Google launches new Gemini Spark AI agent that can work for you

Google just launched Gemini Spark, a new AI agent that can take work off your hands, figure out how to get it done, and come back with the finished product.

The company launched the service on Thursday via Google Cloud, offering enterprises an AI system capable of conducting research, writing code, preparing reports, and completing assignments.

As stated on Google’s website, one is able to assign work to Google Assistant today and get it done way after they have shut down their laptop. This approach is certainly unique, unlike that of assigning an agent a task to write an email, only to do everything else by oneself.

In September, OpenAI came up with dots agents, which work for its users all the time. Meta (META) followed with Muse, which can handle purchases, travel reservations, emails, and payments. Now Alphabet (GOOGL) wants a piece of that business market. Its shares gained 2.2% following Thursday’s announcement.

Google’s new AI can work across your apps without being babysat

The interesting part about Gemini Spark is how much responsibility Google wants businesses to hand over to it.

Say an employee needs information pulled from several departments, organized into a spreadsheet, and turned into a presentation-ready report. Instead of separately requesting each step, the employee can assign the entire job to Gemini.

The agent works out what needs doing, identifies the right software, collects the necessary information, and prepares the requested output.

It can operate inside Gmail, Docs, Sheets, and Calendar, but Google isn’t keeping it locked inside Workspace. Companies using Microsoft 365 or Slack can also access the service.

That flexibility extends to devices. Employees can use Gemini through browsers, iPhones, Android phones, Windows computers, Macs, and even developer command lines.

Software companies can also connect it to their own products without requiring users to open a separate Gemini window.

Another feature tackles one of the more irritating things about working with AI: having to explain yourself again and again.

Since Gemini works via the Google Cloud, it maintains the same working memory and context wherever the employee accesses it from.

An employee could start working on a task using their office PC, track its progress via their smartphone, and continue the next day without having to restart.

Most importantly, the assignments running for a long time period do not rely on the device being constantly turned on.

The employee can schedule assignments to be performed later or have them initiated by the agent at certain events within a connected system.

Google has also built in a way for Gemini to divide difficult assignments between several smaller AI workers.

These temporary sub-agents can tackle different parts of a project simultaneously or complete them in a particular order. Gemini manages their activities and combines the results.

Google is giving AI coworkers their own corporate identities

Google is taking the idea further with what it calls coworker agents, with which companies can create digital workers with ongoing responsibilities, individual storage, and even company email addresses ending in @agents.company.com.

An AI coworker could handle different assignments over several days, communicate with other agents, and maintain its own working history.

There are boundaries, though. Google says each coworker receives access only to the company information employees choose to share with it.

Businesses can also establish dedicated project environments, allowing Gemini to build familiarity with particular tools, instructions, and workflows.

The system uses past interactions to better understand how someone works, rather than treating every request as an entirely new conversation.

Perhaps the more unusual decision is that Google isn’t making Gemini depend exclusively on Google’s own AI models.

The platform currently uses both Gemini and Anthropic’s Claude, selecting whichever is better suited to a particular assignment.

A complicated coding problem might require a more capable model, while something straightforward could run on a cheaper one.

Google argues that this flexibility could help companies control expenses without sacrificing performance on demanding work.

It also intends to introduce other commercial and openly available models as the technology develops. Businesses would therefore be able to switch between AI engines while keeping their existing information, configurations, and work history.