At its Google I/O developer showcase on Tuesday, Google unveiled a slew of new artificial intelligence tools as well as AI models. AI was so central to Google I/O that CEO Sundar Pichai counted the number of times someone said “AI,” which came out to a generous 121. Keeping track of everything can be pretty challenging, so here’s a quick breakdown.
Google AI Tools
- Veo: Project VEO is Google’s video generation tool, which lets you create videos based on video, image or text prompts.
- AI Teammates: Google’s Project Gemini-powered “virtual teammate” — which it called Chip as a sample during an on-stage demo — can be used for situations like checking up on the progress of a work project, as the company demonstrated on stage.
- Project Astra: Google’s Project Astra, which is sort of like an evolution of Google Glasses, employs your phone’s camera and AI to do things like find your lost keys or misplaced glasses and locate things that are making sound.
- Trip planning: Google teased its Gemini AI assistant’s travel planning capabilities — though they don’t constitute an entirely new tool per se — by having the assistant generate a personalized trip, complete with flight itineraries and hotel reservations.
- Fraud-call monitoring: Gemini can monitor incoming phone calls and identify suspicious, scam calls by displaying a warning pop-up message.
- Circle to Search: You can use a gesture to select something on your phone — by circling it or highlighting it, for instance — and get additional information. Circle to Search is poised to become available on 200 million Android devices by the end of 2024.
- TalkBack: To assist low-vision and blind users, the TalkBack screen reader describes what’s on a screen.
- Gmail AI features: The company’s email client gets a bevy of Gemini-powered AI features, including Smart Reply, which offers personalized answers based on an email thread and suggested replies.
- AI Overviews: Google’s Search Generative Experience, or SGE, isn’t new, but it’s rolling out to all users in the US. The feature summarizes information at the top of your Google search results page.
- Imagen 3: Google’s text-to-image generator, Imagen 3, can create pictures from prompts.
- SynthID: SynthID lets you watermark AI-generated pictures and videos.
Google AI models
- Gemini 1.5 Pro model: Google Gemini — Google’s ChatGPT rival — benefits from an update to make it more intelligent, conversational and adept at handling complicated tasks.
- Gemini 1.5 Flash: As its name suggests, Gemini 1.5 Flash prioritizes quick answers because it’s a lighter-weight version of Gemini when compared with 1.5 Pro.
For more information, check out our comparison of Google Gemini 1.5 Pro and Flash.
https://www.cnet.com/rss/all/
Moe Long
