It’s no secret that Google’s flagship AI chatbot Gemini has had some problems. Its production of historically inaccurate images forced Google parent Alphabet to suspend the product temporarily.
But Google is trying to turn the page on its early AI mishaps. The company’s platform Vertex AI, which can manage a range of generative artificial intelligence models through Google Cloud, now includes the ability to “ground” those models. That means responses on Gemini, Google’s own AI model, will be linked to “verifiable sources of information,” the company said Tuesday.
The announcement was part of a range of updates to Gemini, in a push to sell the AI product to corporate customers at the tech giant’s annual Google Cloud Next conference in Las Vegas.
Google Cloud CEO Thomas Kurian told Bloomberg that “grounding” is “improving correctness of answers.” Because of the changes, Kurian said, “Enterprises have been piloting with us a number of scenarios with generative AI; now they’re deploying them in production.”
The latest model of Google’s chatbot, Gemini 1.5 Pro, has been developed by by the company over the last six months and is also now capable of “long context understanding.” That means it has a greater ability to analyze audio, video, and text to offer businesses tips.
“For example, a gaming company could provide a video analysis of a player’s performance, along with tips to improve,” Google said in its announcement Tuesday. “Or an insurance company could combine video, images and text inputs to create an incident report, making the claims process easier.”
This is a developing story and will be updated.
More AI news
Spotify’s new AI playlist tool has been dubbed ‘Spotify’s ChatGPT’
Microsoft says a China-backed group is using AI misinformation to sway foreign elections
Google might charge for AI-powered search
Elon Musk says OpenAI is ‘aggressively recruiting Tesla engineers’ in ‘the craziest talent war I’ve ever seen’