Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

Microsoft has released a major update to Copilot You can now use o3-mini without limit by clicking on “Think Deeper” People sleep on it but Copilot has (for free): → Unlimited reasoning model → Unlimited voice mode → Real-time data w/ GPT-4o → Image gen (Dall-E 3 though)

Paul Couvert

220,799 subscribers

55,877 Aufrufe • vor 1 Jahr •via X (Twitter)

Bildung Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

12 Kommentare

Profilbild von Paul Couvert

Paul Couvertvor 1 Jahr

- Just access Copilot (web or mobile app) - Select "Think Deeper" in the text field - Copilot will use o3-mini to answer you Once again, no limits. →

Profilbild von ARK Electronics

ARK Electronicsvor 2 Jahren

Elevate your drone game with our USA-made, NDAA-compliant flight controllers! Trust in reliable technology that not only enhances your flights but also supports US drone manufacturing capability. 🇺🇸✈️ #USAMade #NDAA #drones #uav #uas #px4 #ardupilot #USA #unmanned #opensource

Profilbild von Sai Rahul

Sai Rahulvor 1 Jahr

Ah yes. I completely forgot the copilot 😅

Profilbild von Paul Couvert

Paul Couvertvor 1 Jahr

Good thing I'm here to remind you of its existence then 😂

Profilbild von Josh Marino

Josh Marinovor 1 Jahr

So people are paying $200 a month for unlimited voice mode but they could get it with Co-Pilot for free?

Profilbild von Paul Couvert

Paul Couvertvor 1 Jahr

I believe this is also the case on ChatGPT Plus but I don't know if it's unlimited or if they just increased the limit.

Profilbild von Shushant Lakhyani

Shushant Lakhyanivor 1 Jahr

There's no need of ChatGPT's subscription now

Profilbild von Paul Couvert

Paul Couvertvor 1 Jahr

Depending on the task, but it can replace it in many situations!

Profilbild von MadMonke.sol

MadMonke.solvor 1 Jahr

accessing copilot opens a world of creativity, doesn’t it? excited to see the insights we'll uncover together.

Profilbild von Paul Couvert

Paul Couvertvor 1 Jahr

Worth a try!

Profilbild von Prometheus

Prometheusvor 1 Jahr

Is this a desktop app? Where to download it?

Profilbild von Paul Couvert

Paul Couvertvor 1 Jahr

More like a progressive web app but a native one is available in preview and should be available soon.

Ähnliche Videos

Copilot has entered the group chat - and it's bringing GPT 4o image gen with it. Go from debate to decisions, from "what if" to "look at this" in seconds just by tagging Microsoft Copilot in GroupMe. For memes, planning, searching, roasting, all without leaving the chat. Try it out!

Copilot has entered the group chat - and it's bringing GPT 4o image gen with it. Go from debate to decisions, from "what if" to "look at this" in seconds just by tagging Microsoft Copilot in GroupMe. For memes, planning, searching, roasting, all without leaving the chat. Try it out!

Mustafa Suleyman

38,220 Aufrufe • vor 1 Jahr

Microsoft has just released an impressive update Copilot on Windows can now see and discuss anything on your screen in real time. Basically ChatGPT Advanced voice mode, but with vision and everywhere on your laptop. You don't need a subscription, it's 100% FREE: 1. Open Copilot 2. Click on the new vision icon 3. Choose the software to share 4. Start the conversation! Genuine game-changer to receive real-time guidance, understand anything on screen, learn faster and more. And the transcript is available right after your session ends.

Microsoft has just released an impressive update Copilot on Windows can now see and discuss anything on your screen in real time. Basically ChatGPT Advanced voice mode, but with vision and everywhere on your laptop. You don't need a subscription, it's 100% FREE: 1. Open Copilot 2. Click on the new vision icon 3. Choose the software to share 4. Start the conversation! Genuine game-changer to receive real-time guidance, understand anything on screen, learn faster and more. And the transcript is available right after your session ends.

Paul Couvert

178,855 Aufrufe • vor 1 Jahr

Microsoft has released a new open source AI model Phi-4-mini is surprisingly powerful for its size (3.8B!) You can use it for free locally on your Windows, MacOS or Linux laptop. Links and tutorial below

Microsoft has released a new open source AI model Phi-4-mini is surprisingly powerful for its size (3.8B!) You can use it for free locally on your Windows, MacOS or Linux laptop. Links and tutorial below

Paul Couvert

209,124 Aufrufe • vor 1 Jahr

Copilot can now control your browser 🔥 Microsoft has just updated Edge to make it fully agentic - Enable Copilot Mode - Turn on Actions in Edge Preview - Give the AI any goal to achieve You can either let it work in the background or watch it perform actions. And there are 3 modes available to make it more or less autonomous (light, balanced, and strict).

Copilot can now control your browser 🔥 Microsoft has just updated Edge to make it fully agentic - Enable Copilot Mode - Turn on Actions in Edge Preview - Give the AI any goal to achieve You can either let it work in the background or watch it perform actions. And there are 3 modes available to make it more or less autonomous (light, balanced, and strict).

Paul Couvert

52,823 Aufrufe • vor 8 Monaten

Microsoft still can’t decide what Copilot should be on Windows 11. Microsoft is testing a new Copilot docking experience, and it feels like Microsoft is quietly going back to the original sidebar idea. Copilot still opens as an app by default, but there’s now a new title bar menu with custom snapping options. You can keep it as a normal app, use picture-in-picture, or dock Copilot to the left or right side of the desktop. When docked, Windows 11 resizes your workspace around Copilot. Your apps move aside, File Explorer adjusts, and Copilot stays pinned to the edge instead of floating like a separate window.

Microsoft still can’t decide what Copilot should be on Windows 11. Microsoft is testing a new Copilot docking experience, and it feels like Microsoft is quietly going back to the original sidebar idea. Copilot still opens as an app by default, but there’s now a new title bar menu with custom snapping options. You can keep it as a normal app, use picture-in-picture, or dock Copilot to the left or right side of the desktop. When docked, Windows 11 resizes your workspace around Copilot. Your apps move aside, File Explorer adjusts, and Copilot stays pinned to the edge instead of floating like a separate window.

Windows Latest

34,824 Aufrufe • vor 1 Monat

Copilot AI in Microsoft 365 is underrated. I created a PowerPoint in 40 seconds. For the example I just described what I wanted to create, and Copilot generated for me: - Slides - Illustrations - Text - Presenter notes All thanks to the power of GPT-4. But you can also modify what's already been created. You can add text, modify existing text or create a new illustration with Dall-E 3. It's like having ChatGPT Plus directly integrated into PowerPoint. People don't realize the impact Copilot will have and what it can do. Note: I've speeded up some of the loading to make the video more pleasant to watch. But it's still very fast.

Copilot AI in Microsoft 365 is underrated. I created a PowerPoint in 40 seconds. For the example I just described what I wanted to create, and Copilot generated for me: - Slides - Illustrations - Text - Presenter notes All thanks to the power of GPT-4. But you can also modify what's already been created. You can add text, modify existing text or create a new illustration with Dall-E 3. It's like having ChatGPT Plus directly integrated into PowerPoint. People don't realize the impact Copilot will have and what it can do. Note: I've speeded up some of the loading to make the video more pleasant to watch. But it's still very fast.

Paul Couvert

100,139 Aufrufe • vor 2 Jahren

I created a free mini-course covering how to use AI for GTM. From deep market research to AI script writing, this course covers our top use cases of AI we’ve used to generate 10,000+ leads for our clients. Here's exactly what this free course will cover: > Performing deep market research with GPT-o3 & Perplexity > TAM & ICP mapping with GPT-o3 > Discovering niche data sources with Perplexity > ICP Keyword generation with Claude 3.7 Sonnet > Campaign strategy development with GPT-o3 > Outbound messaging creation with Claude 3.7 Sonnet > Email personalization at scale with GPT-4o-mini Want to get your hands on it? Comment "AI GTM" and I'll DM you the link directly (Must be connected to receive)

I created a free mini-course covering how to use AI for GTM. From deep market research to AI script writing, this course covers our top use cases of AI we’ve used to generate 10,000+ leads for our clients. Here's exactly what this free course will cover: > Performing deep market research with GPT-o3 & Perplexity > TAM & ICP mapping with GPT-o3 > Discovering niche data sources with Perplexity > ICP Keyword generation with Claude 3.7 Sonnet > Campaign strategy development with GPT-o3 > Outbound messaging creation with Claude 3.7 Sonnet > Email personalization at scale with GPT-4o-mini Want to get your hands on it? Comment "AI GTM" and I'll DM you the link directly (Must be connected to receive)

Christian

50,809 Aufrufe • vor 1 Jahr

Sam Altman recently said: “GPT-OSS has strong real-world performance comparable to o4-mini—and you can run it locally on your phone.” Many believed running a 20B-parameter model on mobile devices was still years away. At Nexa AI, we’ve built our foundation on deep on-device AI technology—turning that vision into reality. Today, GPT-OSS is running fully local on mobile devices through our app, Nexa Studio. Real performance on Snapdragon Gen 5: - 17 tokens/sec decoding speed - < 3 seconds Time-to-First-Token Developers can now use NexaSDK to build their own local AI apps powered by GPT-OSS. What this unlocks: - Real reasoning models running locally - The next wave of AI Agent use cases—without cloud limits - True data privacy: your data never leaves your device Appreciation to OpenAI Sam Altman Greg Brockman and the entire leadership team for pushing the boundaries of the open-source AI community. Thanks to Qualcomm for the partnership, and to Manoj Khilnani and Chun-Po (Jerry) Chang for the incredible collaboration.

Sam Altman recently said: “GPT-OSS has strong real-world performance comparable to o4-mini—and you can run it locally on your phone.” Many believed running a 20B-parameter model on mobile devices was still years away. At Nexa AI, we’ve built our foundation on deep on-device AI technology—turning that vision into reality. Today, GPT-OSS is running fully local on mobile devices through our app, Nexa Studio. Real performance on Snapdragon Gen 5: - 17 tokens/sec decoding speed - < 3 seconds Time-to-First-Token Developers can now use NexaSDK to build their own local AI apps powered by GPT-OSS. What this unlocks: - Real reasoning models running locally - The next wave of AI Agent use cases—without cloud limits - True data privacy: your data never leaves your device Appreciation to OpenAI Sam Altman Greg Brockman and the entire leadership team for pushing the boundaries of the open-source AI community. Thanks to Qualcomm for the partnership, and to Manoj Khilnani and Chun-Po (Jerry) Chang for the incredible collaboration.

NEXA AI

197,453 Aufrufe • vor 9 Monaten

🗣️ ZODs On-Chain AI - Voice Mode (Iteration 3) You can command AI agents with voice commands. Use-case: You're driving, but you're a Solana maxi who can't go a few hours without on-chain activity. 🌍 100+ languages supported ⚛ Model-agnostic, so you can use all leading AI models, from all providers 🤖 All AI agents supported through Voice Mode Coming in iteration 3. Now back to work.

🗣️ ZODs On-Chain AI - Voice Mode (Iteration 3) You can command AI agents with voice commands. Use-case: You're driving, but you're a Solana maxi who can't go a few hours without on-chain activity. 🌍 100+ languages supported ⚛ Model-agnostic, so you can use all leading AI models, from all providers 🤖 All AI agents supported through Voice Mode Coming in iteration 3. Now back to work.

ZODs

36,546 Aufrufe • vor 1 Jahr

Windows Copilot has just been massively improved: It can now understand any file on your PC. You can drag and drop a PDF, for example, and ask it to explain or summarize it. Then it can take it into account for the following answers. Closer and closer to a real AI computer.

Windows Copilot has just been massively improved: It can now understand any file on your PC. You can drag and drop a PDF, for example, and ask it to explain or summarize it. Then it can take it into account for the following answers. Closer and closer to a real AI computer.

Paul Couvert

83,178 Aufrufe • vor 2 Jahren

I'm happy to announce some major upgrades for the RAD Debugger, coming in the next release: - A new splash screen has been added, to visualize initial startup loading, and to allow you to mentally prepare for the program to load. - Formal integration with Microsoft's Copilot™ has been added. Stuck on a bug? Need some advice? Need a friend to talk to about something? Just click the Copilot button in the title bar! - Formal integration with your Microsoft account. Save all your sessions and your Copilot conversations, so that your Copilot can help you anywhere, at any time! (NOTE: Your breakpoints will not be saved across sessions.) I'm so excited to share this with everyone!

I'm happy to announce some major upgrades for the RAD Debugger, coming in the next release: - A new splash screen has been added, to visualize initial startup loading, and to allow you to mentally prepare for the program to load. - Formal integration with Microsoft's Copilot™ has been added. Stuck on a bug? Need some advice? Need a friend to talk to about something? Just click the Copilot button in the title bar! - Formal integration with your Microsoft account. Save all your sessions and your Copilot conversations, so that your Copilot can help you anywhere, at any time! (NOTE: Your breakpoints will not be saved across sessions.) I'm so excited to share this with everyone!

Ryan Fleury

26,442 Aufrufe • vor 1 Jahr

🚨HUGE UPDATE ALERT🚨 Unlimited Mode is now LIVE! Our new filter ability allows you to turn off categories that you may not like. If the normal PokeDoku format is too easy for you, you can allow Unlimited Mode to generate one answer squares and puzzles that can soft lock! We have also finally added the Timer feature so you can see how fast you complete a puzzle! We worked super hard on this game mode because we wanted it to be everything you might have envisioned and more. All we ask in return is a Like and Retweet on this post. Let’s get PokeDoku trending! Tell the world… PokeDoku has Unlimited Mode.

🚨HUGE UPDATE ALERT🚨 Unlimited Mode is now LIVE! Our new filter ability allows you to turn off categories that you may not like. If the normal PokeDoku format is too easy for you, you can allow Unlimited Mode to generate one answer squares and puzzles that can soft lock! We have also finally added the Timer feature so you can see how fast you complete a puzzle! We worked super hard on this game mode because we wanted it to be everything you might have envisioned and more. All we ask in return is a Like and Retweet on this post. Let’s get PokeDoku trending! Tell the world… PokeDoku has Unlimited Mode.

PokeDoku

188,514 Aufrufe • vor 2 Jahren

You think it, invideo Vision shows it. Here's a step-by-step breakdown on how you can use it to convert your story into a frame-by-frame reality👇 FREE + Unlimited for the next 7 days. No subscription required. RT + Comment 'VISION' for a chance to win the highest generative plan on invideo + $1200 in credits.

You think it, invideo Vision shows it. Here's a step-by-step breakdown on how you can use it to convert your story into a frame-by-frame reality👇 FREE + Unlimited for the next 7 days. No subscription required. RT + Comment 'VISION' for a chance to win the highest generative plan on invideo + $1200 in credits.

Invideo

668,579 Aufrufe • vor 6 Monaten

Cerebras inference is very fast. So fast that it changes how we think about configuring our LLMs for voice agent use cases. Kimi K2.6 is a 1T parameter reasoning model that Cerebras serves at 650 - 1,000 tokens per second (end-to-end throughput), with time to first token metrics as low as 150ms (latency). These numbers are two to three times faster than other similarly capable models. The biggest lever we get from this kind of speed is that we can use the model in reasoning mode, and still have excellent "time to first non-thinking token." This solves a big pain point we have in 2026 for voice agent use cases. Almost all recent innovation in post-training has focused on making models good at reasoning ("test time compute"). This is great, but it makes the user-facing model latency much, much slower. Which is a problem for conversational voice agents. We can run Kimi K2.6 with reasoning turned on, and get responses faster than other models produce with reasoning disabled. On my 30-turn voice agent benchmark, Kimi K2.6 with reasoning enabled ties GPT 5.1 and Haiku 4.5 with reasoning disabled, and is still about 200ms seconds faster! On my primary task agent benchmark, Kimi K2.6 is now the #2 model. It ranks just behind Gemini 3.5 Flash in "high" reasoning mode, and tied with GLM 5, Sonnet 4.6, and GPT 5.4 with reasoning set to "low." But Kimi K2.6 completes each turn in the agent loop in under 500ms. The other four models are all at least 3x slower. (Models only qualify for this benchmark if they can complete task turns at a P50 <4s.) A couple of other things that this speed buys us, for production voice agents: - Tool calls happen fast enough that we don't have to work around tool call latency in our pipeline design. - We can prompt the model to output structured data at the beginning of a response, followed by plain text for voice generation. This opens up possibilities like asking the model to do complex classification/generation tasks that influence the rest of the pipeline. For example, the model could create a detailed style prompt for a steerable TTS model, for each individual conversation turn. And, of course, you can use Kimi K2.6 with reasoning turned off. Cerebras calls this "instant" mode. Here's a video of a Cerebras Kimi K2.6 voice agent with voice-to-voice response time, measured at the client, under 500ms. This is the true response latency as perceived by the user, including all network and audio codec overhead, transcription and turn detection, Kimi K2.6 token generation, and voice generation. 500ms is, effectively, instant. So the Cerebras naming for this mode is a propos. :-)

Cerebras inference is very fast. So fast that it changes how we think about configuring our LLMs for voice agent use cases. Kimi K2.6 is a 1T parameter reasoning model that Cerebras serves at 650 - 1,000 tokens per second (end-to-end throughput), with time to first token metrics as low as 150ms (latency). These numbers are two to three times faster than other similarly capable models. The biggest lever we get from this kind of speed is that we can use the model in reasoning mode, and still have excellent "time to first non-thinking token." This solves a big pain point we have in 2026 for voice agent use cases. Almost all recent innovation in post-training has focused on making models good at reasoning ("test time compute"). This is great, but it makes the user-facing model latency much, much slower. Which is a problem for conversational voice agents. We can run Kimi K2.6 with reasoning turned on, and get responses faster than other models produce with reasoning disabled. On my 30-turn voice agent benchmark, Kimi K2.6 with reasoning enabled ties GPT 5.1 and Haiku 4.5 with reasoning disabled, and is still about 200ms seconds faster! On my primary task agent benchmark, Kimi K2.6 is now the #2 model. It ranks just behind Gemini 3.5 Flash in "high" reasoning mode, and tied with GLM 5, Sonnet 4.6, and GPT 5.4 with reasoning set to "low." But Kimi K2.6 completes each turn in the agent loop in under 500ms. The other four models are all at least 3x slower. (Models only qualify for this benchmark if they can complete task turns at a P50 <4s.) A couple of other things that this speed buys us, for production voice agents: - Tool calls happen fast enough that we don't have to work around tool call latency in our pipeline design. - We can prompt the model to output structured data at the beginning of a response, followed by plain text for voice generation. This opens up possibilities like asking the model to do complex classification/generation tasks that influence the rest of the pipeline. For example, the model could create a detailed style prompt for a steerable TTS model, for each individual conversation turn. And, of course, you can use Kimi K2.6 with reasoning turned off. Cerebras calls this "instant" mode. Here's a video of a Cerebras Kimi K2.6 voice agent with voice-to-voice response time, measured at the client, under 500ms. This is the true response latency as perceived by the user, including all network and audio codec overhead, transcription and turn detection, Kimi K2.6 token generation, and voice generation. 500ms is, effectively, instant. So the Cerebras naming for this mode is a propos. :-)

kwindla

40,319 Aufrufe • vor 1 Monat

Copilot Tasks is seriously good?! Even one of the best alternative to Claude Cowork Using a single prompt it was able to: → Use a cloud browser to find the right tool → Interact with the page to enter data → Interpret all the info given by the page → Generate a PowerPoint based on it → Draft an email containing the learnings Because it has access in real time to multiple Microsoft Office apps like Word, PowerPoint, Excel, etc. but also Outlook. Everything was seamless. First try. And you can even ask it to schedule this task (or another one) to run it every Friday, once a month, you name it. As with OpenClaw, the use cases are basically infinite.

Copilot Tasks is seriously good?! Even one of the best alternative to Claude Cowork Using a single prompt it was able to: → Use a cloud browser to find the right tool → Interact with the page to enter data → Interpret all the info given by the page → Generate a PowerPoint based on it → Draft an email containing the learnings Because it has access in real time to multiple Microsoft Office apps like Word, PowerPoint, Excel, etc. but also Outlook. Everything was seamless. First try. And you can even ask it to schedule this task (or another one) to run it every Friday, once a month, you name it. As with OpenClaw, the use cases are basically infinite.

Paul Couvert

153,440 Aufrufe • vor 3 Monaten

A voice agent powered by gpt-oss. Running locally on my macBook. Demo recorded in a Waymo with WiFi turned off. I'm still on my space game voice AI kick, obviously. Code link below. For conversational voice AI, you want to set the gpt-oss reasoning behavior to "low". (The default is "medium".) Notes on how to do that and a jinja template you can use are in the repo. The LLM in the demo video is the big, 120B version of gpt-oss. You can use the smaller, 20B model for this, of course. But OpenAI really did a cool thing here designing the 120B model to run in "just" 80GB of VRAM. And the llama.cpp mlx inference is fast: ~250ms TTFT. Running a big model on-device feels like a time warp into the future of AI.

A voice agent powered by gpt-oss. Running locally on my macBook. Demo recorded in a Waymo with WiFi turned off. I'm still on my space game voice AI kick, obviously. Code link below. For conversational voice AI, you want to set the gpt-oss reasoning behavior to "low". (The default is "medium".) Notes on how to do that and a jinja template you can use are in the repo. The LLM in the demo video is the big, 120B version of gpt-oss. You can use the smaller, 20B model for this, of course. But OpenAI really did a cool thing here designing the 120B model to run in "just" 80GB of VRAM. And the llama.cpp mlx inference is fast: ~250ms TTFT. Running a big model on-device feels like a time warp into the future of AI.

kwindla

202,128 Aufrufe • vor 11 Monaten

We made a thing! Very happy to announce sqlcoder-pro and the Defog Alignment Platform. Available to use immediately without a wait-list, weights will be open-sourced very soon. The video does a quick show and tell comparison against ChatGPT (with gpt-4o). Read on for more details! TLDR 💪 equal (or better) performance on text-to-SQL as the most capable Claude-3.5 or GPT-4 models 🤝 You can use it today on a free plan/free trial, without a waitlist 🪽 self-hostable on a single RTX4090, with 2 second median generation times for SQL queries 🔁 exactly the same output every time, give the same prompt 👨🏻‍🏫 teachable and steerable: show the model what you want it to do 🛞 debuggable – you can understand WTF is going on inside the model, instead of treating it like a black box Let's dig into each of these one-by-one! Performance SQLCoder-8b-pro significantly exceeds the performance of our previous sqlcoder-8b model on Postgres text-to-SQL (from 88.2% to 90.2% accuracy - gpt-4o is at 87.6%, for reference). It is also better at following instructions. This was done via self-merges, hand crafted fine-tuning data, and adapting the training data to fit our tokenizer. Cost You can host this on the model on a single $3,500 RTX4090, and support ~5 requests/second via VLLM. If you're looking to host on the cloud instead, you can run it on a single L4 GPU that costs $300/mo on GCP Repeatability We have a dense 8b model with no MoE shenanigans. For the same prompt with temperature=0, you'll always get the same answer – which is critical in BI. Teachable In our alignment and feedback modes, you can give the model feedback on how it answered certain questions, and it will automatically adapt to the feedback. Debuggable You can use logprobs and attention scores to determine where, exactly is the model paying attention to inside a prompt + what it's getting confused by when generating outputs. Available today You can use Defog on the cloud today by going to docs[dot]defog[dot]ai, and getting an API key. Excited to hear what you think!

We made a thing! Very happy to announce sqlcoder-pro and the Defog Alignment Platform. Available to use immediately without a wait-list, weights will be open-sourced very soon. The video does a quick show and tell comparison against ChatGPT (with gpt-4o). Read on for more details! TLDR 💪 equal (or better) performance on text-to-SQL as the most capable Claude-3.5 or GPT-4 models 🤝 You can use it today on a free plan/free trial, without a waitlist 🪽 self-hostable on a single RTX4090, with 2 second median generation times for SQL queries 🔁 exactly the same output every time, give the same prompt 👨🏻‍🏫 teachable and steerable: show the model what you want it to do 🛞 debuggable – you can understand WTF is going on inside the model, instead of treating it like a black box Let's dig into each of these one-by-one! Performance SQLCoder-8b-pro significantly exceeds the performance of our previous sqlcoder-8b model on Postgres text-to-SQL (from 88.2% to 90.2% accuracy - gpt-4o is at 87.6%, for reference). It is also better at following instructions. This was done via self-merges, hand crafted fine-tuning data, and adapting the training data to fit our tokenizer. Cost You can host this on the model on a single $3,500 RTX4090, and support ~5 requests/second via VLLM. If you're looking to host on the cloud instead, you can run it on a single L4 GPU that costs $300/mo on GCP Repeatability We have a dense 8b model with no MoE shenanigans. For the same prompt with temperature=0, you'll always get the same answer – which is critical in BI. Teachable In our alignment and feedback modes, you can give the model feedback on how it answered certain questions, and it will automatically adapt to the feedback. Debuggable You can use logprobs and attention scores to determine where, exactly is the model paying attention to inside a prompt + what it's getting confused by when generating outputs. Available today You can use Defog on the cloud today by going to docs[dot]defog[dot]ai, and getting an API key. Excited to hear what you think!

Rishabh Srivastava

13,433 Aufrufe • vor 1 Jahr

OptimAI Search MCP brings live context to the agents people already use. Claude. Cursor. Copilot. Codex. One integration layer. Everywhere you build. Under the hood, OptimAI can process hundreds to thousands of real-time signals in parallel across the public web, social platforms, and onchain data - all powered by a decentralized network. That is the difference: not another search tool, but infrastructure for agent-native intelligence. Build with it. 💡 -- OptimAI: the #1 AI on BNB Chain DappBay.

OptimAI Search MCP brings live context to the agents people already use. Claude. Cursor. Copilot. Codex. One integration layer. Everywhere you build. Under the hood, OptimAI can process hundreds to thousands of real-time signals in parallel across the public web, social platforms, and onchain data - all powered by a decentralized network. That is the difference: not another search tool, but infrastructure for agent-native intelligence. Build with it. 💡 -- OptimAI: the #1 AI on BNB Chain DappBay.

OptimAI Network

45,831 Aufrufe • vor 1 Monat

OpenAI just announced API access to o1 (advanced reasoning model) yesterday. I'm delighted to announce today a new short course, Reasoning with o1, built with OpenAI, and taught by Colin Jarvis, Head of AI Solutions at OpenAI, to show you how to use this effectively! Unlike previous language models which generate output directly, o1 “thinks before it responds,” and generates many reasoning tokens before returning a more thoughtful and accurate response. It is great at complex reasoning -- including planning for agentic workflows, coding, and domain-specific reasoning in STEM fields like law. But how you should use it is quite different from other LLMs. I think o1 will be a game changer for many AI applications; and in this course, you'll learn how to use it effectively. In detail, you’ll: - Learn to recognize what tasks o1 is suited for, and when to use a smaller model, or combine o1 with a smaller model - Understand the new principles of prompting reasoning models: Be simple and direct; no explicit chain-of-thought required; use structure; show rather than tell - Implement multi-step orchestration in which o1 plans, and hands tasks over to gpt-4o-mini to execute specific steps; this illustrates a design pattern to optimize intelligence (accuracy) and cost - Use o1 for a coding task to build a new application, edit existing code, and test performance by running a coding competition between o1-mini and GPT 4o - Use o1 for image understanding and learn how it performs better with a "hierarchy of reasoning," in which it incurs the latency and cost upfront, preprocessing the image and indexing it with rich details so it can be used for Q&A later - Learn a technique called meta-prompting, in which you use o1 to improve your prompts. Using a customer support evaluation set, you'll iteratively use o1 to modify a prompt to improve performance You'll also learn about how OpenAI used reinforcement learning to produce a model that uses "test-time compute" to improve performance. I think you'll find this course enjoyable and valuable. Please sign up for it here:

OpenAI just announced API access to o1 (advanced reasoning model) yesterday. I'm delighted to announce today a new short course, Reasoning with o1, built with OpenAI, and taught by Colin Jarvis, Head of AI Solutions at OpenAI, to show you how to use this effectively! Unlike previous language models which generate output directly, o1 “thinks before it responds,” and generates many reasoning tokens before returning a more thoughtful and accurate response. It is great at complex reasoning -- including planning for agentic workflows, coding, and domain-specific reasoning in STEM fields like law. But how you should use it is quite different from other LLMs. I think o1 will be a game changer for many AI applications; and in this course, you'll learn how to use it effectively. In detail, you’ll: - Learn to recognize what tasks o1 is suited for, and when to use a smaller model, or combine o1 with a smaller model - Understand the new principles of prompting reasoning models: Be simple and direct; no explicit chain-of-thought required; use structure; show rather than tell - Implement multi-step orchestration in which o1 plans, and hands tasks over to gpt-4o-mini to execute specific steps; this illustrates a design pattern to optimize intelligence (accuracy) and cost - Use o1 for a coding task to build a new application, edit existing code, and test performance by running a coding competition between o1-mini and GPT 4o - Use o1 for image understanding and learn how it performs better with a "hierarchy of reasoning," in which it incurs the latency and cost upfront, preprocessing the image and indexing it with rich details so it can be used for Q&A later - Learn a technique called meta-prompting, in which you use o1 to improve your prompts. Using a customer support evaluation set, you'll iteratively use o1 to modify a prompt to improve performance You'll also learn about how OpenAI used reinforcement learning to produce a model that uses "test-time compute" to improve performance. I think you'll find this course enjoyable and valuable. Please sign up for it here:

Andrew Ng

357,661 Aufrufe • vor 1 Jahr