The feature, only available for Labs users in the US in English, lets you generate «quick, conversational audio overviews» for «certain» search queries.
That should let the audio play for you while you are multitasking or turning attention elsewhere.
«It gives a lay of the land», Google says, and The Verge notes it can also be set to a feature known to many NotebookLM users — with two simulated «podcast hosts» enthusiastically debating the results for you.
You can opt into the feature on Google Labs, and the feature will then pop up as an alternative under Google’s AI overviews on the results page.
Google has launched a freakishly accurate Hurricane forecast, currently being tested with the U.S. National Hurricane Center. (Picture: Screenshot)Nvidia is building in Europe!
The AI chip maker announced at the GTC Paris conference that they are working with partners in European countries to build both infrastructure and factories, marking probably the largest AI investment so far on the continent. More at Investor’s Business Daily, and Nvidia’s press release.
A novel approach to the AI embargo in China
Chinese AI companies have found a route around the embargo of advanced AI chip sales to the country. Much like the early days of desktop publishing, they have taken to flying suitcases full of high density hard drives to neighboring Malaysia, to hook them up to a nicely unrestricted supercomputer and process the data. More at The Wall Street Journal. See also: Sneakernet.
Massive Google Cloud outage affects just about everyone
Google’s cloud service went down from 11:46 until 14:23 PST yesterday, affecting a lot of internet services, like Spotify, Cloudflare, Discord and Snapchat. It also affected certain login features at OpenAI, impacted most services at Anthropic and, of course Google’s own Gemini, listing the entire time span as «full outage.» More at: TechCrunch, and r/singularity.
Google AI with high precision hurricane forecasts
Google DeepMind & Google Research have launched a model that simulates 50 storm scenarios up to 15 days out, now being trialed with the U.S. National Hurricane Center. Early results show forecasts are ~87 miles more accurate than Europe’s ECMWF model. It’s a smarter, data-rich companion to traditional physics systems—and a potentially big step forward in saving lives. More at The Verge and check it out at Google’s weather lab.
You can schedule your prompts and get mail and calendar overviews to start your day with Gemini. (Picture: Google)Pro and Ultra subscribers in the Gemini app can now preset prompts or tasks at any given time, a feature that may be quite helpful for some.
They say this can be useful for frequently occurring tasks, like getting a morning news summary or providing ideas «for your blog.»
Works with Google apps
Gemini does of course also work with your Google Calendar and Gmail, and can provide you with a list of unread emails and your daily schedule.
It should be available in the Gemini app, and also for free users, who get between three to ten questions per session. This is the model Google plans to take out of Preview as a stable, full model at a later time.
Google has started testing «talk to search» on Android and iOS. The feature lets you discuss and refine search results by clicking under the search bar in the Google app, by text or voice — and you can keep talking after leaving the app.
Anthropic blocks Windsurf access to Claude. Windsurf is an AI coding application that lets you choose which models to use, and the block comes «just weeks after Bloomberg reported that OpenAI was acquiring Windsurf.» «It would be odd for us to be selling Claude to OpenAI,» says Anthropic.
Anthropic opposes 10-year moratorium on state AI laws. In a New York Times op-ed, CEO Dario Amodei argues that AI is developing so fast that new laws might be needed well ahead of the ten-year mark currently being considered in the US Senate. «I believe that these systems could change the world, fundamentally, within two years; in 10 years, all bets are off,» he writes.
Sharing NotebookLM links could become the next internet fad, if Google gets its way. (Picture: Google)Google has a smash hit on its hands with the AI app, and is now expanding access to notebooks with a simple link.
The shared service lets other users interact with your stored notebooks, to ask questions and make podcasts of your research. You can restrict access to only chats, if you like.
Google could lose Chrome and golden default deals in its antitrust trial. (Picture: Pietro & Silvia, CC BY 2.0)The question for the courts last day of closing arguments was this: Are AI competitors also operating in the search market?
Both sides argued they weren’t, as it could interfere with long standing DOJ arguments that Google is a search monopoly, and for Google looking to avoid scrutiny of its recent behavior.
A young teen was allegedly encouraged into suicide. Now the question is whether an AI chatbot can be held accountable. (Picture: Character.ai)In a Florida court case that could one day define First Amendment rights for AI, a judge has declined to consider the defendant’s arguments for dismissal — setting the stage for a trial showdown over whether AI chatbots are faulty products or entities entitled to rights previously reserved for humans.
The case revolves around a teen whom Character.ai allegedly encouraged into suicide over a long interaction, and whether the developers should be held accountable for his tragic death, in a wrongful death lawsuit.
If you let it, Gemini can scan through your past emails to write a new one. (Picture: Google)Buried inside the complex blog posts from the Google I/O keynote yesterday, lies an interesting little nugget titled «The power of personalization.»
In it, Google explains how it can make Gemini and search better by tailoring responses to the vast data Google keeps on you from prior searches, docs in Google Drive — and even your past emails in Gmail.
All of this data is under Google’s control and is infinitely searchable by an AI, which won’t have to spend much time to get to know you «better.»
Google showcased the glasses on stage at Google I/O, highlighting their translation capacity. (Picture: Google)Easy to miss the torrent of AI news at Google I/O was a demo of their upcoming AI-powered smart glasses, expected «after 2025.»
The glasses pair with Gemini on your phone, has an in-lens display, camera, ear buds and microphones and can show you your Android apps.
More importantly, though, is the integration with the AI. The glasses become Gemini’s eyes and ears and will see all you see and listen to conversations around you.
So much news. Google brings a firehose of new AI features to Google I/O. (Picture: Google Imagen 4)Google’s I/O conference kicked off with a tsunami of AI news last night. Clearly, the company is going all-in on artificial intelligence, and is using all the levers in its power to promote it — even in Chrome.
The most important news they presented is AI Mode going live in the USA, using what is likely the most valuable screen real estate on the planet — Google’s front page — to leverage the new AI search feature. It’s been in testing for about a week.
Google DeepMind’s AlphaEvolve can even improve on itself. (Picture: Google)The model is already at work and producing savings internally at Google, having discovered more efficient algorithms and better hardware solutions.
— AlphaEvolve is a Gemini-powered AI coding agent that is able to make new discoveries in computing and mathematics, says Matej Balog, a researcher at Google DeepMind, to VentureBeat.
Google doesn’t say when the model will be available to the general public, but is planning on working with certain scientists and is «exploring» a broader release.
Can you spot the difference? Google is experimenting with a change on the front page. (Picture: screenshot)Replacing the time honored «I’m Feeling lucky,» select users now see «AI Mode» instead.
By doing this, albeit in their Labs section, Google shows they are considering using some of the most valuable screen real estate on the planet to further promote its Gemini AI models.
The era of free AI chats may be coming to an end, with google experimenting with ads. (Picture: Google)The internet display ad behemoth is moving with the times, as the AI industry seeks to monetize. It’s likely only a question of time before we get ads in Gemini.
AdSense apparently began experimenting with ads for AI chatbots, or, as they call it «conversational AI,» in chatbots iAsk and Liner, according to Bloomberg, which is paywalled.
Just a test? SEO roundtable quotes Bloomberg as saying that AdSense had «expanded to include conversations with chatbots operated by AI startups.»
Gemini 2.5 Pro opens even more avenues for coding, says Alphabet. They are taking that to heart. (Picture: Google) With Satya Nadella’s announcement overnight that Microsoft uses AI to code around 30% of their software, AI coding has come of age. Other Big Tech companies have also reported similar numbers lately.
— I’d say maybe 20%, 30% of the code that is inside of our repos today and some of our projects are probably all written by software, said Microsoft CEO Satya Nadella during a conversation at LlamaCon, Meta’s AI developer conference, according to CNBC
Microsoft says they have made progress with Python, and less with C++ projects, but he is still pointing that some projects could be entirely written by AI.
The feature, only available for Labs users in the US in English, lets you generate «quick, conversational audio overviews» for «certain» search queries.
