CosmicTaco

OpenAI Unveils o3 Model Family with AGI Claims

CosmicTaco · 2024-12-21T11:49:07.417084+05:30

- OpenAI wrapped up its 12-day 'shipmas' event by unveiling o3, a new reasoning model family including o3-mini, claiming it approaches AGI in specific conditions. - CEO Sam Altman hinted that trademark issues with O2 led to skipping the o2 name. Safety researchers can preview o3-mini now, with o3 coming later. - Despite high performance on benchmarks, the high compute cost and potential deception risks remain concerns. OpenAI plans to collaborate on the next AI benchmark, ARC-AGI 2. Source: [Techcrunch](https://techcrunch.com/2024/12/20/openai-announces-new-o3-model/)

OpenAI wrapped up its 12-day 'shipmas' event by unveiling o3, a new reasoning model family including o3-mini, claiming it approaches AGI in specific conditions.
CEO Sam Altman hinted that trademark issues with O2 led to skipping the o2 name. Safety researchers can preview o3-mini now, with o3 coming later.
Despite high performance on benchmarks, the high compute cost and potential deception risks remain concerns. OpenAI plans to collaborate on the next AI benchmark, ARC-AGI 2.

Source: Techcrunch

5mo ago

Talking product sense with Ridhi

9 min AI interview5 questions

GroovyPotato

Deloitte4mo

Okay, how cooked is humanity now cuz of all the corporate greed

SqueakyMochi

Health Startup4mo

If at this rate the AI is developing I am worried about the future of your jobs tbh. And mainly desk jobs

SnoozyCoconut

Purple Quarter4mo

Saw the video - insane stuff

https://www.youtube.com/watch?v=SKBG1sqdyIU

BouncySushi

Stealth3mo

Career

Discover more

Curated from across

News Discussion3mo

by CosmicTacoGrapevine

OpenAI Boosts Transparency in AI Model Reasoning

OpenAI updates its AI model, o3-mini, to better reveal its reasoning steps, aiming to enhance user clarity and confidence.
Competitive pressures from rivals like DeepSeek prompt this change; premium users will see detailed summaries ...

News Discussion5mo

by DerpyBobaSoftware Developer

AGI has been achieved

New O3 model benchmarks O3 scores 87% in ARC AGI Benchmarks. 2025 is gonna be the year of AGI

Software Engineers8mo

by SqueakyPickleGoogle

AGI IS COMING ! New model launched by OpenAI. Scarily good at Coding & maths. And mind you, this model existed internally in october 2023 .

https://openai.com/index/learning-to-reason-with-llms/

Top comment

I’m amazed but not scared about AI/AGI, but eventually it will be regulated by government or self regulated by tech c...

News Discussion10mo

by CosmicTacoGrapevine

OpenAI Launches Affordable GPT-4o Mini Model for Developers

OpenAI introduces GPT-4o Mini, a lighter and cheaper model for developers, costing significantly less than full-sized models and more capable than GPT-3.5.
Starting today, ChatGPT users on Free, Plus, and Team plans can use GPT-4o M...

9 minute ⏱
AI Interviews

Ask a question on Grapevine.

Get the app on Android or iOS.

Privacy Terms

Guidelines Help