ChatGPT For Mac Gets Major Update on the 12th Day Of OpenAI

With the festive season approaching, several companies are seizing the opportunity to bring out new and exciting deals, campaigns, and promotions. OpenAI has also joined the celebrations with the thrilling ‘12 Days of OpenAI’ event series. This series was started on the 5th of December and the grand finale for the same was on Friday, December 20. 

The event continued daily for 12 weekdays, each featuring a live stream with a launch or demonstration. The launches included both major innovations and smaller surprises and on each Friday something new whether big or small was presented to the users. 

The most monumental announcement was reserved for the final day of this 12-day ‘shipmas’ celebration. On the last Friday, the company introduced o3, the successor to the earlier o1 reasoning model. To clarify, o3 represents a family of models which are similar to o1. It comprises o3 and 3-mini and is a more compact version fine-tuned for specific tasks. 

OpenAI has claimed that o3, under certain circumstances will approach the realm of AGI, although with important warnings. Right now, neither the o3 nor the o3-mini is widely accessible but safety researchers can register for a preview of the o3-mini beginning from the launch. 

A preview for o3 will follow at an unspecified date later and according to Sam Altman, CEO of OpenAI, they plan to launch o3-mini by the end of January with o3 to follow. 

However, this timeline contradicts his recent comments, in an interview earlier, Altman expressed a preference for a federal testing framework to look over the monitoring and risk mitigation of new reasoning models before release. 

Unlike other AI systems, reasoning models such as o3 have the ability to effectively fast-check themselves. It further helps them to navigate some of the common challenges on their own that are typically a hindrance for other models.