OpenAI released its new o1-preview series of AI models, designed to spend more time thinking before they respond. These models can reason through complex tasks and solve harder problems than previous models in science, coding and math. In initial tests, the updated reasoning model performed on par with PhD students. In a qualifying exam for the International Mathematics Olympiad, the model scored 83%, compared to GPT-4o’s 13%.
After the new update’s preview models were released, users took to the internet to share their innovative projects using o1. Ten such projects that stood out are as follows.
Develop Games on the Go!
Karina Nguyen, a user made an AISteroid game with retro sci/fi vibe.
o1 is really good at making fun small games! for example, i made AISteroid Game w/ retro scifi vibes 🙂 https://t.co/cUZQaVnCEi pic.twitter.com/ZOMpCVkPIj
— Karina Nguyen (@karinanguyen_) September 12, 2024
Another user named Akhaliq combined o1 with Replit and Gardio to build a chess game.
The users were able to code and build games using the o1 module despite the previous GPT modules being not equipped to do so. Subham Saboo also created a space shooter game which he then ran on Reptile claiming o1 has changed coding and AI forever.
OpenAI o1 creates a fully interactive space shooter game in less than 2 minutes and Replit lets me run it in seconds.
AI and coding has changed forever. pic.twitter.com/toRmzaHock
— Shubham Saboo (@Saboo_Shubham_) September 13, 2024
Create Your Own App
A user, Ammar Reshi combined o1 with Cursor Composer and built an ios weather predicting app, with accurate predictions from scratch in 10 minutes with animation features. This module computed the coding and UI generating a response that tailor made the app from scratch.
Just combined @OpenAI o1 and Cursor Composer to create an iOS app in under 10 mins!
o1 mini kicks off the project (o1 was taking too long to think), then switch to o1 to finish off the details.
And boom—full Weather app for iOS with animations, in under 10 🌤️
Video sped up! pic.twitter.com/hc9SCZ52Ti
— Ammaar Reshi (@ammaar) September 12, 2024
Economics Essay Slay
Tyler Cowen, researcher at OpenAI asked o1 to write a college essay and unlike any other previous GPT module, OpenAI o1 responded with ease generating an in depth answer for the given prompt.
Watch “Tyler Cowen: OpenAI o1 & Economics” on #Vimeo https://t.co/dgaIwLIjoW
— tylercowen (@tylercowen) September 12, 2024
Genetics Solving
Catherine Brownstein, another researcher tested o1 to help her reason through “n of 1” cases; medical cases that nobody has ever seen and o1 was able to step up to the occasion and assist with the cases. o1 was able to understand complex genetic related queries and even solve equations for it generating positive answers.
7. Genetics pic.twitter.com/6iX3jo6ZDT
— Min Choi (@minchoi) September 12, 2024
Quantum Physics Computing
Mario Krenn used o1 to draft and reason through complex quantum physics equations, o1 responded better than any other version of GPT module generating quotations that fit the case. It decoded the problem, generated equations and solved them too. This module solved equations that renowned academics require brain power to do so and proved its competence to other GPT models.
2. Quantum Physics pic.twitter.com/4DGCiwGvOW
— Min Choi (@minchoi) September 12, 2024
Solving Puzzles
Jerry Tworke, a researcher ran a complex puzzle on the o1 module, trying to guess a person’s age in question to say a person’s relation to another person’s age and o1 took time to process the information and then solved the equation like a math problem revealing the answer with reasoning.
6. Solving complex logic puzzlepic.twitter.com/YIRy6RBFT0
— Min Choi (@minchoi) September 12, 2024
Translations Worldwide
Hyung Won Chung tried translating a korean script on the previous GPT module and it responded with no answer as it interpreted the language as gibberish and wasn’t able to understand the prompt. But with o1 it not only understood the prompt but also identified the language, decoded the text, enhanced translation by computing the meaning of the text and delivered impressive results.
8. Translates corrupted sentencespic.twitter.com/5x1m0BfOsB
— Min Choi (@minchoi) September 12, 2024
Reasoning Out Data
Jerry Tworke also tested the model for reasoning, typing a prompt based on physics and the physical relationship of objects, he concluded that humans would be able to logically reason out the outcome but LLM models aren’t trained to do so before, however, o1 after thinking the outcome stated the step by step process and answered positively.
4. Reasoning pic.twitter.com/9JZKGgUPRL
— Min Choi (@minchoi) September 12, 2024
Solving Complex Math Equations
Mason Meyer, a researcher used o1 module to generate a 5×5 nonogram, a pretty complex task for a mathematician or a LLM based AI program, but o1 after procession the question did generate the nanogram and prepared a grid layout for it, giving Meyer a puzzle that Meyer later asked o1 to solve for and it succeeded in doing so.
10. Coding pic.twitter.com/6KaC1RWrVb
— Min Choi (@minchoi) September 12, 2024
Coding For Beginners
It takes a lot of effort to run code that consistently runs well and o1 can potentially make it easier for every layman to code with its new upgrade. Scott Wu, a programmer tested o1 on human reasoning and code building and reported the results were positive.
10. Coding pic.twitter.com/6KaC1RWrVb
— Min Choi (@minchoi) September 12, 2024
