Google’s Gemini is now testing a brand new characteristic referred to as “Tasks.” This will probably be much like OpenAI’s Undertaking Function for ChatGPT.
With Tasks, you may add information, paperwork or your code. Then, you may ask Gemini to reference these information in undertaking conversations.
Google describes this as a characteristic the place you “begin by including information to the undertaking. Gemini can reference these information in undertaking conversations and use them to generate new paperwork and code.”
With Tasks, you may group related duties collectively and create your individual office inside Gemini.
It is unclear when the characteristic will start rolling out, but it surely might be restricted to enterprise clients initially.
Along with Tasks, Google is engaged on Gemini 3, which might be the corporate’s strongest mannequin thus far, and it might ship a giant blow to OpenAI.
Google rolls out Genie 3
Google not too long ago introduced Genie 3, which is developed by Google DeepMind, the identical workforce behind Gemini.
With Genie 3, you may simulate the actual world and switch a textual content immediate into a brand new, playable, interactive world you may transfer by way of in actual time.
As one consumer on Reddit discovered, you may even look down and see the way you’re strolling in a world simulation created utilizing Genie 3.
Google says Genie 3 renders the world at 720p and 24 fps, and retains the scene bodily constant for a number of minutes.
46% of environments had passwords cracked, practically doubling from 25% final yr.
Get the Picus Blue Report 2025 now for a complete take a look at extra findings on prevention, detection, and information exfiltration traits.
