Google launches Gemini 2.0 faster with ‘deep research’

Welcome to the new era of AI. Wait, you’re probably assuming we were already basking in the glow of artificial intelligence? Yes and no, at least according to Google. The technological monolith behind Gemini is ready to show us Gemini 2.0which Google says will allow agents to operate your phone or device and all of its apps on your behalf with “agent” AI.

What makes 2.0 a complete upgrade over the first instance of Gemini introduced in December last year (although it took several months to arrive on Android and iPhone with Gemini 1.5)? On the one hand, it is supposed to offer faster responses based on your prompts. While Gemini 1.5 can generate AI images with the company’s Image 3 model, the new version will also have AI audio output. The most important aspect of 2.0, Google saidis that it is made for AI agents.

If you haven’t heard the buzzword before, think of agents (or “agentification”) as multiple AI models that communicate with each other. The idea is that AI can essentially take over your phone when you ask it to. If you want AI to check your emails, make your reservation for your date, and then put it on your calendar, an AI agent should be able to take care of that.

This “universal assistant,” as Google calls it, starts with Gemini 2.0 Flash, which should be available starting Wednesday for all Gemini users. If you have Gemini Advanced, you get a new tool called Deep investigationa kind of artificial intelligence agent that does all the research on the Internet for you and then generates a large book report. Google’s tool is supposed to allow users to “do their own research” and generate reports in a few minutes. We’ll have to wait and see how many students try to pass these reports off as their own work.

Deep Research is technically an agent, but Google said more developers are also working to make agent AI work in their own apps. As for examples, the Mountain View, California company showed how a Gemini 2.0-based AI agent could create, evaluate, merge, and execute code on the fly. It’s a new “Jules” tool similar to Microsoft’s current GitHub Copilot.

The company also promoted a video of Gemini 2.0 interacting with several Supercell mobile games such as clash of clans. The AI ​​could read the screen and offer advice on the current Squad Busters meta. You could also remind players to complete their daily challenges to earn that sweet in-game currency. Is all that too exciting? No, not necessarily. The AI ​​training seemed superficial and didn’t offer any advice or strategy beyond creating picks that you could search for yourself between matches.

The real “agent” AI may be Google DeepMind’s long-awaited Project Astra. Gizmodo used it in an earlier version at Google I/O earlier this year. The tool is similar to Gemini Live but with many more viewing and interpretation capabilities with your phone’s camera. The new version is supposed to result in better, more conversational dialogue. It’s also supposed to remember your conversations and can now work with Google Search, Google Lens, and Google Maps.

It is not yet known when Astra will be available to more users. For now, the feature is simply in the testing phase and parts of it will likely be ported to various Gemini products in the future.

Google’s ‘deep dive’ model is supposed to make you feel like a Google search pro with Gemini 2.0

The Internet rabbit hole is deep, but Google says it has a new tool to dig through. If you can trust everything, leave the network. The company’s Deep Research tool first creates a “research plan” that is essentially a summary of an overall report. Then, you review a list of websites you think are applicable before laying it all out in a multi-page report, complete with some tables and graphs. It shows where you got your information from at the bottom.

The tool is available to all Gemini Advanced users in English starting today. It’s currently only available on desktop devices like Chromebooks or through the browser, although the mobile version should be available next year.

This whole process may take some time. For example, I asked Gemini Advanced to research Deep Mind’s history with Google before its acquisition in 2014. Gemini laid out its research plan, including DeepMind’s early stages and funding, through its academic publications. Google says you should select “1.5 Pro with Deep Research” from the drop-down menu, although it wasn’t available in my account at the time of writing. The AI ​​chatbot claimed it knew what it was supposed to do, so I asked it to finish the report; I began to wait… and wait.

Google’s blog mentions that the tool should take “a few minutes” to refine its analysis. I asked Gemini Advanced and asked how long the investigation would take. Gemini said it would take “between 6 and 12 hours to complete the report.”

The tool is obviously new and the AI ​​may not be accurate when telling me the schedule, but I can’t help but laugh. The AI ​​takes about the same time as it would take an average college student to create a last-minute report the night before it is due.

Google’s Deep Research runs on the larger Gemini model with a 1 million token context window, but that doesn’t mean it’s necessarily accurate. Google’s AI search may often select some websites that may not be the most accurate or may present sanitized information from untrustworthy sites. That’s why Gemini Tools reminds users to “double check” all their responses below the text message.