Google launched Gemini 2.0 Flash Thinking, its “most thoughtful model yet” that’s capable of reasoning its way through complex tasks, in areas like programming, math, and physics. A bit like OpenAI’s o1 can. Powered by Google’s recently released LLM, Gemini Flash 2.0, Google’s reasoning model is supposedly quicker at responding than others (reasoning models take their time to ‘’think’’ and work their way through problems) and, although it appears to be similar in design to o1, it has one benefit that addresses a major concern with AI functioning as a ‘’black box, shrouded in mystery: It shows you what it’s “thinking”. It displays its workings out and explains why it’s come to the conclusion it has, offering users a more transparent experience and a clearer insight into how it works.
