Google releases the latest version of their Gemini AI
model
They also released an AI researcher, coding agent, and more.
Google unveiled their latest AI model, Gemini 2.0 Flash.
2.0 Flash is simultaneously more powerful and faster than previous Gemini models.
Google's releases strongly convey a focus on autonomous AI agents.
It may be OpenAI’s 12 Days of Shipmas, but don’t tell Google that.
The search giant released a whole bunch of AI heat on Wednesday, headlined by the latest version of their flagship model, Gemini 2.0 Flash, which Google describes as an “AI model for the agentic era.”
According to Google’s testing, Gemini 2.0 Flash outperforms even 1.5 Pro, and it does so at twice the speed.
Add in the fact that 2.0 Flash can natively use tools like Google Search and is multimodal, meaning it can generate text, images, and audio, and you have a model that is a significant improvement.
And you don’t have to just take Google’s word for it. The AI leaderboard Chatbot Arena also agrees.
The other releases
However, Gemini 2.0 Flash is not the only thing Google blessed us with. They also unveiled:
Deep Research, a tool that uses Gemini to search the web and write up a report based on its findings. Basically, it can do your research for you.
An updated version of Project Astra, an all-seeing, all-hearing and all-remembering experimental AI assistant.
Jules, an AI coding assistant that can autonomously fix software bugs.
Project Mariner, an AI agent that can take control of your Chrome browser.
Overall, it is clear that Google’s AI efforts are focused on agents, which are AI systems that can operate autonomously. They aren’t alone. OpenAI is also pouring a lot of energy into agents.
If AI agents are successfully created, you’ll suddenly have an AI that can truly do your work for you. Code. Write emails. Book vacations. You name it, and chances are it can do it.
That’s the future we’re racing towards. Now we’ll just have to see who can get there first.
My eyes tend to glaze over for AI updates these days but Project Mariner is the first AI agent to catch my attention in a while. Apparently it takes control over Chrome, moves your cursor, fills out forms, clicks buttons, etc. etc.
The devil's in the details but if this agent is actually competent I've got a half dozen different web apps to give it routine workflows for, Notion most of all.
It's fascinating to see how far AI has come. Google's latest releases are a testament to their innovation.
Some real promising advancements that I think are long overdue from Google, many were expecting much more from them, so this is very interesting to see!
Google's latest AI model, Gemini 2.0 Flash, sounds like a game-changer with its speed and ability to use tools like Google Search natively. The addition of tools like Deep Research and Jules shows how apps powered by AI are becoming more capable of handling complex tasks autonomously. It's clear Google is pushing hard in the race for advanced AI agents!
Woww! Google's Gemini 2.0 Flash is a major leap in speed and power, with multimodal capabilities and tool integration like Google Search. The focus on autonomous AI agents, like Jules for bug fixing and Project Mariner for browser control, shows the future of AI is here. Exciting times as Google and OpenAI race ahead!