Загрузка видео...

Не удалось загрузить видео

На главную

we built blazing fast computer/browser use with Jev + Stagehand. this task cost $0.001 and executed at near instant speed (in a remote browser btw) the loop: observe the page, send a11y tree as state + actions as questions, Jev decides the next action, then Stagehand executes it.

kylejeong's profile picture

Kyle Jeong

8,084 subscribers

63,359 просмотров • 1 день назад •via X (Twitter)

Наука и технологии
LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

We built a plugin that gives Jev a browser in Cline, and have been blown away by the results. 1. Install it in our new desktop app: Customize > Marketplace > Plugins > search 'jev-browser' 2. Create a Vercel AI Gateway API key, then save it to ~/.cline/plugins/cline-jev-browser.config.json as {"gateway": {"apiKey": "..."}} and restart Cline. 3. Ask any browser task and it will launch Chrome in the background to complete it.
0:17

We built a plugin that gives Jev a browser in Cline, and have been blown away by the results. 1. Install it in our new desktop app: Customize > Marketplace > Plugins > search 'jev-browser' 2. Create a Vercel AI Gateway API key, then save it to ~/.cline/plugins/cline-jev-browser.config.json as {"gateway": {"apiKey": "..."}} and restart Cline. 3. Ask any browser task and it will launch Chrome in the background to complete it.

Cline

54,259 просмотров • 17 часов назад

Built a tax document classifier with Jev. We ingest thousands of tax documents using an LLM pipeline I built last tax season. I read multiple articles as late as April this year claiming AI fails at tax document classification. Wasn't the case back then, and it's proved wrong again now. Jev classifies 100% of our tax document corpus at $0.001 per page. 34x cheaper and 6x faster than the LLM setup. Open sourcing it:
0:12

Built a tax document classifier with Jev. We ingest thousands of tax documents using an LLM pipeline I built last tax season. I read multiple articles as late as April this year claiming AI fails at tax document classification. Wasn't the case back then, and it's proved wrong again now. Jev classifies 100% of our tax document corpus at $0.001 per page. 34x cheaper and 6x faster than the LLM setup. Open sourcing it:

Nakshatra Saxena

184,610 просмотров • 23 часов назад

Jev + Kimi K3 for fraud detection! TLDR: Jev classified 100 emails in 1.42 seconds, then I routed the uncertain cases to Kimi K3. The full pipeline got 96/100 correct for only ~$0.07. Video is not sped up, check out the live run! Here was my process: I gave Jev 100 emails to classify (a mix of 50 legit & 50 fraudelent emails). It classified all of them in 1.42 seconds. An underrated feature about Jev is it will give you the confidence score for a classification, so I routed any prediction under 95% confidence to Kimi K3 to be fully sure. 31 emails fell below that threshold. After routing those to Kimi K3, the combined pipeline reached 96% accuracy. The full run took 16 seconds & ~$0.07 in inference costs: - $0.068 from Kimi K3 on Together AI - $0.003 (1/3 of a cent) from Jev on TypeSafe AI. I think this is a really interesting pattern: use a fast specialized model like Jev for the narrow task, then route the uncertain cases to a larger LLM. I feel like this kind of approach could be a game changer for use cases like fraud or anything realtime. You can use the speed & low cost of Jev while having a larger LLM as a fallback to ensure high accuracy.
0:12

Jev + Kimi K3 for fraud detection! TLDR: Jev classified 100 emails in 1.42 seconds, then I routed the uncertain cases to Kimi K3. The full pipeline got 96/100 correct for only ~$0.07. Video is not sped up, check out the live run! Here was my process: I gave Jev 100 emails to classify (a mix of 50 legit & 50 fraudelent emails). It classified all of them in 1.42 seconds. An underrated feature about Jev is it will give you the confidence score for a classification, so I routed any prediction under 95% confidence to Kimi K3 to be fully sure. 31 emails fell below that threshold. After routing those to Kimi K3, the combined pipeline reached 96% accuracy. The full run took 16 seconds & ~$0.07 in inference costs: - $0.068 from Kimi K3 on Together AI - $0.003 (1/3 of a cent) from Jev on TypeSafe AI. I think this is a really interesting pattern: use a fast specialized model like Jev for the narrow task, then route the uncertain cases to a larger LLM. I feel like this kind of approach could be a game changer for use cases like fraud or anything realtime. You can use the speed & low cost of Jev while having a larger LLM as a fallback to ensure high accuracy.

Hassan

52,368 просмотров • 1 день назад

The Shopify Live Globe is back this Black Friday! And we built a pinball game out of it. It runs at 120fps in a browser with a full 3d environment, physics engine, VR support and more. 👀
0:15

The Shopify Live Globe is back this Black Friday! And we built a pinball game out of it. It runs at 120fps in a browser with a full 3d environment, physics engine, VR support and more. 👀

Daniel Beauchamp

84,002 просмотров • 9 месяцев назад

We are investing in the frontiers of agentic capabilities with a few early prototypes. Project Mariner is built with Gemini 2.0 and is able to understand and reason across information - pixels, text, code, images + forms - on your browser screen, and then uses that info to complete tasks for you. When evaluated against the WebVoyager benchmark, it achieved a state-of-the-art result of 83.5% working as a single agent setup.
0:16

We are investing in the frontiers of agentic capabilities with a few early prototypes. Project Mariner is built with Gemini 2.0 and is able to understand and reason across information - pixels, text, code, images + forms - on your browser screen, and then uses that info to complete tasks for you. When evaluated against the WebVoyager benchmark, it achieved a state-of-the-art result of 83.5% working as a single agent setup.

Sundar Pichai

219,170 просмотров • 1 год назад

Welcome to 2025! This year, we’re making major progress with the tomi SuperApp, putting more power in your hands: ✅ Q1: Chat with built-in payments - send crypto as easily as a message. ✅ Q2: Encrypted private groups to keep your conversations secure. ✅ Q3: Customizable profiles to manage your digital identity your way. ✅ Q4: Decentralized browser + VPN for private and open web access. And it doesn’t stop there - additional developments are already planned, as shown in the animation below. The future of the internet is taking shape, and it’s just the beginning.
0:12

Welcome to 2025! This year, we’re making major progress with the tomi SuperApp, putting more power in your hands: ✅ Q1: Chat with built-in payments - send crypto as easily as a message. ✅ Q2: Encrypted private groups to keep your conversations secure. ✅ Q3: Customizable profiles to manage your digital identity your way. ✅ Q4: Decentralized browser + VPN for private and open web access. And it doesn’t stop there - additional developments are already planned, as shown in the animation below. The future of the internet is taking shape, and it’s just the beginning.

TOMI

38,478 просмотров • 1 год назад

Next time when someone asks how fast $kas is or they are saying other PoW/PoS are faster, give them this link I've created this small tool to showcase the speed of Kaspa versus other PoWs or PoSs. There is a button there that allows u to toggle between 1bps and 10bps. It didn't make sense to add the 100 bps option as it was too fast even for my browser :) Enjoy and long live $kas
0:10

Next time when someone asks how fast $kas is or they are saying other PoW/PoS are faster, give them this link I've created this small tool to showcase the speed of Kaspa versus other PoWs or PoSs. There is a button there that allows u to toggle between 1bps and 10bps. It didn't make sense to add the 100 bps option as it was too fast even for my browser :) Enjoy and long live $kas

Kaspador

49,013 просмотров • 1 год назад

Today we’re introducing Copilot Mode in Edge, our first step in reinventing the browser for the AI age. My favorite feature is multi-tab RAG. You can use Copilot to analyze your open tabs, like I do here with papers our team has published in nature journals over the last year. And there is a lot more to come, including built-in actions so you can delegate tasks as you browse.
0:19

Today we’re introducing Copilot Mode in Edge, our first step in reinventing the browser for the AI age. My favorite feature is multi-tab RAG. You can use Copilot to analyze your open tabs, like I do here with papers our team has published in nature journals over the last year. And there is a lot more to come, including built-in actions so you can delegate tasks as you browse.

Satya Nadella

705,461 просмотров • 1 год назад

What is the best video editing agent for short form social? Does it actually work? We watched professional video editors, step by step, as they built short-form social reels in Adobe Premiere Pro. Today we're open-sourcing this preview dataset on Hugging Face, to make AI agents better at editing videos. The data set is 234 annotated steps across 4 computer-use trajectories. Editors narrated their reasoning aloud as they worked, so every step pairs a screenshot with the expert's own thought, a structured action, and executable grounding: >a Premiere MCP tool call, keyboard shortcut, menu path, or coordinate click. >The format follows the AgentNet trajectory schema, extended with a Premiere action taxonomy and multi-path execution. ***That makes it directly usable for computer-use agent SFT, reasoning mid-training, tool-use and function calling, and benchmarking agents against a human expert baseline. Enjoy!
0:13

What is the best video editing agent for short form social? Does it actually work? We watched professional video editors, step by step, as they built short-form social reels in Adobe Premiere Pro. Today we're open-sourcing this preview dataset on Hugging Face, to make AI agents better at editing videos. The data set is 234 annotated steps across 4 computer-use trajectories. Editors narrated their reasoning aloud as they worked, so every step pairs a screenshot with the expert's own thought, a structured action, and executable grounding: >a Premiere MCP tool call, keyboard shortcut, menu path, or coordinate click. >The format follows the AgentNet trajectory schema, extended with a Premiere action taxonomy and multi-path execution. ***That makes it directly usable for computer-use agent SFT, reasoning mid-training, tool-use and function calling, and benchmarking agents against a human expert baseline. Enjoy!

ben

40,140 просмотров • 2 месяцев назад

THIS GUY VIBE CODED A BROWSER TOOL WITH CLAUDE CODE THAT DRAWS PROPERTY LINES ONTO DRONE FOOTAGE AND KEEPS THEM LOCKED TO THE GROUND AS THE DRONE MOVES real estate agents kept asking for those property boundary overlays on drone videos so buyers can see exactly where a lot ends. the only real option was after effects, which is a massive pro tool for one small repeatable job so he built a browser tool that does just that one thing > drop in a drone clip and draw the parcel outline once on the first frame > gpu motion tracking keeps that boundary stuck to the ground as the drone flies and the perspective shifts > if the line drifts anywhere, you drop a keyframe to correct it and it re-tracks the segments around it > style the line however you want, then export the finished video > free to use, you dont even need an account until you export he built one tool that solves one annoying job perfectly, for people who were never going to learn after effects anyway
0:13

THIS GUY VIBE CODED A BROWSER TOOL WITH CLAUDE CODE THAT DRAWS PROPERTY LINES ONTO DRONE FOOTAGE AND KEEPS THEM LOCKED TO THE GROUND AS THE DRONE MOVES real estate agents kept asking for those property boundary overlays on drone videos so buyers can see exactly where a lot ends. the only real option was after effects, which is a massive pro tool for one small repeatable job so he built a browser tool that does just that one thing > drop in a drone clip and draw the parcel outline once on the first frame > gpu motion tracking keeps that boundary stuck to the ground as the drone flies and the perspective shifts > if the line drifts anywhere, you drop a keyframe to correct it and it re-tracks the segments around it > style the line however you want, then export the finished video > free to use, you dont even need an account until you export he built one tool that solves one annoying job perfectly, for people who were never going to learn after effects anyway

Om Patel

404,227 просмотров • 1 месяц назад

Getting a bunch of questions about this! It's a full gpu rendering pipeline with built in shader editor for Rive, which can interleave 2D + 3D. Think WebGPU (minus compute, for now) that runs anywhere, not just the browser. You can implement anything at all. Cell shading. Background blurs. Deformation effects. And yes, even Rive vector components as textures. Shaders are precompiled and ship with your .riv. Minimal runtime size impact. Still WIP.
0:13

Getting a bunch of questions about this! It's a full gpu rendering pipeline with built in shader editor for Rive, which can interleave 2D + 3D. Think WebGPU (minus compute, for now) that runs anywhere, not just the browser. You can implement anything at all. Cell shading. Background blurs. Deformation effects. And yes, even Rive vector components as textures. Shaders are precompiled and ship with your .riv. Minimal runtime size impact. Still WIP.

Guido Rosso

22,346 просмотров • 5 месяцев назад

I built a clone of the Yeezy store with Next.js. This was a fun challenge — the site has some smooth animations and feels very fast. But it was bothering me that I couldn't use the browser back button. Can we do better? So I rebuilt the site with v0 and Motion. Here's how it works: 1. When you click on a product, Motion is able to animate the original position of the product in the grid, to the zoomed in product detail page. 2. During this transition, we also shallow update the URL with the `/p/slug` route for the page. 3. If you press the back button in the navbar, or use the browser back button, or press escape — all options will take you back to the main product listing page. 4. If you reload the page while looking at a product, or someone sends you a link to a specific product, it still works! This is the best parts of a SPA and MPA mixed together. In the future, I can make this even better with View Transitions (I wasn't able to get the product animation just right, but if you can I'd love to see it!). I also took some creative liberties from the original design. The whole 1/2/3 size thing, where you needed to click the "?" to see SM/MD/LG was strange, so I just went directly to those sizes. Similarly, I prefered the more traditional style sheet/modal with the background color change, versus the full screen takeover. If you wanted to actually hook this up to Shopify now, you can swap the cart implementation with Next.js Commerce, which has all the APIs you need + optimistic writes 🔥 Should I make a video walking through the code?
0:16

I built a clone of the Yeezy store with Next.js. This was a fun challenge — the site has some smooth animations and feels very fast. But it was bothering me that I couldn't use the browser back button. Can we do better? So I rebuilt the site with v0 and Motion. Here's how it works: 1. When you click on a product, Motion is able to animate the original position of the product in the grid, to the zoomed in product detail page. 2. During this transition, we also shallow update the URL with the `/p/slug` route for the page. 3. If you press the back button in the navbar, or use the browser back button, or press escape — all options will take you back to the main product listing page. 4. If you reload the page while looking at a product, or someone sends you a link to a specific product, it still works! This is the best parts of a SPA and MPA mixed together. In the future, I can make this even better with View Transitions (I wasn't able to get the product animation just right, but if you can I'd love to see it!). I also took some creative liberties from the original design. The whole 1/2/3 size thing, where you needed to click the "?" to see SM/MD/LG was strange, so I just went directly to those sizes. Similarly, I prefered the more traditional style sheet/modal with the background color change, versus the full screen takeover. If you wanted to actually hook this up to Shopify now, you can swap the cart implementation with Next.js Commerce, which has all the APIs you need + optimistic writes 🔥 Should I make a video walking through the code?

Lee Robinson

148,937 просмотров • 1 год назад

THIS GUY DID A SPECIAL FLIGHT TRACKING RADAR USING CLAUDE THAT SHOWS EVERY AIRCRAFT NEAR WHERE YOU ARE USB-powered. AMOLED touchscreen. Tap any dot on the screen - it tells you the airline, the speed, the registration, whether it's climbing or descending. It sits on his desk. He glances out the window and knows exactly what's flying overhead. He built it in a week, just as a project for himself. Then it blew up. Now people want one. Total cost in parts: about $100. Most people look up and see a dot. He built a window into the whole sky. He built his radar for what's in the air. I built mine for what's on the ticket - using Claude to find the cheapest flight before you book.
0:12

THIS GUY DID A SPECIAL FLIGHT TRACKING RADAR USING CLAUDE THAT SHOWS EVERY AIRCRAFT NEAR WHERE YOU ARE USB-powered. AMOLED touchscreen. Tap any dot on the screen - it tells you the airline, the speed, the registration, whether it's climbing or descending. It sits on his desk. He glances out the window and knows exactly what's flying overhead. He built it in a week, just as a project for himself. Then it blew up. Now people want one. Total cost in parts: about $100. Most people look up and see a dot. He built a window into the whole sky. He built his radar for what's in the air. I built mine for what's on the ticket - using Claude to find the cheapest flight before you book.

kiosa

15,179 просмотров • 2 месяцев назад

"Probably the single best time we had as a company was the last time we were in a position like this, everybody was just super focused, feeling urgency. That's really when we built options and then it's when we built a lot of other features and functions for our customers." Tom Farley, CEO of Bullish, at Consensus Hong Kong. CoinDesk Live in partnership with Hex Trust
0:15

"Probably the single best time we had as a company was the last time we were in a position like this, everybody was just super focused, feeling urgency. That's really when we built options and then it's when we built a lot of other features and functions for our customers." Tom Farley, CEO of Bullish, at Consensus Hong Kong. CoinDesk Live in partnership with Hex Trust

CoinDesk

10,453 просмотров • 7 месяцев назад

I say this, and in no means throwing shade at Cursor, but using the new design mode really makes me appreciate the speed of interactions in Figma and Framer. I will absolutely be testing it out in my own workflow, and love that an IDE finally is adding a browser view and selection UX for local dev support, especially one as powerful as cursor... But i'll wait until undo works. I mean... how am I supposed to design without cmd+z?
0:19

I say this, and in no means throwing shade at Cursor, but using the new design mode really makes me appreciate the speed of interactions in Figma and Framer. I will absolutely be testing it out in my own workflow, and love that an IDE finally is adding a browser view and selection UX for local dev support, especially one as powerful as cursor... But i'll wait until undo works. I mean... how am I supposed to design without cmd+z?

Tom Johnson

38,225 просмотров • 9 месяцев назад

Most people still think the internet is free. It isn’t. Every click, search, scroll, and workflow is training data. Users generate it. Platforms keep it. AI gets smarter. You get nothing. Action Model flips that. The browser extension doesn’t ask you to “farm tasks.” It verifies real actions while you use the product ... then lets you participate in the value those actions create. That’s the model. Train it. Earn it. Own it. And right now there’s a live proof of that model: NeoSoul x ActionFi $5K still on the table Create your first agent. Meet the criteria. Get verified. No vanity clicks. No empty follows. If the extension is already on, you can start now. If it isn’t, you’re watching a closing window from the sidelines. This is the last stretch. Not later. Now. 👉 Action Model #ActionFi #NeoSoul
0:15

Most people still think the internet is free. It isn’t. Every click, search, scroll, and workflow is training data. Users generate it. Platforms keep it. AI gets smarter. You get nothing. Action Model flips that. The browser extension doesn’t ask you to “farm tasks.” It verifies real actions while you use the product ... then lets you participate in the value those actions create. That’s the model. Train it. Earn it. Own it. And right now there’s a live proof of that model: NeoSoul x ActionFi $5K still on the table Create your first agent. Meet the criteria. Get verified. No vanity clicks. No empty follows. If the extension is already on, you can start now. If it isn’t, you’re watching a closing window from the sidelines. This is the last stretch. Not later. Now. 👉 Action Model #ActionFi #NeoSoul

Trathoa

19,704 просмотров • 10 дней назад

🌱 GET INVOLVED 🌱 We need to understand that our planet's future is in our hands. We can't be silent bystanders. We need your help - we need you! 🫵 Let's celebrate Earth Day together and connect with nature! Get involved and receive a FREE MINT or WL! 1. Perform one of the actions below 2. Tag your community and #EarthDay2024 3. RT this post 🎁 FREE Mint (Plant a Tree) - First 20 People Only Plant a tree and reply with a picture of it. Use at least one asset from the comments in the picture. You get a FREE Mint, and the tagged community will get 5 GTD + 15 FCFS spots. 🌳 GTD Spot (Touch the Tree) Reply with a picture of you touching a tree. Use at least one asset from the comments in the picture. You get a GTD spot, and the tagged community will get 5 GTD + 15 FCFS spots. 🪴 GTD Spot (Nature at Home) Reply with a picture of your plants at home. Use at least one asset from the comments in the picture. You get a GTD spot, and the tagged community will get 5 GTD + 15 FCFS spots. 🍀 Random Giveaway RT this post and have a chance to get a WL spot - 5 GTD + 15 FCFS pool ⏳ The campaign will last 10 days. Disclaimer: Tagged communities can claim spots only once.
0:19

🌱 GET INVOLVED 🌱 We need to understand that our planet's future is in our hands. We can't be silent bystanders. We need your help - we need you! 🫵 Let's celebrate Earth Day together and connect with nature! Get involved and receive a FREE MINT or WL! 1. Perform one of the actions below 2. Tag your community and #EarthDay2024 3. RT this post 🎁 FREE Mint (Plant a Tree) - First 20 People Only Plant a tree and reply with a picture of it. Use at least one asset from the comments in the picture. You get a FREE Mint, and the tagged community will get 5 GTD + 15 FCFS spots. 🌳 GTD Spot (Touch the Tree) Reply with a picture of you touching a tree. Use at least one asset from the comments in the picture. You get a GTD spot, and the tagged community will get 5 GTD + 15 FCFS spots. 🪴 GTD Spot (Nature at Home) Reply with a picture of your plants at home. Use at least one asset from the comments in the picture. You get a GTD spot, and the tagged community will get 5 GTD + 15 FCFS spots. 🍀 Random Giveaway RT this post and have a chance to get a WL spot - 5 GTD + 15 FCFS pool ⏳ The campaign will last 10 days. Disclaimer: Tagged communities can claim spots only once.

SANA: Ordinals 🟧

27,534 просмотров • 2 лет назад