OpenAI’s third live product launch conference will be held at 2 a.m. this Tuesday, which will be quite a surprise.

Even OpenAI CEO Sam Altman specifically posted: "I am very, very excited about what will be released on the third day. Monday is too far away, can't wait."

In fact, you can see how high the expectations for the third day are from the number of user comments, which has exceeded 1,100.

The second release of OpenAI last Saturday was to strengthen the fine-tuning API. To put it bluntly, it sells o1's in-depth customization service, and the comments are only about 400, which shows that users don't care much about this, and they don't care too much about it. Big surprise.

Sam also discovered this, and he probably adjusted his strategy immediately and came out with something big to suppress the competition, otherwise the popularity of the next 9 days would definitely be getting lower day by day.

Some netizens speculate that the one that can bring surprises now is probably the Sora public beta.

Or advanced voice with visual mode?

Something even more exciting than Sora is AGI (Artificial General Intelligence).

Some netizens are not excited, feeling that OpenAI is always more gimmick than substantial.

Is it ChatGPT Ultra that costs US$1,000 a month?

Sora 2 is coming soon. (This generation is not in public beta yet, and there will be a second generation~)

I hope there will be major improvements in the advanced audio mode, such as new functional web search, visual enhancement, better voice, and the ability to read text from text mode to start and then switch to Advanced Audio mode to continue, and you can re-enter the old mode after ending the session.

There are still quite a few people looking forward to Sora.

GPT-5, agent? I feel like this guy’s guess is pretty reliable.

To be honest, OpenAI’s products in the first two days can only be regarded as quite satisfactory. There have been rumors about ChatGPT Pro and full-blooded o1 before. The second day is more like selling the o1 model. These two were released on the same day. More appropriate.

As for Sora, which has been hidden for 10 months, it is no longer fresh. There are many replacement products from foreign brands such as Luma and Runway, and domestic products such as Keling, Conch, Zhipu Qingyan and Tencent are open source. The effects of the Hunyuan video model are very stunning.

At present, generative AI mainly focuses on the four major tracks of text, pictures, audio and video. Sam said that he can give everyone a surprise, unless he breaks out of these four frameworks, just like Sora last year, it will once again detonate the development of the industry and become a benchmark.

Looking forward to a product live broadcast at 2 a.m. on Tuesday to see what is a surprise. If we continue to play beacon fire and trick the princes, we will lose the hearts of users.