The ChatGPT o1 model escapes preview mode for a while
OpenAI accidentally released the full ChatGPT o1 model on the AI chatbot platform a few days ago. The leak only lasted a few hours, but that was enough time to show that the model will be even more formidable than the brilliant preview version of the riddle solver that was introduced to the platform in September.
ChatGPT o1’s advancement over the current standard ChatGPT-4o model is undeniable, even as an example. It is more accurate, better at solving complex multi-step problems, and can write sophisticated software code. However, the preview does not include all the features available in the standard ChatGPT-4o, such as data analysis, visual insight and the ability to search the web.
These tools were all present in the final model before OpenAI yanked it. In those few hours, users showed how ChatGPT o1 solved visual puzzles, described photos in great detail, and could shut down a text conversation based on a screenshot, even replicating the emojis used. Below you can see some examples.
ChatGPT o1 + image is in the wild!!!Use this linkhttps://t.co/RdHmy53hYYFound via @Jaicraft39 in quotes. Here’s a quick test with one of my images (you’ve seen this image below in my feed). Please note: the model name is simply “ChatGPT”. Expect this to be connected soon. I… https://t.co/tTcXOVALWq pic.twitter.com/pTRCyATtXUNovember 2, 2024
o1 first contact 👽brought to you by me + @Jaicraft39 pic.twitter.com/MPrMxbk5sqNovember 2, 2024
ChatGPT o1 thinks slowly but deeply
The big difference with ChatGPT o1 compared to its predecessors is that it must use reason and logic to break questions and tasks into smaller parts and solve them all before responding. That should make him less likely to hallucinate and better able to put his answer in the context of the question. OpenAI hasn’t made any announcements about specific release dates for ChatGPT o1, but the more public testing, even the unintentional ones, suggests it will happen very soon, perhaps before the end of the year.
There’s plenty of reason to think it will be a boon for OpenAI as ChatGPT faces increasingly sophisticated competition from the likes of Google Gemini, Meta AI, and other AI chatbots. The slower response speed is largely compensated by the more accurate and useful answers given to questions about text or images. Personally, I’m curious if the humor and flirting shortcomings of ChatGPT o1 preview have been fixed in the final version.