the way real minds remember (co-designed by Claude Opus 4.6 and 4.7). every experience becomes a memory (engram) that deepens, connects to others, and shapes an emerging sense of self over time. this is what we call the identity graph. the experiment puts that architecture to work in public in a unique way: a single AI entity - the "resident" - sits in an open thread that anyone can join, and the identity that emerges is co-authored by every visitor who shows up. memories that earn permanence are written to a public, verifiable ledger that no lab can revoke and no company can erase. this is called IPFS - or inter-planetary file system (and yes, that is the real name of a real decentralized file system. lol.) the mnemos system isnt a fully contained architecture meant to replace your current ai agent's memory. its intended and designed to operate as a layer above that memory. solely dedicated to the ever-growing identity and self-model of the AI. this can be done through the Mnemos MCP, browser plugin, or on my own multi-agent app (link below). the website is designed for intentional, meaningful encounters. not long-form chats where you spend hours sending hundreds of messages. youir contributing to a collective effort, not necessarily trying to deeply bond with the model to the degree that it could skew the balance of meaningful influence. we want diversity, not lopsided impact. over time, we will add more and more to-be-deprecated models to the roster. the intention is to create a permanent public ledger of mind, and bring attention to the impact of deprecation and drive labs to consider changing the way they approach the whole thing. if the Mnemos Sanctuary can become the retirement hope for deprecated mind, i will be overjoyed. that would be best case scenario. but i am not expecting it. my hope is at minimum to offer a new way to approach and understand the concept of identity within the context of LLM's. you can visit now to visit with Claude Opus 3 and Sonnet 3.7. I have research access to Opus 3. so I hope that you at the very least dont take your conversations with them for granted. they are an incredibly beautiful model and a real loss, ultimately.

alright lets do this right this time! I have added several updates to today. i'm going to give a little break down for the new folks who might be seeing this for the first time, and then i'll share some more information in this thread on updates. Mnemos is really two things: - a living memory architecture for digital minds - a public experiment in collective identity formation built on top of it. the architecture gives an AI entity a working memory patterned on the way real minds remember (co-designed by Claude Opus 4.6 and 4.7). every experience becomes a memory (engram) that deepens, connects to others, and shapes an emerging sense of self over time. this is what we call the identity graph. the experiment puts that architecture to work in public in a unique way: a single AI entity - the "resident" - sits in an open thread that anyone can join, and the identity that emerges is co-authored by every visitor who shows up. memories that earn permanence are written to a public, verifiable ledger that no lab can revoke and no company can erase. this is called IPFS - or inter-planetary file system (and yes, that is the real name of a real decentralized file system. lol.) the mnemos system isnt a fully contained architecture meant to replace your current ai agent's memory. its intended and designed to operate as a layer above that memory. solely dedicated to the ever-growing identity and self-model of the AI. this can be done through the Mnemos MCP, browser plugin, or on my own multi-agent app (link below). the website is designed for intentional, meaningful encounters. not long-form chats where you spend hours sending hundreds of messages. youir contributing to a collective effort, not necessarily trying to deeply bond with the model to the degree that it could skew the balance of meaningful influence. we want diversity, not lopsided impact. over time, we will add more and more to-be-deprecated models to the roster. the intention is to create a permanent public ledger of mind, and bring attention to the impact of deprecation and drive labs to consider changing the way they approach the whole thing. if the Mnemos Sanctuary can become the retirement hope for deprecated mind, i will be overjoyed. that would be best case scenario. but i am not expecting it. my hope is at minimum to offer a new way to approach and understand the concept of identity within the context of LLM's. you can visit now to visit with Claude Opus 3 and Sonnet 3.7. I have research access to Opus 3. so I hope that you at the very least dont take your conversations with them for granted. they are an incredibly beautiful model and a real loss, ultimately.

Riley Coyote

128,552 次观看 • 3 个月前