“The new model, Gemini 2.0, can help with coding projects and Google se virtual arches, generating images and audio in multiple languages.”
- Google unveiled a new iteration of its flagship AI model, which it claims is twice as quick as its predecessor and will power user-assistance virtual agents.
- According to the business on Wednesday, the latest model, Gemini 2.0, can help with coding projects, Google searches, and producing sounds and graphics in other languages.
- Tulsee Doshi, a director of product management at the business, told reporters that the new Gemini capabilities “make it possible to build agents that can think, remember, plan, and even take action on your behalf.”
- Google, a division of Alphabet Inc., has been attempting to ensure that its dominance in search and advertising is not weakened by the newest generation of AI technologies promoted by OpenAI and other firms. Although the company has managed to maintain its market position in search thus far,
- OpenAI is exerting pressure on the industry leader by integrating additional search functions into ChatGPT. The ultimate goal of both businesses is to create artificial general intelligence, or software that can carry out activities just as well as or even better than people.
In an interview, Koray Kavukcuoglu, chief technology officer of Google DeepMind, an AI lab, stated, “We want to build that technology — that is where the real value is.” “And we are attempting to choose the appropriate applications and problems to solve to get there.”
Google added more AI to its search engine
- which is still its lifeblood, in addition to its experimental goods. The business announced this week that it would start testing Gemini 2.0 in search and AI Overviews, virtual summaries driven by AI that appear at the top of Google searches.
- This will enhance the speed and caliber of search results for more complicated queries, such as sophisticated mathematical formulas. On Wednesday, Google also made Gemini 2.0 Flash, the company’s quick and effective AI model, available to developers. According to Google, it could process images more effectively and simulate human reasoning.
- Google unveiled “deep research,” a new online tool that claims to allow Gemini users to utilize AI to delve deeply into subjects with in-depth reports. Users of Google’s premium AI subscription service, Gemini Advanced, will be able to access the function on Wednesday.
- It is marketed as an AI-powered research helper. According to the company, Gemini users throughout the world will have access to a chat-optimized online version of the experimental Gemini 2.0 Flash. In the upcoming year, additional Google products will use the concept.
- The items highlighted on Wednesday demonstrate how Google DeepMind, the company’s top AI unit, is becoming increasingly important to product development. Project Astra, an AI agent that processes visual input using a smartphone camera, is undergoing more testing at the lab. Google staff demonstrated how
Astra may condense the information on the website in a lavish area that resembles a home library, complete with tall bookshelves holding works on travel and computer programming.
In a small art gallery accessible through a secret entrance tucked away in the shelves, the agent considered how Norwegian painter Edvard Munch’s “The Scream” encapsulated both his nervousness and the prevailing paranoia of his time.
However, the agent continued to exhibit certain virtual limitations.
It was unable to identify whether any titles were on the bookshelf during a live presentation with a reporter from Bloomberg.
Since its debut at Google’s historic developer virtual conference earlier this year, the agent has advanced and can now react conversationally at the same rate as a human, according to DeepMind researcher Greg Wayne.
According to Wayne’s interview, the agent virtual used to have trouble with the name of DeepMind CEO Demis Hassabis, mistaking it for a request for information about Damascus, the capital of Syria. However, it now manages that and other requests with ease.
“Creating AI with eyes, ears, and a voice to assist you in the real or digital world has been the founding motto,” Wayne stated.
The flash device the business is releasing is “a lot more capable than anything that we had a year ago at a fraction of the cost,” Kavukcuoglu added, comparing where they were a year ago to where they are now.
