Google recently unveiled a new AI model focused on “reasoning,” though it is still in the early experimental phase.
This new model “Gemini 2.0 Flash Thinking Experimental” can be found in AI Studio, Google’s platform for AI prototyping. According to its description, it is designed for strong “multimodal understanding, reasoning, and coding,” with the aim of tackling very complex problems in areas like programming, math and physics.
In a post on X, Logan Kilpatrick, who leads product for AI Studio, called Gemini 2.0 Flash Thinking Experimental “the first step in [Google’s] reasoning journey.” Jeff Dean, chief scientist for Google DeepMind, Google’s AI research division, said in his own post that Gemini 2.0 Flash Thinking Experimental is “trained to use thoughts to strengthen its reasoning.”
“We see promising results when we increase inference time computation,” Dean said, referring to the amount of computing used to “run” the model as it considers a question.
This new “Flash Thinking Experimental” version is built on Google’s freshly launched Gemini 2.0 Flash model and seems to share design similarities with OpenAI’s o1 and other models built around the concept of “reasoning.” A key difference with these types of models is that they essentially double-check their own work which helps them sidestep some of the common mistakes that plague regular AI.
However, this added scrutiny comes at a cost: reasoning models tend to take significantly longer, sometimes several seconds or even minutes to produce results.
When you give Gemini 2.0 Flash Thinking Experimental a prompt it does not immediately respond. Instead, it seems to pause and consider a range of related prompts, explaining its thought process as it goes. Eventually it pulls together what it believes is the most accurate answer.
This article was first posted by me on ''Medium.com.''
Detailed Article link: https://medium.com/@sadozye86/a-new-reasoning-based-ai-model-has-been-released-by-google-b6e0fc9825fa?sk=v2%2F402748a9-b75a-4755-85b0-7576630c7f29