Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

I was skeptical of GitHub Copilot auto mode...Then Nhu explained how the routing actually works and I can't go back. Turns out a system that picks the right model every time doesn't just save cost, it outperforms using one model alone. give auto another shot 👇🏽

Andrea

6,610 subscribers

13,229 görüntüleme • 4 ay önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

My AI agent was mass robbing me and I didn't even know. Checked my bill. $1,400 last month. For a side project. This idiot was using the most expensive model for EVERYTHING. A news summary? $3/M token model. A basic thumbnail? Premium tier. I could've used models 10x cheaper. Same output. Couldn't tell the difference. So I built a fix. One toggle called Save Mode. Flip it — system auto-picks the cheapest model that works. 264/mo→264/mo→27/mo. Same workflow. Same results. Go check your AI bill right now. You're welcome. #openclaw #claudcode #vibecoding #aiagent

My AI agent was mass robbing me and I didn't even know. Checked my bill. $1,400 last month. For a side project. This idiot was using the most expensive model for EVERYTHING. A news summary? $3/M token model. A basic thumbnail? Premium tier. I could've used models 10x cheaper. Same output. Couldn't tell the difference. So I built a fix. One toggle called Save Mode. Flip it — system auto-picks the cheapest model that works. 264/mo→264/mo→27/mo. Same workflow. Same results. Go check your AI bill right now. You're welcome. #openclaw #claudcode #vibecoding #aiagent

SkillBoss

30,885 görüntüleme • 6 ay önce

The model is not enough - it doesn't come with an understanding of how your business or team works. That context is spread across the work happening every day: files, conversations, emails, databases, and apps. Figuring out what matters for a particular task, and what the model actually needs, is a hard & unsolved problem. That's why Work IQ + Copilot is special... so I white boarded a bit how it works under the hood.

The model is not enough - it doesn't come with an understanding of how your business or team works. That context is spread across the work happening every day: files, conversations, emails, databases, and apps. Figuring out what matters for a particular task, and what the model actually needs, is a hard & unsolved problem. That's why Work IQ + Copilot is special... so I white boarded a bit how it works under the hood.

Charles Lamanna

42,104 görüntüleme • 4 gün önce

Can we just pause for a moment and marvel at the time we live in? I wanted a SQLite editor in the GitHub Copilot app, but that doesn't exist. So I just asked Opus for it. And got one WITH INTELLISENSE. It's just...I can't even believe it sometimes

Can we just pause for a moment and marvel at the time we live in? I wanted a SQLite editor in the GitHub Copilot app, but that doesn't exist. So I just asked Opus for it. And got one WITH INTELLISENSE. It's just...I can't even believe it sometimes

Burke Holland

74,785 görüntüleme • 2 ay önce

Every dev knows unit tests are important ... and every dev has a project missing them. 😅 With GitHub Copilot CLI, you can quickly generate a robust test suite right from the terminal: 1️⃣ Enter plan mode (Shift-Tab) 2️⃣ Launch a fleet of agents on autopilot 3️⃣ Monitor progress with /tasks Check out how it works, and then try it yourself. 👇

Every dev knows unit tests are important ... and every dev has a project missing them. 😅 With GitHub Copilot CLI, you can quickly generate a robust test suite right from the terminal: 1️⃣ Enter plan mode (Shift-Tab) 2️⃣ Launch a fleet of agents on autopilot 3️⃣ Monitor progress with /tasks Check out how it works, and then try it yourself. 👇

GitHub

54,852 görüntüleme • 6 ay önce

THE FUTURE OF AI APPLICATIONS Jensen Huang: “Another breakthrough happened… the first time I realized Perplexity was using multiple models at the same time, I thought it was completely genius.” The future isn’t one great model… It’s orchestrating multiple great models at every step of the reasoning chain. Multi-model. Multimodal. Multi-cloud.

THE FUTURE OF AI APPLICATIONS Jensen Huang: “Another breakthrough happened… the first time I realized Perplexity was using multiple models at the same time, I thought it was completely genius.” The future isn’t one great model… It’s orchestrating multiple great models at every step of the reasoning chain. Multi-model. Multimodal. Multi-cloud.

Computer

217,875 görüntüleme • 9 ay önce

Threadguy on the moment he realized Zcash privacy actually works "I was DMing one of the guys from the Hyperdash team and I was like, can you find this wallet? He's like yeah, give me a PNL and I can find anyone's wallet. I'm like all right, can you find this? He's like, yo, that's shielded Zcash, I can't. I was like, wait, it really works like that? He's like brother, it's in the shielded pool. And I was like, wait, it actually works. And he's like, yeah, of course it works. And I'm like, raise my fucking targets. I didn't realize it actually worked. I didn't realize you could just shield it like that."

Threadguy on the moment he realized Zcash privacy actually works "I was DMing one of the guys from the Hyperdash team and I was like, can you find this wallet? He's like yeah, give me a PNL and I can find anyone's wallet. I'm like all right, can you find this? He's like, yo, that's shielded Zcash, I can't. I was like, wait, it really works like that? He's like brother, it's in the shielded pool. And I was like, wait, it actually works. And he's like, yeah, of course it works. And I'm like, raise my fucking targets. I didn't realize it actually worked. I didn't realize you could just shield it like that."

Kai

35,932 görüntüleme • 10 gün önce

Tried rigging a model myself for the first time and animating something out of it... I can’t believe it actually works 😆 Thanks to 𝐸𝓊 for making this adorable model for me! Finally… I have my own Machi rig ✨

Tried rigging a model myself for the first time and animating something out of it... I can’t believe it actually works 😆 Thanks to 𝐸𝓊 for making this adorable model for me! Finally… I have my own Machi rig ✨

Mochi 🐸

147,772 görüntüleme • 7 ay önce

GitHub just released a technical preview of the "GitHub Copilot App" - a new agentic development tool that brings agents and @GitHub together in one pane of glass. I've been using this daily for a few weeks. I think you're going to love it.

GitHub just released a technical preview of the "GitHub Copilot App" - a new agentic development tool that brings agents and @GitHub together in one pane of glass. I've been using this daily for a few weeks. I think you're going to love it.

Burke Holland

23,672 görüntüleme • 4 ay önce

Xiaomi just dropped a FREE CLAUDE CODE with memory that actually persists between sessions and nobody is talking about it MiMo Code dropped 4 days ago. 8k GitHub stars. Built-in free model called MiMo Auto - you just install it and it works, zero configuration, zero API keys What's inside: > persistent memory across sessions - agent picks up exactly where it stopped > subagents that spawn automatically and run in parallel > /dream extracts knowledge from your session into project memory > /distill turns your repeated workflows into reusable skills > /goal with a judge model that decides if the work is actually done - not just if the agent thinks it is How to install: > Mac / Linux: curl -fsSL | bash > Windows npm install -g MIMO-ai/cli Then run: > mimo video of me installing and running it is attached github(.)com/XiaomiMiMo/MiMo-Code

Xiaomi just dropped a FREE CLAUDE CODE with memory that actually persists between sessions and nobody is talking about it MiMo Code dropped 4 days ago. 8k GitHub stars. Built-in free model called MiMo Auto - you just install it and it works, zero configuration, zero API keys What's inside: > persistent memory across sessions - agent picks up exactly where it stopped > subagents that spawn automatically and run in parallel > /dream extracts knowledge from your session into project memory > /distill turns your repeated workflows into reusable skills > /goal with a judge model that decides if the work is actually done - not just if the agent thinks it is How to install: > Mac / Linux: curl -fsSL | bash > Windows npm install -g MIMO-ai/cli Then run: > mimo video of me installing and running it is attached github(.)com/XiaomiMiMo/MiMo-Code

monokern

64,582 görüntüleme • 3 ay önce

I lived in the US for a year back in 2003. Matrix Reloaded came out in May, and the hype around it was seriously insane, beyond imagination. In the movie, the main characters are using this super cool phone, and then in a close-up shot, I spotted the 'Samsung' logo right there on it. It felt absolutely mind-blowing and magical. I came back to Korea full of hope that this thing would actually go on sale worldwide. But later I found out it was only released in the US as a super limited edition of just 5,000 units (or so I heard back then). I was hugely disappointed… but eventually, I tracked one down and bought it from a collector. Even now, looking at it, the design is still so damn cool. I took it to a carrier to try using it for real, but they told me it’s a 2G phone and can’t be used anymore. That was another big letdown, I remember it clearly. Do you guys like the Matrix movies too? I’m a total huge fan!

Gamer_lafan

564,651 görüntüleme • 8 ay önce

Here's a time-lapse of the art piece! This was actually a older piece I made back in 2021. So I wanted to give it another go to my current style!

Here's a time-lapse of the art piece! This was actually a older piece I made back in 2021. So I wanted to give it another go to my current style!

Taplaos

89,892 görüntüleme • 13 gün önce

Fable is my favorite model ever. Stuff that never worked suddenly works. The factory is running! Feedback from Slack -> auto-processed -> PRs written –> merged while I sleep. Showing how it runs using Claude and Cursor 👇

Fable is my favorite model ever. Stuff that never worked suddenly works. The factory is running! Feedback from Slack -> auto-processed -> PRs written –> merged while I sleep. Showing how it runs using Claude and Cursor 👇

Kieran Klaassen

22,575 görüntüleme • 3 ay önce

🚨 GMI estimated ~$0.11 in savings on one 3-step Python coding workflow. I gave GMI Router the same e-commerce codebase and split the work into 3 requests: 👉 Code analysis → Cost Mode → DeepSeek-V4-Flash → ~$0.046 saved 👉 Debugging → Balanced Mode → DeepSeek-V4-Flash → ~$0.042 saved 👉 Implementation + testing → Quality Mode → DeepSeek-V4-Flash → ~$0.019 saved Total: ~$0.11 estimated savings vs Claude Opus 4.8, based on GMI's estimated token usage. But here's what caught my attention: Router selected the same model for all 3 tasks — even though I changed the mode each time. And that's actually the point. GMI Cloud 's Router is cache-aware. When the same model can handle the next request, sticking with it can help preserve cache reuse and reduce cost. There's another reason: These tasks weren't especially difficult. So even in Quality Mode, Router didn't need to move to a heavier model. The result: 1 codebase. 3 tasks. 3 routing modes. 1 lightweight model. ~$0.11 estimated savings. This is why request-level routing is more interesting than simply assigning a different model to every task.

🚨 GMI estimated ~$0.11 in savings on one 3-step Python coding workflow. I gave GMI Router the same e-commerce codebase and split the work into 3 requests: 👉 Code analysis → Cost Mode → DeepSeek-V4-Flash → ~$0.046 saved 👉 Debugging → Balanced Mode → DeepSeek-V4-Flash → ~$0.042 saved 👉 Implementation + testing → Quality Mode → DeepSeek-V4-Flash → ~$0.019 saved Total: ~$0.11 estimated savings vs Claude Opus 4.8, based on GMI's estimated token usage. But here's what caught my attention: Router selected the same model for all 3 tasks — even though I changed the mode each time. And that's actually the point. GMI Cloud 's Router is cache-aware. When the same model can handle the next request, sticking with it can help preserve cache reuse and reduce cost. There's another reason: These tasks weren't especially difficult. So even in Quality Mode, Router didn't need to move to a heavier model. The result: 1 codebase. 3 tasks. 3 routing modes. 1 lightweight model. ~$0.11 estimated savings. This is why request-level routing is more interesting than simply assigning a different model to every task.

Anthara Fairooz

28,520 görüntüleme • 1 ay önce

How to use GitHub Actions to build a pipeline that tests your code, trains a model, and publishes a new version of it. Bonus: I let GitHub Workspace Copilot do a lot of the work. (This is the best "AI developer" I've tested so far by a mile.) Here is the full video:

How to use GitHub Actions to build a pipeline that tests your code, trains a model, and publishes a new version of it. Bonus: I let GitHub Workspace Copilot do a lot of the work. (This is the best "AI developer" I've tested so far by a mile.) Here is the full video:

Santiago

148,109 görüntüleme • 2 yıl önce

Production traffic is not uniform. You get a few requests that need your best model, but most are simple questions and lookups you can solve with cheaper, faster models. The most expensive mistake you can make today is sending every request to your strongest model. You need routing. Period. This is the simplest trick to improve the architecture of whatever you are building. Please, don't implement routing yourself. You don't have to. I'm currently working with TrueFoundry's Auto Routing. It reads each request, classifies it as simple, medium, or complex, and sends it to the model assigned to that tier. You have two choices: 1. Send every request to the free heuristic classifier to score signals such as technical vocabulary, code, prompt length, and multi-step reasoning. 2. Send the request to an LLM classifier when its difficulty requires a more nuanced judgment. The beauty of using routing is that nothing changes in your code. You still call a single endpoint model, but routing works behind the scenes to pair every request with the best possible model. TrueFoundry ran several experiments with two different setups: 1. Send every request to Claude Opus 2. Send every request to a router with Haiku, Sonnet, and Opus The first experiment ran 550 deterministically graded academic prompts through every setup. Auto Routing was 69% cheaper while retaining 98% of the baseline quality. The second experiment ran three production-shaped workloads through every setup, using user chats, developer chats, and long agent tasks. Auto Routing was 80% cheaper. Thanks to the TrueFoundry team for partnering with me on this post.

Production traffic is not uniform. You get a few requests that need your best model, but most are simple questions and lookups you can solve with cheaper, faster models. The most expensive mistake you can make today is sending every request to your strongest model. You need routing. Period. This is the simplest trick to improve the architecture of whatever you are building. Please, don't implement routing yourself. You don't have to. I'm currently working with TrueFoundry's Auto Routing. It reads each request, classifies it as simple, medium, or complex, and sends it to the model assigned to that tier. You have two choices: 1. Send every request to the free heuristic classifier to score signals such as technical vocabulary, code, prompt length, and multi-step reasoning. 2. Send the request to an LLM classifier when its difficulty requires a more nuanced judgment. The beauty of using routing is that nothing changes in your code. You still call a single endpoint model, but routing works behind the scenes to pair every request with the best possible model. TrueFoundry ran several experiments with two different setups: 1. Send every request to Claude Opus 2. Send every request to a router with Haiku, Sonnet, and Opus The first experiment ran 550 deterministically graded academic prompts through every setup. Auto Routing was 69% cheaper while retaining 98% of the baseline quality. The second experiment ran three production-shaped workloads through every setup, using user chats, developer chats, and long agent tasks. Auto Routing was 80% cheaper. Thanks to the TrueFoundry team for partnering with me on this post.

Santiago

15,543 görüntüleme • 22 gün önce

I found out that there are actually no limits to the amount of Voidgrubs you can kill, so you can just keep spawning them in Practice Tool. It deals around 2500 true damage with one auto if you have 60 Voidgrub stacks. Also don't mind the model I was testing something else LMAO

I found out that there are actually no limits to the amount of Voidgrubs you can kill, so you can just keep spawning them in Practice Tool. It deals around 2500 true damage with one auto if you have 60 Voidgrub stacks. Also don't mind the model I was testing something else LMAO

aryasdemise 🇹🇷

348,207 görüntüleme • 1 yıl önce

I finally sat down and watched the first two episodes of "Dutton Ranch".. and l've gotta say, I'm hooked already. That Taylor Sheridan sure knows how to pull you in. What I can't figure out is how in the world that man finds the time to crank out all these shows. Seems like every time I turn around, he's got another hit lined up and ready to go. Not complaining one bit though... just hoping he doesn't run out of stories anytime soon. 👍

I finally sat down and watched the first two episodes of "Dutton Ranch".. and l've gotta say, I'm hooked already. That Taylor Sheridan sure knows how to pull you in. What I can't figure out is how in the world that man finds the time to crank out all these shows. Seems like every time I turn around, he's got another hit lined up and ready to go. Not complaining one bit though... just hoping he doesn't run out of stories anytime soon. 👍

G-PA

21,961 görüntüleme • 4 ay önce

Kimi K3 dropped 3 days ago and i already used it to build my own PUBG the model that just took the number one coding spot on earth and it runs on a free tier no team no engine i just fed it free open models and assets that are already public and told it what i wanted it built the entire game itself the shooting the recoil the hit detection all of it came straight out of the model i barely wrote a line what youre watching here is the deathmatch mode and its already live battle royale is next im deep in it and the graphics get reworked right after i know it looks raw right now i wont pretend otherwise but the endgame is a build that runs perfect on phones and the moment its on mobile you drop ads in and it earns on its own the part that still gets me is my whole cost for this was a single subscription

Kimi K3 dropped 3 days ago and i already used it to build my own PUBG the model that just took the number one coding spot on earth and it runs on a free tier no team no engine i just fed it free open models and assets that are already public and told it what i wanted it built the entire game itself the shooting the recoil the hit detection all of it came straight out of the model i barely wrote a line what youre watching here is the deathmatch mode and its already live battle royale is next im deep in it and the graphics get reworked right after i know it looks raw right now i wont pretend otherwise but the endgame is a build that runs perfect on phones and the moment its on mobile you drop ads in and it earns on its own the part that still gets me is my whole cost for this was a single subscription

savip.

68,984 görüntüleme • 2 ay önce