正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

How does O1 break down problems into steps? While prior models memorized content, O1 adds a layer where it thinks about how to arrive at the right sequence of steps to do more complex tasks and get the right answer.

Garry Tan

564,857 subscribers

44,219 次观看 • 1 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

10 条评论

Garry Tan 的头像

Garry Tan1 年前

Why OpenAI's o1 Is A Huge Deal | YC Decoded

Romain Lacombe 的头像

Romain Lacombe1 年前

Just a note Garry: it's not a layer, CoT is all happening in-context.

SLOMP 的头像

SLOMP1 年前

keeping yc founders on the bleeding edge

protomemetic 的头像

protomemetic1 年前

Appreciate this video. Just last night I asked o1 what it meant by "thinking" and it told me to stop asking.

rudy@FURE.Cab 🌐 - e/acc 的头像

[email protected] 🌐 - e/acc1 年前

Sim to First Principal thinking 🤔

djcows 的头像

djcows1 年前

i see pizza math problem solving, i press like

Steven Musielski 的头像

Steven Musielski1 年前

This is incredible Garry. WOW.

Dennis Hackethal 的头像

Dennis Hackethal1 年前

@ycombinator No model ever ‘memorized’ or ‘thought’ about anything. Stop using misleadingly humanizing verbiage and spreading this false advertising

DWA 🌊 🚀 的头像

DWA 🌊 🚀1 年前

Keep hyping guys…

Jon_D 的头像

Jon_D1 年前

I wonder what the next thinking process upgrade will be! Definitely a fascinating breakthrough. What do you think is next?

相关视频

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

357,401 次观看 • 1 年前

o1 is here! Alistair sat down with shyamal at the OpenAI office to discuss how our unique techniques have unlocked new levels of performance. We teach LLMs to mimic human reasoning and benefit dramatically from new models, including o1.

o1 is here! Alistair sat down with shyamal at the OpenAI office to discuss how our unique techniques have unlocked new levels of performance. We teach LLMs to mimic human reasoning and benefit dramatically from new models, including o1.

Cosine

23,517 次观看 • 1 年前

OpenAI’s newest model is finally here: o1. o1 represents an entirely new class of models designed to reason or “think through” complex problems— and it's already making huge leaps in domains like math and coding. For the very first episode of YC Decoded, we took a look inside.

OpenAI’s newest model is finally here: o1. o1 represents an entirely new class of models designed to reason or “think through” complex problems— and it's already making huge leaps in domains like math and coding. For the very first episode of YC Decoded, we took a look inside.

Y Combinator

92,680 次观看 • 1 年前

o1-pro is probably the best model i've used for coding, hands down i gave it a pretty complicated codebase and asked it to refactor while referencing docs the difference between claude/gemini/o1 and o1 pro is night and day. first time in a while i've been this impressed. full comparison in the video + code.

o1-pro is probably the best model i've used for coding, hands down i gave it a pretty complicated codebase and asked it to refactor while referencing docs the difference between claude/gemini/o1 and o1 pro is night and day. first time in a while i've been this impressed. full comparison in the video + code.

Sully

255,348 次观看 • 1 年前

🔥 🎥 ➡️ o1 FULL is available in Cursor NOW!!! watch to see how you can use this workflow in Cursor: >>> o1 is the ARCHITECT: brainstorm with o1 and it creates a step-by-step plan for implementation and then... >>> >>> Sonnet 3.5 is the ENGINEER: takes the o1 generated plan and implements it FAST shame that there's no access to o1 BEAST mode via API yet, but it's very useful to have o1 in Cursor and be able to connect it to your full codebase PSA: the o1 Cursor requests are charged on a pay as you go basis, so don't go too crazy and lose your house

🔥 🎥 ➡️ o1 FULL is available in Cursor NOW!!! watch to see how you can use this workflow in Cursor: >>> o1 is the ARCHITECT: brainstorm with o1 and it creates a step-by-step plan for implementation and then... >>> >>> Sonnet 3.5 is the ENGINEER: takes the o1 generated plan and implements it FAST shame that there's no access to o1 BEAST mode via API yet, but it's very useful to have o1 in Cursor and be able to connect it to your full codebase PSA: the o1 Cursor requests are charged on a pay as you go basis, so don't go too crazy and lose your house

Dan Mac

38,294 次观看 • 1 年前

Is Web3 Gaming the future or just hype? Let's break down what Web3 gaming promises, how some have given it a bad name, and how to do it right! Go to for more!

Is Web3 Gaming the future or just hype? Let's break down what Web3 gaming promises, how some have given it a bad name, and how to do it right! Go to for more!

Upland

44,233,734 次观看 • 1 年前

Gemini Agent uses Gemini 3’s advanced reasoning to break up complex tasks into smaller steps. With your permission, it can connect to your apps like Gmail or Calendar or even browse the web for you, under your guidance. Here’s how it works ↓

Gemini Agent uses Gemini 3’s advanced reasoning to break up complex tasks into smaller steps. With your permission, it can connect to your apps like Gmail or Calendar or even browse the web for you, under your guidance. Here’s how it works ↓

Google Gemini

105,637 次观看 • 6 个月前

Here's a tutorial on how to take footage and turn it into a fictive character using Kling O1 on Freepik

Here's a tutorial on how to take footage and turn it into a fictive character using Kling O1 on Freepik

Martin LeBlanc

12,540 次观看 • 6 个月前

Watch Illia (Illia (root.near) (🇺🇦, ⋈)) break down how NEAR Intents works in 4 easy steps at Solana Breakpoint NEAR Intents - The Universal Liquidity Layer

Watch Illia (Illia (root.near) (🇺🇦, ⋈)) break down how NEAR Intents works in 4 easy steps at Solana Breakpoint NEAR Intents - The Universal Liquidity Layer

NEAR Protocol

24,685 次观看 • 5 个月前

I went to the steps of USAID to tell the truth about how USAID protects America and the real reason Elon Musk wants to shut it down.

I went to the steps of USAID to tell the truth about how USAID protects America and the real reason Elon Musk wants to shut it down.

Chris Murphy 🟧

1,247,662 次观看 • 1 年前

ROUTE LLM - SUPER INTELLIGENCE IN THE MAKING RouteLLM combines the superpowers of o1-preview, Sonnet 3.5, Gemini Flash, o1-mini, 4o, and 4o-mini! Based on your query, it optimizes for latency, cost, and quality of the response. The cool thing about RouteLLM is it combines. - coding ability of Sonnet 3.5 - reasoning ability of o1-preview - hard-coding ability of o1-mini - 1M context of Gemini - speed of 4o-mini Similar to ensemble models that are very effective in classical ML, RouteLLM and similar models will be the most effective in the LLM ecosystem

ROUTE LLM - SUPER INTELLIGENCE IN THE MAKING RouteLLM combines the superpowers of o1-preview, Sonnet 3.5, Gemini Flash, o1-mini, 4o, and 4o-mini! Based on your query, it optimizes for latency, cost, and quality of the response. The cool thing about RouteLLM is it combines. - coding ability of Sonnet 3.5 - reasoning ability of o1-preview - hard-coding ability of o1-mini - 1M context of Gemini - speed of 4o-mini Similar to ensemble models that are very effective in classical ML, RouteLLM and similar models will be the most effective in the LLM ecosystem

Bindu Reddy

66,374 次观看 • 1 年前

Steps to get #NBADraft ready: 👨‍💻 do your research 👔 get the fit right

Steps to get #NBADraft ready: 👨‍💻 do your research 👔 get the fit right

Washington Wizards

15,038 次观看 • 1 年前

🔥 💻 🎥 How to provide o1 Pro with your FULL CODEBASE through the ChatGPT Cursor connection yesterday I recorded a video connecting o1 Pro to Cursor through the ChatGPT Desktop app more than one person commented on how limiting it is for o1 Pro to only have access to a single open file in Cursor here's a walkthrough of how you can provide o1 Pro with your FULL CODEBASE as context instead of just a single file: 1. Write a Python function that concatenates full codebase into a single snapshot in a .txt file (link in comment) 2. Open that "codebase_snapshot.txt" in a separate Cursor pane 3. Go to the ChatGPT app and ask it what context it has access to through Cursor - should say it has two files with one of them being the codebase snapshot and boom there ya go. o1 Pro has access to everything in your codebase full walkthrough here 👇

🔥 💻 🎥 How to provide o1 Pro with your FULL CODEBASE through the ChatGPT Cursor connection yesterday I recorded a video connecting o1 Pro to Cursor through the ChatGPT Desktop app more than one person commented on how limiting it is for o1 Pro to only have access to a single open file in Cursor here's a walkthrough of how you can provide o1 Pro with your FULL CODEBASE as context instead of just a single file: 1. Write a Python function that concatenates full codebase into a single snapshot in a .txt file (link in comment) 2. Open that "codebase_snapshot.txt" in a separate Cursor pane 3. Go to the ChatGPT app and ask it what context it has access to through Cursor - should say it has two files with one of them being the codebase snapshot and boom there ya go. o1 Pro has access to everything in your codebase full walkthrough here 👇

Dan McAteer

99,614 次观看 • 1 年前

Being a good trader has a lot to do with how your mind thinks. Below is a question i was asked during my interview for the fund I worked at right out of college. Can you get it right? Comment your answer and reasoning why. #TTG #stocks #trading

Being a good trader has a lot to do with how your mind thinks. Below is a question i was asked during my interview for the fund I worked at right out of college. Can you get it right? Comment your answer and reasoning why. #TTG #stocks #trading

Michael Edward

123,712 次观看 • 3 年前

Billionaire breaks Down "How to Become a Millionaire?" into Simple Steps

Billionaire breaks Down "How to Become a Millionaire?" into Simple Steps

Empire Mindset | Wealth Architect

88,521 次观看 • 8 个月前