Loading video...

Video Failed to Load

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

MIT PhD student Alex Zhang on why simple harness abstractions beat models trained around their own harness: "When these abstractions are implemented in a really simple way, basically what you see with Prime Agent is that you can slot in any model. We don't train any model around Prime... show more

MTS

447,459 subscribers

148,427 views • 1 month ago •via X (Twitter)

Science & Technology Education

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

MIT PhD student Alex Zhang admits current AI models might already be AGI with the right harness around them: "To an extent, the answer is kind of yes. The strongest thing I can say is that we can do a lot more with the current models that we have." "Whether that means actually automating things, I'm not actually sure, but one of the obvious things that we find with even harnesses outside of Prime Agent is that harnesses can greatly increase the capabilities and generalization of models." "There's a lot of cases where we think models can't be used or they can't be applied for these settings, but the likely answer is that this actually isn't true. You probably can apply a model to the setting if you have the right harness." alex zhang Seth Karten Prime Intellect

MIT PhD student Alex Zhang admits current AI models might already be AGI with the right harness around them: "To an extent, the answer is kind of yes. The strongest thing I can say is that we can do a lot more with the current models that we have." "Whether that means actually automating things, I'm not actually sure, but one of the obvious things that we find with even harnesses outside of Prime Agent is that harnesses can greatly increase the capabilities and generalization of models." "There's a lot of cases where we think models can't be used or they can't be applied for these settings, but the likely answer is that this actually isn't true. You probably can apply a model to the setting if you have the right harness." alex zhang Seth Karten Prime Intellect

MTS

28,659 views • 1 month ago

MIT PhD student Alex Zhang reveals the scaling result where a model trained on short tasks generalizes to problems 100x longer for free: "If you're very clever about the design of your harness or how you use the language model, you can almost get scaling gains for free." "If you train a model naively, there's no tricks. It's just the same way you train a model on these RL environments. You just roll it out, and then you just get some reward." "If you train it on only short tasks, like only tasks that are 10,000 tokens long, and then you were to run it on a similar domain, but at a million tokens, or 10 million tokens, or 100,000 tokens, it generalizes really, really well. If you look at it compared to even the base transformer, you get way better generalization properties." "When the model uses an RLM (Recursive Language Model) after it's trained on these short tasks, it will see some kind of trajectory of actions that it does. Between these two problems of different lengths, the RLM learns to see them as almost the same problem." "Token for token, they're almost the same. You can describe it in code. In one code setting, maybe the for loop is a little bigger, but it's the same kind of code and it derives the constants from the data. There's no hard coding, so they literally look the same." alex zhang

MIT PhD student Alex Zhang reveals the scaling result where a model trained on short tasks generalizes to problems 100x longer for free: "If you're very clever about the design of your harness or how you use the language model, you can almost get scaling gains for free." "If you train a model naively, there's no tricks. It's just the same way you train a model on these RL environments. You just roll it out, and then you just get some reward." "If you train it on only short tasks, like only tasks that are 10,000 tokens long, and then you were to run it on a similar domain, but at a million tokens, or 10 million tokens, or 100,000 tokens, it generalizes really, really well. If you look at it compared to even the base transformer, you get way better generalization properties." "When the model uses an RLM (Recursive Language Model) after it's trained on these short tasks, it will see some kind of trajectory of actions that it does. Between these two problems of different lengths, the RLM learns to see them as almost the same problem." "Token for token, they're almost the same. You can describe it in code. In one code setting, maybe the for loop is a little bigger, but it's the same kind of code and it derives the constants from the data. There's no hard coding, so they literally look the same." alex zhang

MTS

100,519 views • 2 months ago

$DEEPSEEK JUST DROPPED HARNESS, AND IT ALREADY BECAME ONE OF THE FASTEST-GROWING PROJECTS ON GITHUB: 130,000 STARS IN ONLY 3 DAYS. COMPLETELY FREE Core idea: almost anything can become a plugin ➜ models, sessions, skills, sandboxes, loops, even the interface. Put the pieces in, get a working agent out 1\ Architecture is built on Cordis, so you can swap components at any time 2\ Formula is simple: any model + any component = ready agent 3\ Community already shipped 6,000 ready skills for it You get a flexible environment where you control the building blocks instead of being locked into one rigid setup Installation is simple. Run these commands one by one: ``` git clone cd deepseek-harness pnpm install pnpm run build pnpm dsh web ```$

DEEPSEEK JUST DROPPED HARNESS, AND IT ALREADY BECAME ONE OF THE FASTEST-GROWING PROJECTS ON GITHUB: 130,000 STARS IN ONLY 3 DAYS. COMPLETELY FREE Core idea: almost anything can become a plugin ➜ models, sessions, skills, sandboxes, loops, even the interface. Put the pieces in, get a working agent out 1\ Architecture is built on Cordis, so you can swap components at any time 2\ Formula is simple: any model + any component = ready agent 3\ Community already shipped 6,000 ready skills for it You get a flexible environment where you control the building blocks instead of being locked into one rigid setup Installation is simple. Run these commands one by one: ``` git clone cd deepseek-harness pnpm install pnpm run build pnpm dsh web ```

beamnxw ./

13,050 views • 1 month ago

Today is Day 11 of 12 Days of Agent Two. Agent Two isn't just Claude under the hood - we get asked this a lot. The truth is - while we use all the state of the art models out there, the Agent's real intelligence comes from its harness: a domain-specific layer built from continuous creative research that shapes how every model is actually used - more about this in the video. But this harness is what makes Agent Two frontier intelligence for creative work - at 4x the speed, and half the price.

Today is Day 11 of 12 Days of Agent Two. Agent Two isn't just Claude under the hood - we get asked this a lot. The truth is - while we use all the state of the art models out there, the Agent's real intelligence comes from its harness: a domain-specific layer built from continuous creative research that shapes how every model is actually used - more about this in the video. But this harness is what makes Agent Two frontier intelligence for creative work - at 4x the speed, and half the price.

Invideo

2,560,489 views • 1 month ago

Gave Prime Intellect's new self-improving agent an E2B sandbox and told it to play Factorio. Factorio is a long-horizon test of world models: the agent has to learn a new system, predict the effects of its actions, and adapt when its assumptions are wrong. That requires a persistent sandbox like E2B that can keep the game running and retain its state while the agent works. Spin up a lot of sandboxes and benchmark prime agent or run your own evals with E2B. Try it yourself: e2b sbx create prime

Gave Prime Intellect's new self-improving agent an E2B sandbox and told it to play Factorio. Factorio is a long-horizon test of world models: the agent has to learn a new system, predict the effects of its actions, and adapt when its assumptions are wrong. That requires a persistent sandbox like E2B that can keep the game running and retain its state while the agent works. Spin up a lot of sandboxes and benchmark prime agent or run your own evals with E2B. Try it yourself: e2b sbx create prime

Vasek Mlejnsky

23,085 views • 1 month ago

Shopify CEO Tobi Lütke called Pi "the most interesting agent harness" - a tiny, open-source framework built by a single developer that powers OpenClaw and is changing how builders think about AI agents. In this clip, Dave Killeen (Field CPO at Pendo) breaks down why Pi matters, how it compares to Manus, and why the real magic in AI isn't the language model itself - it's the orchestration layer around it. Pi is a lightweight agent harness that lets developers build smarter AI experiences by leveraging models like Claude, GPT, and Gemini through better orchestration, not bigger models. If you're building AI products, coding agents, or exploring the new world of harness engineering, this is essential context. We discuss why Manus gave users a far better experience than talking directly to an LLM, and why open-source tools like Pi could be the future of agent development.

Shopify CEO Tobi Lütke called Pi "the most interesting agent harness" - a tiny, open-source framework built by a single developer that powers OpenClaw and is changing how builders think about AI agents. In this clip, Dave Killeen (Field CPO at Pendo) breaks down why Pi matters, how it compares to Manus, and why the real magic in AI isn't the language model itself - it's the orchestration layer around it. Pi is a lightweight agent harness that lets developers build smarter AI experiences by leveraging models like Claude, GPT, and Gemini through better orchestration, not bigger models. If you're building AI products, coding agents, or exploring the new world of harness engineering, this is essential context. We discuss why Manus gave users a far better experience than talking directly to an LLM, and why open-source tools like Pi could be the future of agent development.

Aakash Gupta

10,774 views • 17 days ago

this is huge news from Cursor, they've pulled of the impossible and turned their ai-wrapper into a in-destructible moat sam altman literally called it cursor's ai agent harness is available for anyone to build on, which means ai models are now a commodity > 1-time install and now anyone can run cursor's agent locally or via cloud. use any model (e.g. gpt 5.5) but with the added cursor harness that makes it 10X better. > its so good that 3 of cursors biggest competitors are embedding it into their products. > now cursor DOESN'T DEPEND on anthropic or openai. their own model (composer 2) competes directly! yesterday sam altman said the ai model and harness are one and the same and today cursor turned their harness into a self-owned moat fucking masterclass (coming from a former cursor / ai wrapper hater)

this is huge news from Cursor, they've pulled of the impossible and turned their ai-wrapper into a in-destructible moat sam altman literally called it cursor's ai agent harness is available for anyone to build on, which means ai models are now a commodity > 1-time install and now anyone can run cursor's agent locally or via cloud. use any model (e.g. gpt 5.5) but with the added cursor harness that makes it 10X better. > its so good that 3 of cursors biggest competitors are embedding it into their products. > now cursor DOESN'T DEPEND on anthropic or openai. their own model (composer 2) competes directly! yesterday sam altman said the ai model and harness are one and the same and today cursor turned their harness into a self-owned moat fucking masterclass (coming from a former cursor / ai wrapper hater)

Ejaaz

950,452 views • 5 months ago

A lot of people are calling Hermes Agent the end of OpenClaw. BRUH! It's not... Nous Research trains actual models and they built an agent around that expertise. The local model routing is solid, but the part that matters for your business is that your conversations become fine-tuning data. You can train a model on how you actually work. 00:00 The Problem with Local AI Models 00:25 Introduction to Nous Research 01:04 Cross-Platform Agent Capabilities 01:44 Deep Local Model Integration 02:30 Routing Tasks to Different Models 03:06 Conversations as Training Data 03:50 Hermes Agent vs. OpenClaw 04:15 Future Plans and Series Overview

A lot of people are calling Hermes Agent the end of OpenClaw. BRUH! It's not... Nous Research trains actual models and they built an agent around that expertise. The local model routing is solid, but the part that matters for your business is that your conversations become fine-tuning data. You can train a model on how you actually work. 00:00 The Problem with Local AI Models 00:25 Introduction to Nous Research 01:04 Cross-Platform Agent Capabilities 01:44 Deep Local Model Integration 02:30 Routing Tasks to Different Models 03:06 Conversations as Training Data 03:50 Hermes Agent vs. OpenClaw 04:15 Future Plans and Series Overview

Ray Fernando

42,398 views • 6 months ago

OpenAI researcher David Holtz reveals workers who don't adapt their prompting to newer models leave 50% of the alpha on the table because half of AI's value comes from how you interact with it: "We ran a lab experiment. We give people the same task to do, but we give them models of different capabilities. How much of the improvement when we give people more capable models comes from the fact that we gave them a better model vs the fact that they actually update and adapt the way they prompt and interact with these models?" "For tasks where precision matters, where you have some output you're trying to get and you're trying to steer the model, which is different from a creative task like brainstorming." "For these steering-based tasks, when you give people a better model, about 50% of the improvement we see in how well they're able to complete the task comes from changes in how they interact with the model." "If you just kept the way you interact with the model static and just made the capabilities better and better, you'd leave 50% of the alpha basically on the table." OpenAI

OpenAI researcher David Holtz reveals workers who don't adapt their prompting to newer models leave 50% of the alpha on the table because half of AI's value comes from how you interact with it: "We ran a lab experiment. We give people the same task to do, but we give them models of different capabilities. How much of the improvement when we give people more capable models comes from the fact that we gave them a better model vs the fact that they actually update and adapt the way they prompt and interact with these models?" "For tasks where precision matters, where you have some output you're trying to get and you're trying to steer the model, which is different from a creative task like brainstorming." "For these steering-based tasks, when you give people a better model, about 50% of the improvement we see in how well they're able to complete the task comes from changes in how they interact with the model." "If you just kept the way you interact with the model static and just made the capabilities better and better, you'd leave 50% of the alpha basically on the table." OpenAI

MTS

15,590 views • 1 month ago

Anthropic Technical Staff Jess Yan says the harness and the model cannot be pulled apart without giving up performance, though she flags her own bias. "I think I am quite biased but I also think that it is impossible to get the maximum possible performance without tying together the harness and the model." "Now the components of the harness and maybe the thickness of the harness will change over time as models get more and more capable." "However, when we test our models and when we are assessing their performance we always have to test it in conjunction with a harness." "And are we going to test it with all the different harnesses of the world? We're going to select the harnesses that we have built." "And so there is an aspect of the necessity of building models is that you have to be testing them with harnesses, and that sort of keeps them paired together."

Anthropic Technical Staff Jess Yan says the harness and the model cannot be pulled apart without giving up performance, though she flags her own bias. "I think I am quite biased but I also think that it is impossible to get the maximum possible performance without tying together the harness and the model." "Now the components of the harness and maybe the thickness of the harness will change over time as models get more and more capable." "However, when we test our models and when we are assessing their performance we always have to test it in conjunction with a harness." "And are we going to test it with all the different harnesses of the world? We're going to select the harnesses that we have built." "And so there is an aspect of the necessity of building models is that you have to be testing them with harnesses, and that sort of keeps them paired together."

Fireside Alpha

584,074 views • 1 month ago

Small Language Models (SML) are the future of AI. "Small" (SML) instead of "Large" (LLM). These small models are highly specialized models with superhuman abilities on specific tasks. Here are two techniques to build these models: • Spectrum • Model Merging I give you a short introduction in the attached video, but here is a quick summary: Spectrum helps us identify the most relevant layers to solve one specific task. We can ignore everything else and focus on fine-tuning these layers. Using Spectrum, we can fine-tune models in a heartbeat. Model Merging combines multiple models into a unique, much better model than any of the individual input models. You can also combine models specialized in different tasks and get a model with multiple abilities. This is the state of the art of productizing models. It's what Arcee.ai's platform does behind the scenes. Arcee collaborated with me on this post and is sponsoring it. There are three main steps to produce a model for your particular use case: 1. You create a dataset by uploading your data. 2. You train a model. At this step, Arcee uses Spectrum and Model Merging to produce a highly specialized model for your task. 3. You can deploy that model to any environment you want. Three important notes: • Training process is 2x faster and 2x cheaper than regular fine-tuning. • Resultant models are smaller and have higher accuracy. • They create these specialized models from open-source models. Check this site so you can fully appreciate how this works: If you want to fine-tune an open-source model, consider Arcee's platform. This is the state of the art.

Small Language Models (SML) are the future of AI. "Small" (SML) instead of "Large" (LLM). These small models are highly specialized models with superhuman abilities on specific tasks. Here are two techniques to build these models: • Spectrum • Model Merging I give you a short introduction in the attached video, but here is a quick summary: Spectrum helps us identify the most relevant layers to solve one specific task. We can ignore everything else and focus on fine-tuning these layers. Using Spectrum, we can fine-tune models in a heartbeat. Model Merging combines multiple models into a unique, much better model than any of the individual input models. You can also combine models specialized in different tasks and get a model with multiple abilities. This is the state of the art of productizing models. It's what Arcee.ai's platform does behind the scenes. Arcee collaborated with me on this post and is sponsoring it. There are three main steps to produce a model for your particular use case: 1. You create a dataset by uploading your data. 2. You train a model. At this step, Arcee uses Spectrum and Model Merging to produce a highly specialized model for your task. 3. You can deploy that model to any environment you want. Three important notes: • Training process is 2x faster and 2x cheaper than regular fine-tuning. • Resultant models are smaller and have higher accuracy. • They create these specialized models from open-source models. Check this site so you can fully appreciate how this works: If you want to fine-tune an open-source model, consider Arcee's platform. This is the state of the art.

Santiago

164,162 views • 2 years ago

“don’t train your own model” is common ai advice. it's wrong. your token bill's the proof. today, we’re excited to launch castform into open preview. castform is the easiest way for you to train your own model, on your own data. open-weights models are performant and much cheaper. when trained on your task & proprietary data, they beat closed models. the thing standing between you and that was weeks of plumbing & years of ml expertise. with castform, model training is as simple as prompt engineering. castform bring your agent traces or raw corpora. castform turns it into training data, picks the right algorithmic recipes, manages gpus, and gives you an ide to watch and chat with your model as it learns. see what you can build with castform👇

“don’t train your own model” is common ai advice. it's wrong. your token bill's the proof. today, we’re excited to launch castform into open preview. castform is the easiest way for you to train your own model, on your own data. open-weights models are performant and much cheaper. when trained on your task & proprietary data, they beat closed models. the thing standing between you and that was weeks of plumbing & years of ml expertise. with castform, model training is as simple as prompt engineering. castform bring your agent traces or raw corpora. castform turns it into training data, picks the right algorithmic recipes, manages gpus, and gives you an ide to watch and chat with your model as it learns. see what you can build with castform👇

girish

462,836 views • 3 months ago

Perplexity CEO Aravind Srinivas on the brutal truth about who actually makes money in AI (and why it's not who you think): Aravind argues that the real value in AI comes from orchestration. He points to products like Codex, Claude Code, and Perplexity Computer: "What is that? It's an orchestration system. It takes a model, pairs it with an agent harness." And what is an agent harness? "The simplest way of describing it is like rules for how the agent loop should run. What are all the skills and sub-agents and connectors and tools it accesses? Without the harness, you don't necessarily capture and convert the intrinsic intelligence in the model into valuable output tokens." This leads to a blunt conclusion about who has a real business in AI, and who doesn't: "If you're literally just a reseller of model tokens, you have no business, because the model will get commoditized. So even if you're a model builder, you don't have a business. As an infra layer, you have some business on serving those output tokens. But as an application layer or model builder, you don't really have a business if you're just a reseller of tokens that come directly out of the model." So where does the value accrue? "You have a business if you know how to take the model, ground it in valuable context, orchestrate it with a really good agent harness, connected to the right set of tools and connectors (whether it's personal connectors or business connectors) and provide the experience to people in one single unified system." Aravind Srinivas then explains Perplexity's specific edge: Beyond orchestrating across tools, files, and connectors, they also orchestrate across models. "That is the differentiation that Anthropic and OpenAI cannot claim, because you wouldn't find GPT-5 inside the Claude Code harness. You wouldn't find Claude Opus inside the Codex harness. These are competing with each other. Whereas you would find both these models inside Perplexity Computer." Why does this matter? Because it all comes down to power. In Aravind's framing, the fundamental cost driver in AI is watts (the one input nobody can subsidize except the government). "Whoever provides the most valuable output tokens with the least amount of power expended to produce them generates the greatest value to the end user, has the most pricing power, has the most value. That is the orchestration problem to solve." His conclusion: "The one single most important metric in AI is token value per watt per user."

Perplexity CEO Aravind Srinivas on the brutal truth about who actually makes money in AI (and why it's not who you think): Aravind argues that the real value in AI comes from orchestration. He points to products like Codex, Claude Code, and Perplexity Computer: "What is that? It's an orchestration system. It takes a model, pairs it with an agent harness." And what is an agent harness? "The simplest way of describing it is like rules for how the agent loop should run. What are all the skills and sub-agents and connectors and tools it accesses? Without the harness, you don't necessarily capture and convert the intrinsic intelligence in the model into valuable output tokens." This leads to a blunt conclusion about who has a real business in AI, and who doesn't: "If you're literally just a reseller of model tokens, you have no business, because the model will get commoditized. So even if you're a model builder, you don't have a business. As an infra layer, you have some business on serving those output tokens. But as an application layer or model builder, you don't really have a business if you're just a reseller of tokens that come directly out of the model." So where does the value accrue? "You have a business if you know how to take the model, ground it in valuable context, orchestrate it with a really good agent harness, connected to the right set of tools and connectors (whether it's personal connectors or business connectors) and provide the experience to people in one single unified system." Aravind Srinivas then explains Perplexity's specific edge: Beyond orchestrating across tools, files, and connectors, they also orchestrate across models. "That is the differentiation that Anthropic and OpenAI cannot claim, because you wouldn't find GPT-5 inside the Claude Code harness. You wouldn't find Claude Opus inside the Codex harness. These are competing with each other. Whereas you would find both these models inside Perplexity Computer." Why does this matter? Because it all comes down to power. In Aravind's framing, the fundamental cost driver in AI is watts (the one input nobody can subsidize except the government). "Whoever provides the most valuable output tokens with the least amount of power expended to produce them generates the greatest value to the end user, has the most pricing power, has the most value. That is the orchestration problem to solve." His conclusion: "The one single most important metric in AI is token value per watt per user."

Big Brain AI

42,540 views • 1 month ago

The easiest way to run your agent harnesses using local models: Whether you're using Claude Code, Codex, OpenCode, or Pi, the local setup itself isn't really the hard part. The harder problem is figuring out which local model your machine can actually handle well. You have to think about RAM, model size, quantization, context length, KV cache, speed, accuracy, and a bunch of other trade-offs before you even start. Magnitude is an open-source solution that removes all of that guesswork for you. It profiles your machine, benchmarks what it can realistically run, recommends the best models for your hardware, and then lets you connect them to your preferred agent harness. The entire setup takes just two commands. In this video, I walk through the full process from hardware profiling to running a harness on a local model. Chapters: 00:00 Intro 00:17 What Magnitude is 00:45 Setup in two commands 01:29 Why picking a local model gets confusing fast 02:11 Hardware profiling + model recommendations 02:59 Choosing your harness 03:22 Connecting a harness that's already running 04:18 Outro Get started: (don't forget to star 🌟)

The easiest way to run your agent harnesses using local models: Whether you're using Claude Code, Codex, OpenCode, or Pi, the local setup itself isn't really the hard part. The harder problem is figuring out which local model your machine can actually handle well. You have to think about RAM, model size, quantization, context length, KV cache, speed, accuracy, and a bunch of other trade-offs before you even start. Magnitude is an open-source solution that removes all of that guesswork for you. It profiles your machine, benchmarks what it can realistically run, recommends the best models for your hardware, and then lets you connect them to your preferred agent harness. The entire setup takes just two commands. In this video, I walk through the full process from hardware profiling to running a harness on a local model. Chapters: 00:00 Intro 00:17 What Magnitude is 00:45 Setup in two commands 01:29 Why picking a local model gets confusing fast 02:11 Hardware profiling + model recommendations 02:59 Choosing your harness 03:22 Connecting a harness that's already running 04:18 Outro Get started: (don't forget to star 🌟)

Akshay 🚀

18,722 views • 13 days ago

The founder of LangChain says both models and harnesses have gotten really good between December and now. According to Harrison Chase, the core idea of an agent before Christmas was a model running in a loop and calling tools. This had been the north star for 3 years. - langchain had this when it launched - autogpt was the same idea - openclaw is kind of a future version of it Then about a year ago, they started getting really good. Claude Code, Manus, and Deep Research were all launched around the same time. All of them use the same pattern: running in a loop with harnesses (planning tools, file systems, code execution, etc) Harness engineering became a thing. Then Opus came out in November and really unlocked it. - the harness let the model do more and more - less hardcoded logic - way more control Then everyone went on vacation, played around, and realized that the model and the harness finally worked reliably.

The founder of LangChain says both models and harnesses have gotten really good between December and now. According to Harrison Chase, the core idea of an agent before Christmas was a model running in a loop and calling tools. This had been the north star for 3 years. - langchain had this when it launched - autogpt was the same idea - openclaw is kind of a future version of it Then about a year ago, they started getting really good. Claude Code, Manus, and Deep Research were all launched around the same time. All of them use the same pattern: running in a loop with harnesses (planning tools, file systems, code execution, etc) Harness engineering became a thing. Then Opus came out in November and really unlocked it. - the harness let the model do more and more - less hardcoded logic - way more control Then everyone went on vacation, played around, and realized that the model and the harness finally worked reliably.

Ivan Burazin

33,948 views • 5 months ago

We are releasing 4M-21 with a permissive license, including its source code and trained models. It's a pretty effective multimodal model that solves 10s of tasks & modalities. See the demo code, sample results, and the tokenizers of diverse modalities on the website. IMO, the multitask learning aspect of multimodal models has really taken a step forward. We can train a single model on many diverse tasks with ~SOTA accuracy. But a long way to go in terms of transfer/emergence. 🌐 ⌨️ Joint work w/ EPFL Apple.

We are releasing 4M-21 with a permissive license, including its source code and trained models. It's a pretty effective multimodal model that solves 10s of tasks & modalities. See the demo code, sample results, and the tokenizers of diverse modalities on the website. IMO, the multitask learning aspect of multimodal models has really taken a step forward. We can train a single model on many diverse tasks with ~SOTA accuracy. But a long way to go in terms of transfer/emergence. 🌐 ⌨️ Joint work w/ EPFL Apple.

Amir Zamir

69,604 views • 2 years ago

Harnesses often get dismissed as just scaffolding, just prompt engineering, and not real research. But that couldn't be farther from the truth. The same model weights that score 30% on ARC-AGI score 95% with a better harness. So we gathered a group of researchers and founders working at the frontier to do a deep dive into the state of harnesses. We cover how we got to this point, the case for making your harness as expressive as possible, and what YC learned building an agent for every employee in the company. 00:00 - Francois Chaubard: Why harnesses matter 04:27 - Building an auto-researcher by accident 07:13 - A five minute history of harnesses 13:56 - Self-improving harnesses 18:35 - Seth Karten: Prime Agent, a self-improving RLM harness 21:50 - Context as an L1, L2, L3 cache 24:51 - From Turing machine to von Neumann computer 28:33 - Messaging between agents 30:04 - ARC-AGI results 33:09 - Emulator Bench and GPU kernels 37:30 - Jon Saad-Falcon: OpenJarvis, personal AI on personal devices 38:26 - How far behind are local models 39:21 - The five primitives of a personal AI stack 42:47 - Letting cloud models optimize your local stack 43:53 - 800x cheaper than the cloud 45:58 - Josh France and Regan Bell: QM, YC's agent harness for work 47:29 - A history of YC's internal agents 49:24 - OpenClaw and a fleet of 50 agents 51:04 - Pulling the brain out of the sandbox 54:43 - Letting the agent choose its own sandbox and model 57:16 - The grind tool: budgets on goals 58:50 - Agents don't understand social context

Harnesses often get dismissed as just scaffolding, just prompt engineering, and not real research. But that couldn't be farther from the truth. The same model weights that score 30% on ARC-AGI score 95% with a better harness. So we gathered a group of researchers and founders working at the frontier to do a deep dive into the state of harnesses. We cover how we got to this point, the case for making your harness as expressive as possible, and what YC learned building an agent for every employee in the company. 00:00 - Francois Chaubard: Why harnesses matter 04:27 - Building an auto-researcher by accident 07:13 - A five minute history of harnesses 13:56 - Self-improving harnesses 18:35 - Seth Karten: Prime Agent, a self-improving RLM harness 21:50 - Context as an L1, L2, L3 cache 24:51 - From Turing machine to von Neumann computer 28:33 - Messaging between agents 30:04 - ARC-AGI results 33:09 - Emulator Bench and GPU kernels 37:30 - Jon Saad-Falcon: OpenJarvis, personal AI on personal devices 38:26 - How far behind are local models 39:21 - The five primitives of a personal AI stack 42:47 - Letting cloud models optimize your local stack 43:53 - 800x cheaper than the cloud 45:58 - Josh France and Regan Bell: QM, YC's agent harness for work 47:29 - A history of YC's internal agents 49:24 - OpenClaw and a fleet of 50 agents 51:04 - Pulling the brain out of the sandbox 54:43 - Letting the agent choose its own sandbox and model 57:16 - The grind tool: budgets on goals 58:50 - Agents don't understand social context

Y Combinator

491,785 views • 21 days ago

MICROSOFT JUST OPEN SOURCED THE ENTIRE AGENT RUNTIME, SO YOU NEVER HAVE TO BUILD ONE AGAIN a raw model only writes text. to make it a real agent it needs a loop, tools, memory, planning, approvals, and a way to know when it is done. Microsoft's Agent Harness is that whole layer, and it ships with everything on by default. you bring the model and the tools, it wires up the rest: > an automatic tool calling loop with iteration limits > history saved after every call, so a crash resumes instead of starting over > context compaction that summarizes when the window fills > a persistent todo list plus plan and execute modes > file memory that survives across turns > tool approval with standing rules and auto approval > OpenTelemetry tracing and web search built in > Agent Skills that load domain expertise on demand the mental model is simple: the model is the brain, the harness is the body around it. and it works with whatever you already run: > Azure OpenAI, OpenAI, Anthropic, and GitHub Copilot, and it hosts on Microsoft Foundry you spin one up in a single line: agent = create_harness_agent(client, name, instructions) open source, Python, .NET, and Go, stable since mid 2026. bookmark this. Microsoft already built the hard part of every AI agent, you just plug in the model.

MICROSOFT JUST OPEN SOURCED THE ENTIRE AGENT RUNTIME, SO YOU NEVER HAVE TO BUILD ONE AGAIN a raw model only writes text. to make it a real agent it needs a loop, tools, memory, planning, approvals, and a way to know when it is done. Microsoft's Agent Harness is that whole layer, and it ships with everything on by default. you bring the model and the tools, it wires up the rest: > an automatic tool calling loop with iteration limits > history saved after every call, so a crash resumes instead of starting over > context compaction that summarizes when the window fills > a persistent todo list plus plan and execute modes > file memory that survives across turns > tool approval with standing rules and auto approval > OpenTelemetry tracing and web search built in > Agent Skills that load domain expertise on demand the mental model is simple: the model is the brain, the harness is the body around it. and it works with whatever you already run: > Azure OpenAI, OpenAI, Anthropic, and GitHub Copilot, and it hosts on Microsoft Foundry you spin one up in a single line: agent = create_harness_agent(client, name, instructions) open source, Python, .NET, and Go, stable since mid 2026. bookmark this. Microsoft already built the hard part of every AI agent, you just plug in the model.

NO1ennn

69,163 views • 1 month ago

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

kwindla

331,698 views • 1 month ago

.L3Harris says Palantir helped them beat frontier AI models in less than 2 days, at 95% lower cost: "When we fine-tuned open source models trained on our own data, we were able to outperform the frontier models in less than 48 hours." "The cost of our fine-tuned open source model was 95% lower than the frontier models we were using." "AI is a commodity. It's all about the data. We view our data as a corporate asset. It's our unique hard-earned knowledge." "We believe American defense companies should not be a vassal for frontier AI labs, handing over our data and institutional knowledge, hoping to rent back the intelligence it creates." " We own the model, we own the compute, we own the advantage."

.L3Harris says Palantir helped them beat frontier AI models in less than 2 days, at 95% lower cost: "When we fine-tuned open source models trained on our own data, we were able to outperform the frontier models in less than 48 hours." "The cost of our fine-tuned open source model was 95% lower than the frontier models we were using." "AI is a commodity. It's all about the data. We view our data as a corporate asset. It's our unique hard-earned knowledge." "We believe American defense companies should not be a vassal for frontier AI labs, handing over our data and institutional knowledge, hoping to rent back the intelligence it creates." " We own the model, we own the compute, we own the advantage."

Jawwwn

1,103,566 views • 18 days ago