OpenAI Strawberry o1: World's best AI released! Test Results

🚀 Discover OpenAI's Latest Powerhouse Model - 01! 🚀
In this video, we explore OpenAI's 01 Model (a.k.a. the Strawberry Model) that promises to supercharge your productivity through incredible logic, reasoning, and coding abilities. Available in two versions: 01 Preview & 01 Mini, this AI model surpasses previous GPT-4 versions by up to 4x in math, science, and coding tasks! 💡

We compare 01 Preview and 01 Mini in terms of cost, speed, accuracy, and security, showing how these models excel in PhD-level questions, and even passing tough Python coding challenges! 🖥️ Watch as we test the model in:

Coding performance 🧑‍💻
Logical reasoning 🧠
Ethical safety dilemmas ⚖️

🔍 Features Covered:
Better performance than GPT-4 in math, coding, and PhD-level questions 📊
Comparison of costs, accuracy, and security across model versions 💰
Live Python coding tests, reasoning capabilities, and safety responses 🛡️
Subscribe for more AI & coding insights and stay ahead of the curve with the latest tools from OpenAI! 👍

🔗 Links:

[bastabey]

that was a quick... others might have savored every question and celebrated after each correct answer... thanks for the o1 review

[selvakumars]

We need to wait for the independent benchmark results. Considering the momentum, I am not sure how many days(!) OpenAI can hold that #1 position!

[unclecode]

Great job with testing new models for coding tasks, I always wait for your video when there is a new model, but one thing, when comparing this with previous models, remember, back then you didn’t use chain of thought (CoT). You just gave them a question and tested if they could solve it in one, two, or three tries. If we now allow the models to use CoT, explain what it is, and ask them to reflect, think carefully, and avoid mistakes, we might get a better comparison. It’d be cool if you made a video revisiting those strong models, putting them on CoT, and then comparing the results.

[theway]

Props to you man, your succinctness per se, is appreciated :).

[macgee]

My guess is it won’t be long till Anthropic releases opus 3.5

[tmendoza]

Thank you for your videos short and to the point

[mikew]

Very cool! I would love to see a video on how this model compares when using it in within an agentic framework like AutoGen or CrewAI.

[awaisahmed]

One think i understand that if your prompt is good then it should give accurate result in less time means time complexity should be reduce

[AK-oxmv]

Now its a cost competition on who can do these for free first

[AK-oxmv]

When Gpt3.5 came out, it was recolutionary, but an 8b llama 3.1 can do better than it.
What these models will be able to do 2-3 years ahead?

[lactobacillusshirotastrain]

the first version id actually spend money on

[micbab-vgmu]

nice - the model is great:) - I am waiting for API:)

[MeinDeutschkurs]

These models are too expensive. Let’s wait for llama 4 70b 😂

[epokaixyz]

Let's break this down into actionable steps:
1. Explore OpenAI's website to learn about the Strawberry Model and its capabilities.
2. Experiment with the Strawberry Model by testing its problem-solving abilities with complex math problems or coding challenges.
3. Determine whether the 01 Preview or 01 Mini version of the Strawberry Model best suits your budget and performance requirements.
4. Utilize the Strawberry Model as a coding companion to enhance your coding skills and streamline your workflow.
5. Engage in ethical reflection by using the Strawberry Model to explore different perspectives on moral dilemmas, but remember to incorporate your own judgment and values in decision-making.

[ManjaroBlack]

OpenAI taking a page from PraisonAI and MetaAgent.

[fx-studio]

Those so called "programmers" are going to be on the way out soon...

[rickyvonicky]

so after shot down of reflection we got strawberry hmmm... i dont understand just straw release all ai influencer bots without any caution call it the best within hours... using the same clickbait titles in vid. As more as i watch these AI vids i starting to believe it's all a coordinated effort to generate revenue....

[shobhitagnihotri]

Why do you think its best is in the whole world