OPENAI has changed the way its new Chatgpt O3-Mini model displays its chain of thought (COT) to “allow people to understand how the model”, but it faced an almost immediate backlash of users and accusations that He is a copy of the way Deepseek’s R1 model displays his reasoning.
Addressing Techradar, Openai said: “Users have told us that understanding the model model thanks to a response not only supports more enlightened decision -making, but also helps strengthen their confidence in his responses.”
“Although the raw bed of the model remains hidden because it is difficult to understand, we have found a balance: the model may think freely, then it organizes these thoughts so that they are easy to read. To improve clarity and security , we have added an additional post-treatment step where the model reviews the gross chain of thought, by deleting any dangerous content, then simplifies all complex ideas.
The new approach actually provides summaries of the model’s reasoning instead of showing you raw data. However, although the new approach is available in the O3-Mini for free and paid users and O3-Mini-High for paid users, the answer to X to its modifications was not completely positive.
Mayo Osshin replied on X, publishing: “We would appreciate if you have shown the complete thought channel, not just the summary version … Thank you”, and Conor replied: “It’s always a summary and not a real bed, which is disappointing. “”
Some other users have replied by suggesting that Optai simply responded to the threat offered by the new Deepseek by copying the way he presented the reasoning chain in his R1 model. “Finally, Deepseek changes the o-world for us,” replied Hamza. “So Openai copied the functionality of Deepseek’s thought chain?” said Ignis Rex, and “this moment when China is the only innovation, and the United States is the copier,” said Josip Tomo Licardo.
A big step forward
Personally, I prefer the new way of presenting the reasoning in Chatgpt O3-Mini. If you try the model now (just press the reasoning button) before entering your prompt, you will find that many more information is provided on how Chatgpt arrives at its conclusion. The previous scarcity of reasoning information, compared to Deepseek, was something that I criticized 03-Mini in recent days. So from my point of view, I think the new approach is a big step forward for Chatgpt O3-Mini.
I am not the only one to welcome new feature either. The pshichenko novel user said: “hilarious, the chain of thought is more endearing and funny than the rigid response” and “it’s a great improvement! I would love to see some examples of how this update affects his reasoning process, “said Jesus Vazquezon X.
