Google introduces its new AI model: Gemini 2.0 Flash Thinking Experimental
Google has announced Gemini 2.0 Flash Thinking Experimental, a new "reasoning" AI model currently in the experimental phase. The model promises to offer reasoning capabilities for complex problems, though it still requires further development. What is Google's new reasoning model? When will we be able to use it?
12punto
Google has announced a new AI model called "reasoning," though this model is currently in the testing phase.
The new model, Gemini 2.0 Flash Thinking Experimental, is available for use on Google’s AI prototyping platform, AI Studio.
It is stated that this model is optimized for multimodal understanding, reasoning, and coding, and is capable of logically solving the most complex problems in fields such as programming, mathematics, and physics.
AI Studio product manager Logan Kilpatrick described Gemini 2.0 Flash Thinking Experimental in a post on X as "the first step in Google's reasoning journey." Jeff Dean, chief scientist at DeepMind, Google’s AI research division, said in his own post that this model was "trained to use thoughts to strengthen its logic."
<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Want to see Gemini 2.0 Flash Thinking in action? Check out this demo where the model solves a physics problem and explains its reasoning. <a href="https://t.co/Nl0hYj7ZFS">pic.twitter.com/Nl0hYj7ZFS</a></p>— Jeff Dean (@JeffDean) <a href="https://twitter.com/JeffDean/status/1869790032296579169?ref_src=twsrc%5Etfw">December 19, 2024</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Built upon the recently announced Gemini 2.0 Flash model, Gemini 2.0 Flash Thinking Experimental appears to have a design similar to OpenAI’s o1 and other reasoning models.
Gemini 2.0 Flash Thinking Experimental pauses before responding, evaluates a series of relevant requests, and reaches a conclusion by explaining its logic. After a while, it summarizes the answer it deems to be the most accurate.
Although it currently produces errors in some tests, it is considered promising. It has not been definitively stated when the testing period will end.