I think AI work to date is not enough; many companies are merely trying to copy OpenAI instead of innovating.

AI video models are finally starting to work effectively, with new entrants like Hyo, Clling, and OpenAI's Sora making significant strides in the market.

The current AI boom is largely due to the Transformer architecture, and I believe there will be future breakthroughs of similar significance, though I can't predict them.

Meta views OpenAI as their biggest competitor, and I respect their aggressive approach to innovation, even if their current AI efforts haven't met expectations.

OpenAI may gain a temporary market lead, but competitors like Llama, Claude, and Gemini will quickly catch up with updates and improvements.

Many users have reported that the latest AI model, Operator with o3, feels very much like AGI due to its improved capabilities in using a computer effectively.

I think we have Jarvis today in the form of Operator, which is the most underappreciated new OpenAI launch. It allows anyone to use the web as a control surface to do anything, making every UI essentially an API.

Google's technology may surpass OpenAI's, but they struggle with execution and polish.

The development of neural operators allows us to create AI that can solve complex real-world problems, such as predicting hurricane paths, much faster and more accurately than traditional methods.