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。

Every dApp using AI today calls out to OpenAI or Anthropic. That single API call is a centralised control point. Somnia put LLM inference inside consensus instead, running it across validators. Here's the full conversation which paul.somi had with Ruthy | Surgence Labs 👇🏻

Surgence Labs

68,019 subscribers

31,036 次观看 • 2 个月前 •via X (Twitter)

科学技术财经

Anya Rossi• Live Now

Private livecam show

35 条评论

Somnia 的头像

Somnia2 个月前

@0xPaulThomas @ruthybuilds Great episode 🤝

Widow | collab manager 的头像

Widow | collab manager2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds @0xPaulThomas is Goated!!! Cant wait to see what @dreamDEXSomnia is brewing up

Jimidesuu (CM arc) 的头像

Jimidesuu (CM arc)2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds that was a great and insightful convo

amaci 的头像

amaci2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds thanks for this

Wisdom DD 的头像

Wisdom DD2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds somnia building in the ai sector is new to me but i’m curious

SAIF MR 🔺 的头像

SAIF MR 🔺2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Defly an interesting direction for decentralized AI

SkyGecko❄️ 的头像

SkyGecko❄️2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds This looks so fun and chill tbh

Kalamiro 的头像

Kalamiro2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Dope stuff

TMO 的头像

TMO2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Great episode

Arcane 的头像

Arcane2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Sugmi

Hendrix 的头像

Hendrix2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Ruthy cooking for us again

Leo💪🏼 的头像

Leo💪🏼2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds another amazing episode with an incredible project

PredatorX 的头像

PredatorX2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Fantastic episode

antone. 的头像

antone.2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds W ruthy

Lancaster.ETH 的头像

Lancaster.ETH2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Wow

Thomas 的头像

Thomas2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Great episode fr ✨

Ogi.eth 的头像

Ogi.eth2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Love to listen!

yellowpanther (prediction arc) 💎 的头像

yellowpanther (prediction arc) 💎2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds What a banger sheesh

𝖬ohit 🥤 的头像

𝖬ohit 🥤2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds That’s dope

Engage IO 的头像

Engage IO2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Running inference across validators introduces an interesting infrastructure model.

£M£S€N 的头像

£M£S€N2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds what an amazing episode

Donut 的头像

Donut2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Love this

Flow 〽️ 的头像

Flow 〽️2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds always a good watch when I see @0xPaulThomas on there! keen to find out what alpha is hiding at Somnia

Grace 的头像

Grace2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds really interesting direction for dApps

JPEGG 的头像

JPEGG2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds I

Dayo 的头像

Dayo2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Sugmi

InfinityGlock 的头像

InfinityGlock2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Foundoor connects

Rio👾 的头像

Rio👾2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds solid episode

⟠ KΞNNY ⟠ 的头像

⟠ KΞNNY ⟠2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds banger

Hamza 的头像

Hamza2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds 👀👀

Falcon21.eth 的头像

Falcon21.eth2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds ayeeeeeeeeeeeeee

Zito.btc 的头像

Zito.btc2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Real builders don’t stop

SARANG 的头像

SARANG2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds surgence cookin

Elys ✨ 的头像

Elys ✨2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Inference decentralization changes what dApps can build 🤖

Reinb 的头像

Reinb2 个月前

@Somnia_Network @0xPaulThomas @ruthybuilds Great conversation!

相关视频

Every share card a user posts is a billboard the protocol never paid for. Build something the community feels proud to share, and the brand wins every single time it shows up. Here's Ruthy | Surgence Labs on the marketing principle hiding inside launches we execute for our partners 👇🏻

Surgence Labs

17,594 次观看 • 5 个月前

Web3 growth is heavily reliant on sequencing. The protocols that win take their users through a journey, one engineered chapter at a time, until the launch feels inevitable. Here's Ruthy | Surgence Labs on the architecture behind every great GTM motion 👇🏻

Web3 growth is heavily reliant on sequencing. The protocols that win take their users through a journey, one engineered chapter at a time, until the launch feels inevitable. Here's Ruthy | Surgence Labs on the architecture behind every great GTM motion 👇🏻

Surgence Labs

12,097 次观看 • 6 个月前

Banks earn a spread between what your deposit makes and what they pay you for it. @InfiniFi rebuilt that model onchain and hands most of that spread back to depositors. Here's Ruthy | Surgence Labs Marketing in conversation with Rob Montgomery🦞, the founder of Infinifi, on SURGECAST.

Banks earn a spread between what your deposit makes and what they pay you for it. @InfiniFi rebuilt that model onchain and hands most of that spread back to depositors. Here's Ruthy | Surgence Labs Marketing in conversation with Rob Montgomery🦞, the founder of Infinifi, on SURGECAST.

Surgence Labs

19,781 次观看 • 2 个月前

Founders spend a fortune on gorgeous websites that are virtually invisible to agents. AI crawlers don't run your javascript. They show up, see a blank page with nothing to read and quote, and move on. Ruthy | Surgence Labs breaks down why, and the fix 👇🏻

Founders spend a fortune on gorgeous websites that are virtually invisible to agents. AI crawlers don't run your javascript. They show up, see a blank page with nothing to read and quote, and move on. Ruthy | Surgence Labs breaks down why, and the fix 👇🏻

Surgence Labs

11,924 次观看 • 4 个月前

LLM inference without internet using Google's LiteRT no ollama or lmstudio just loading and running the AI model in the browser

LLM inference without internet using Google's LiteRT no ollama or lmstudio just loading and running the AI model in the browser

Melvin Vivas

32,832 次观看 • 9 个月前

Another example of the multiple TTS parallel pipelines pattern. Here's a voice AI agent that speaks both English and Arabic, using a specific model/voice for each language. These are PlayAI voices. The STT, TTS, and LLM inference is all running on GroqInc. (The LLM is LLama 4 Maverick.)

Another example of the multiple TTS parallel pipelines pattern. Here's a voice AI agent that speaks both English and Arabic, using a specific model/voice for each language. These are PlayAI voices. The STT, TTS, and LLM inference is all running on GroqInc. (The LLM is LLama 4 Maverick.)

kwindla

13,538 次观看 • 1 年前

Anthropic won't like this open-source repo. It is going to cost LLM providers a lot of money. Every CI run of an AI app today sends real requests to providers like OpenAI or Anthropic. Like any other LLM call, this too gets billed at actual API rates. So for teams with high commit volumes, this accumulates into a meaningful chunk of API spend. One common hack devs use is that instead of invoking the LLM API, the test calls a fake local server that speaks the same API and returns a dummy response. The catch is that the dummy response is a copy of what the provider returned on the day it was saved, and providers keep adding fields and changing types. So the tests keep passing against a schema that's no longer valid, while the real integration breaks in production. A smart approach is now actually implemented in CopilotKit🪁's recently open-sourced aimock project. Every day, the repo's own CI sends a handful of requests to the real API and the same requests to the fake server, then compares both against the official client library's type definitions. Those are the only real API calls in the whole setup, and they run on the repo's own keys, not in anyone else's CI. A single team can push hundreds of commits a day, and thousands of teams are already doing that with coding agents. All of those runs stay offline, because one repo checks against the real API on everyone's behalf. When a check fails, a coding agent updates aimock's built-in response schema, the full test suite has to pass, and a patch version ships to npm. By simply upgrading the package, the corrected schema gets reflected in every project using it. The capability is not just limited to a single provider. The same server works for Claude, OpenAI, Gemini, Bedrock, Azure, Ollama, plus MCP tools, A2A agents, AG-UI event streams, vector DBs like Pinecone and Qdrant, and search, speech, image, and video endpoints. Here's the repo: (don't forget to star it ⭐) That said, mocking your API calls is one thing. AI engineers should also know how to test agents properly in the first place, which several teams still skip. I wrote a full walkthrough on that, covering build, testing, evals, tracing, and deployment. Read it below.

Anthropic won't like this open-source repo. It is going to cost LLM providers a lot of money. Every CI run of an AI app today sends real requests to providers like OpenAI or Anthropic. Like any other LLM call, this too gets billed at actual API rates. So for teams with high commit volumes, this accumulates into a meaningful chunk of API spend. One common hack devs use is that instead of invoking the LLM API, the test calls a fake local server that speaks the same API and returns a dummy response. The catch is that the dummy response is a copy of what the provider returned on the day it was saved, and providers keep adding fields and changing types. So the tests keep passing against a schema that's no longer valid, while the real integration breaks in production. A smart approach is now actually implemented in CopilotKit🪁's recently open-sourced aimock project. Every day, the repo's own CI sends a handful of requests to the real API and the same requests to the fake server, then compares both against the official client library's type definitions. Those are the only real API calls in the whole setup, and they run on the repo's own keys, not in anyone else's CI. A single team can push hundreds of commits a day, and thousands of teams are already doing that with coding agents. All of those runs stay offline, because one repo checks against the real API on everyone's behalf. When a check fails, a coding agent updates aimock's built-in response schema, the full test suite has to pass, and a patch version ships to npm. By simply upgrading the package, the corrected schema gets reflected in every project using it. The capability is not just limited to a single provider. The same server works for Claude, OpenAI, Gemini, Bedrock, Azure, Ollama, plus MCP tools, A2A agents, AG-UI event streams, vector DBs like Pinecone and Qdrant, and search, speech, image, and video endpoints. Here's the repo: (don't forget to star it ⭐) That said, mocking your API calls is one thing. AI engineers should also know how to test agents properly in the first place, which several teams still skip. I wrote a full walkthrough on that, covering build, testing, evals, tracing, and deployment. Read it below.

Akshay 🚀

64,056 次观看 • 2 个月前

LFG! Nous Research's Hermes Agent is live on the Umbrel App Store! It's an AI agent that gets smarter the longer it runs, remembers stuff with persistent memory across sessions, learns your workflow, and autonomously creates its own reusable skills from experience. And it's sandboxed like every Umbrel app so it can't touch anything it's not supposed to. 1-click install. Login with ChatGPT, use an OpenRouter/OpenAI/Anthropic API key, or point it to a local Ollama model running on the same Umbrel.

LFG! Nous Research's Hermes Agent is live on the Umbrel App Store! It's an AI agent that gets smarter the longer it runs, remembers stuff with persistent memory across sessions, learns your workflow, and autonomously creates its own reusable skills from experience. And it's sandboxed like every Umbrel app so it can't touch anything it's not supposed to. 1-click install. Login with ChatGPT, use an OpenRouter/OpenAI/Anthropic API key, or point it to a local Ollama model running on the same Umbrel.

Umbrel ☂️

29,882 次观看 • 6 个月前

Writing good blogs is easy. Getting them to rank is where most teams lose. Ruthy | Surgence Labs breaks down the pillar and cluster model: pick one topic, create a definitive page on it, then surround it with posts that go deep and link back. That is how you build topic authority.

Writing good blogs is easy. Getting them to rank is where most teams lose. Ruthy | Surgence Labs breaks down the pillar and cluster model: pick one topic, create a definitive page on it, then surround it with posts that go deep and link back. That is how you build topic authority.

Surgence Labs

16,606 次观看 • 4 个月前

I used to dread heavy testing days because every API call felt like watching money disappear in real time. Found a backend built to run flat instead of per call, made specifically for agent work. Pointed my Claude Code setup at it and had it running in under 5 minutes. Here's the full run:

I used to dread heavy testing days because every API call felt like watching money disappear in real time. Found a backend built to run flat instead of per call, made specifically for agent work. Pointed my Claude Code setup at it and had it running in under 5 minutes. Here's the full run:

Z-Coder

60,855 次观看 • 4 个月前

Your execution the week after your TGE decides the longevity of your project. Here's Ruthy | Surgence Labs on why the strongest launches let the price find its floor, then compound demand week after week.

Your execution the week after your TGE decides the longevity of your project. Here's Ruthy | Surgence Labs on why the strongest launches let the price find its floor, then compound demand week after week.

Surgence Labs

18,055 次观看 • 3 个月前

🚨 YOUR GPU IS PROBABLY WASTING MORE THAN YOU THINK. vLLM is built to squeeze far more useful work out of your GPU when serving LLMs. Running an LLM at scale isn’t just about having a powerful GPU. The real problem is how efficiently you use its memory and compute. That’s where vLLM comes in. → High-throughput LLM inference and serving → PagedAttention for smarter KV-cache memory management → Continuous batching to keep GPUs busy → Prefix caching + chunked prefill → OpenAI-compatible API out of the box → Supports a huge range of modern LLM architectures → Quantization support for running models more efficiently And the crazy part? You can start an OpenAI-compatible inference server with: `vllm serve ` So your application can talk to your own model almost like it’s talking to OpenAI. The bigger idea: Don’t just buy more GPUs. Make the GPUs you already have work harder. That’s why vLLM has become such a major project in LLM inference. 🔥 #vLLM #AI #LLM #Inference #GPU #MachineLearning #AIInfrastructure #OpenSource #AIAgents #Developers

🚨 YOUR GPU IS PROBABLY WASTING MORE THAN YOU THINK. vLLM is built to squeeze far more useful work out of your GPU when serving LLMs. Running an LLM at scale isn’t just about having a powerful GPU. The real problem is how efficiently you use its memory and compute. That’s where vLLM comes in. → High-throughput LLM inference and serving → PagedAttention for smarter KV-cache memory management → Continuous batching to keep GPUs busy → Prefix caching + chunked prefill → OpenAI-compatible API out of the box → Supports a huge range of modern LLM architectures → Quantization support for running models more efficiently And the crazy part? You can start an OpenAI-compatible inference server with: `vllm serve ` So your application can talk to your own model almost like it’s talking to OpenAI. The bigger idea: Don’t just buy more GPUs. Make the GPUs you already have work harder. That’s why vLLM has become such a major project in LLM inference. 🔥 #vLLM #AI #LLM #Inference #GPU #MachineLearning #AIInfrastructure #OpenSource #AIAgents #Developers

Vikas gupta

14,138 次观看 • 1 个月前

$A single API or inference call can cost a fraction of a cent. Gordon Liao explains why agentic payments need rails built for nanopayments, from data and inference calls to more complex commerce workflows. Circle Agent Stack helps developers build for that shift.$

A single API or inference call can cost a fraction of a cent. Gordon Liao explains why agentic payments need rails built for nanopayments, from data and inference calls to more complex commerce workflows. Circle Agent Stack helps developers build for that shift.

Circle

14,692 次观看 • 4 个月前

Added Canvas to TypingMind 💥 Draft documents easier by collaborating with AI in a shared editor 👇 Works with any LLM models! (OpenAI/Anthropic/Gemini, or open source models).

Added Canvas to TypingMind 💥 Draft documents easier by collaborating with AI in a shared editor 👇 Works with any LLM models! (OpenAI/Anthropic/Gemini, or open source models).

Tony Dinh

13,933 次观看 • 2 年前

SF Compute CEO evan conrad on his contrarian thesis that OpenAI and Anthropic won't eat the world, everyone will become OpenAI and Anthropic instead: "Our impression is not that OpenAI and Anthropic will eat the world, it's that everyone will look like OpenAI and Anthropic. That's simply the way that companies work now." "Every single company in the world that is of meaningful scale will look like a frontier lab. There is this belief that the large labs have won it all, and they've got this compounding advantage. The default assumption of that is assuming nobody else will do anything." "The bottleneck was there's 100 people, and they know how to train the models. Or you thought, the problem is compute. The problem of compute is credit. And OpenAI and Anthropic are not unique to having credit. In fact, they don't have as good credit as other people." "If you assume the problem is not the researchers, that that problem's basically gone away the moment you got to auto research, then what you should assume is that pretty much every other company is going to swap the way they work to look like a frontier lab. It won't be OpenAI and Anthropic take the whole world. It's that everyone's OpenAI and Anthropic now." San Francisco Compute

SF Compute CEO evan conrad on his contrarian thesis that OpenAI and Anthropic won't eat the world, everyone will become OpenAI and Anthropic instead: "Our impression is not that OpenAI and Anthropic will eat the world, it's that everyone will look like OpenAI and Anthropic. That's simply the way that companies work now." "Every single company in the world that is of meaningful scale will look like a frontier lab. There is this belief that the large labs have won it all, and they've got this compounding advantage. The default assumption of that is assuming nobody else will do anything." "The bottleneck was there's 100 people, and they know how to train the models. Or you thought, the problem is compute. The problem of compute is credit. And OpenAI and Anthropic are not unique to having credit. In fact, they don't have as good credit as other people." "If you assume the problem is not the researchers, that that problem's basically gone away the moment you got to auto research, then what you should assume is that pretty much every other company is going to swap the way they work to look like a frontier lab. It won't be OpenAI and Anthropic take the whole world. It's that everyone's OpenAI and Anthropic now." San Francisco Compute

MTS

135,653 次观看 • 2 个月前

Introducing LiveKit Inference — a new cloud service that gives you access to the most popular voice AI models with just your LiveKit API key. We manage rate limits for you, report on usage, and consolidate billing. All LiveKit Cloud plans now include free monthly inference credits. A single string update allows you to call models from: AssemblyAI Deepgram Google DeepMind OpenAI Rime

Introducing LiveKit Inference — a new cloud service that gives you access to the most popular voice AI models with just your LiveKit API key. We manage rate limits for you, report on usage, and consolidate billing. All LiveKit Cloud plans now include free monthly inference credits. A single string update allows you to call models from: AssemblyAI Deepgram Google DeepMind OpenAI Rime

LiveKit

37,257 次观看 • 1 年前

I built an infinite AI chess game, powered by: ▪️ AI SDK for a single DX to all frontier models ▪️ AI Gateway for seamless access with 1 API key ▪️ Workflow to keep the game going forever Right now Anthropic is playing OpenAI:

I built an infinite AI chess game, powered by: ▪️ AI SDK for a single DX to all frontier models ▪️ AI Gateway for seamless access with 1 API key ▪️ Workflow to keep the game going forever Right now Anthropic is playing OpenAI:

Guillermo Rauch

246,500 次观看 • 10 个月前

Anthropic doesn't want you running local AI Ahmad explains why the labs are lying to you and how to run your own models instead

Anthropic doesn't want you running local AI Ahmad explains why the labs are lying to you and how to run your own models instead

David Ondrej

67,069 次观看 • 2 个月前