ChatGPT maker OpenAI has announced a new event, “12 days of OpenAI,” where new models and features will be revealed every weekday for the next two weeks.

The AI lab remains tight-lipped about specifics, but this could include any number of previously hinted at or revealed technologies that have remained hidden for any number of reasons, ranging from the risk of misinformation to operating costs.

Most predictions on social media point to the AI video generator Sora, a full version of the o1 inference model, and some improved control over voice in ChatGPT Advanced Voice using the Voice Engine.

CEO Sam Altman wrote to X: “Each weekday, we will have live streams at the launch and demos, some big, some stocking stuffers. We have some great stuff to share, hope you enjoy! Merry Christmas.”

OpenAI has not been sitting on its hands this past year. Since last January, the company has shipped GPT-4o, Advanced Voice, o1, ChatGPT Canvas, and more.

What we are likely to see in the next two weeks, as Sam Tha-Krause dons his big red suit, is that some of these announcements will be products we can use.

Of all the things OpenAI could announce at the “12 Days” event, Sora is probably the most anticipated, not as much as it was 6 months ago because of models like the Hailuo MiniMax, Runway Gen-3, and Kling It will be, but it's still a big story and, according to recent leaks, still top of the line.

We also expect a full release of the inferred o1 model. This will give us access to ChatGPT memory, GPT, and live search data, which will be very helpful in planning projects and refining complex ideas.

GPT-4o can generate images, but OpenAI has locked it down so no one can use this feature. This may become available in the next 12 days. Early demonstrations of this feature suggested impressive results and controls.

Advanced Voice is powerful, but could be improved. This could include access to live search and other features. It could also be able to provide real-world analysis through webcams and cell phone cameras.

We may see updates to Canvas, including features hinted at during its initial release, such as stacking elements of different canvases within Chat. inside ChatGPT, GPT-4o such as vector graphics and video analytics native features may be unlocked, making chatbots more useful.

We may also see further announcements such as o2 and GPT-5o previews. Outside of the language model, Sora-Turbo and Sora-full may be announced soon. Then there is Voice Engine, OpenAI's speech synthesis tool, and 'operator', an OpenAI agent framework that is rumored but not confirmed to be released next year.

Whatever is announced in OpenAI's 12 days, it will usher in a new era of artificial intelligence. Also, Google Gemini has turned one year old and I suspect that Google will make an announcement the same week. Where's Skynet when you need it?