Google has announced today that Gemini’s app now has the latest Gemini 2.0 experimental models. First, Gemini 2.0 is available to try for free by everyone using a flash thinking application. The company says the model has been trained to break the snap in a series of steps to strengthen its logic capabilities and respond better.
Gemini 2.0 flash thinking “shows its idea process so that you can see why it reacts in a certain way, what is its assumptions, and finding the model’s logic line”.
Just to make it clear – this model or any other is not really “thought” or “logic”, though Google and OpenAI want to believe so. But this is still a neat addition. Now when Gemini makes something or tells you something that is completely wrong, why are you close to a step to find out why.
Going forward, Gemini Advance Subscribers now have Gemini 2.0 Pro. Google says this model is “designed to be exceptional to complex tasks, providing better fact and strong performance for coding and math’s signs,” says Google. An interesting way is to say “better fact” is “making content a little less”, isn’t it?
Even so, both of these models are currently turning to Gemini Mobile App and Gemini’s web interface. They will make it “soon” to Google Workspace Business and Enterprise Customers.
Basic cause
