AI

Continuing to push the widely unpopular data center projects, Trump said in a Truth Social post that some of them “have become extremely popular when a financial deal is struck.” His plan sounds lot like Meta’s approach on the Hyperion project in Louisiana, where Mark Zuckerberg has touted $50,000 bonus checks for teachers despite a secretive approval process residents didn’t know much about.

The Wall Street Journal published a big profile of the Google DeepMind co-founder, who says his timeline for the arrival of artificial general intelligence (AGI) is becoming “nearer and nearer.”

Xbox CTO Scott Van Vliet has been “cooking up retro vibes” as part of Microsoft’s new Xbox Labs effort. The result is what looks like a vibecoded option to pick your dashboard on an Xbox console. This is simply an experiment right now and it’s not clear if we’ll ever see such an option, but Xbox CEO Asha Sharma asks “blades on the web. Do we like this experiment?”

OpenAI describes Orchestra Bio as “a platform that helps biopharma teams coordinate the complex process of drug development.” The company will bring Orchestra’s team and tech “together with Rosalind, our AI models and tools for life sciences research.”

Over the last several months, the AI industry has set its sights on some of the problems that have stymied mathematicians for years. In some cases, AI models have made huge progress — though exactly how much isn’t always obvious. The Verge’s Robert Hart explains what OpenAI and others have actually accomplished, the disastrously bad ways the companies have interacted with researchers, and whether there’s a future left for human mathematics.

The feature analyzes video and audio signals to determine whether they “may have been altered outside of Zoom.” The video conferencing app is also launching a tool that uses AI to make presenters “appear more engaged,” as well as an AI-powered feature that smooths out voices and accents in real time.

The AI assistant, which launched in September, is only the 12th app to have reached this milestone within 40 days, according to market intelligence firm Sensor Tower. Part of its success is likely down to what Sensor Tower describes as a sizable “advertising push by Meta.”

This update allows you to perform tasks across your apps and files without transmitting data to the cloud, but Perplexity recommends 36GB of RAM or more. It uses Apple Silicon to run Qwen 3.6 or PPLX Pro (a post-trained version of Qwen) directly on your Mac, while allowing you to transfer certain tasks to the cloud for access to updated information and browser use.

Yelp’s new “Check for Me” feature is starting to roll out on iOS as part of Yelp’s fall product release. The launch also includes updates to Yelp AI’s assistant, including more personalized recommendations and the ability for it to book reservations on OpenTable.

ByteDance, Alibaba, and Tencent have reportedly collectively ordered more than 3 million of Nvidia’s RTX Pro 5500 GPUs, among the most powerful still compliant with US export controls. Those orders far exceed planned production, however, and Reuters reports that Nvidia has warned customers they might only get 5 to 10 percent of what they requested.

A European Commission page noting tech acquisitions has a listing under Apple dated June 17th for a company called Samplab GmbH, as reported by AppleInsider. An archived “About” page for Samplab notes that “AI will revolutionize the music industry.”

Creator life involves using a lot of software. Bad software. Stephen Robles is one of the creators who decided to build something better, for himself and for his community. He tells David how he thinks about his responsibility as a developer, how he makes apps, and how AI is changing the way he builds things for his audience.

Digital Sun thanked Steam players for signing up for a first playtest of an upcoming game called ReVamp, but said that:

The AI note-taking app will use your sources to suggest topics for a slide deck or video, while allowing you to confirm decisions on formatting, visual style, and scope before it starts creating. You can also turn on a new “Review” toggle that lets you look over a final outline before Gemini Notebook’s final rendering.

Microsoft CEO Satya Nadella said earlier this month that we should assume that all AI models are “compromised,” and now Microsoft is going one step further. “Unfortunately, frontier models, regardless of their origin, cannot be trusted,” says Hayete Gallot, executive vice president of Microsoft Security. Microsoft is naturally positioning its new Project Perception as a solution to securing AI models. It’s an agentic security system designed to help cybersecurity professionals secure and control AI agents.

The Department of Homeland Security ordered the agencies to remove any technology created by the Chinese AI company over concerns that its models may transfer “​sensitive data to the PRC,” according to Reuters. The ban also prevents civilian agencies from using third-party services that integrate DeepSeek’s AI models.

Reuters reports that Microsoft pulled out of the deal with British company Nscale to build the AI data center after local electricity grid officials said the site won’t be supplied with power until the 2030s. Construction is still expected to go ahead without Microsoft’s support, and Nscale says it’s partnership with Microsoft on other projects “remains strong.”

The suspect, Abdulrahman Alfozan, used WhatsApp to communicate with the unnamed royal throughout 2025 and 2026, according to a court filing spotted earlier by CNBC.

Wired reports that Character filing for bankruptcy and agreeing to sell assets to Disney last week was the second of “two interrelated and integrated” transactions. The first, which took place in September, involved Disney hiring a “majority” of the chatbot startup’s employees, a familiar move from the “acquihire” playbook.

The new Anthropic Public Benefit Fund will be a wholly owned LLC, which the company says will give it “flexibility.” It offered little information about how it would use the money, other than saying it wouldn’t fund lobbying or political donations. It offered limited examples of what it might do:

The big problem with smart TVs is that they start to slow down, stutter, and generally suck pretty quickly. Ansub Khan figured out a way to get Claude to clean up the mess and make his old Android TV feel new again. Best part: he lays out step-by-step how to do it on your own TV. Though, as with anything involving AI, proceed cautiously.

In February 2025, US District Court of Delaware Judge Stephanos Bibas ruled against Ross Intelligence, which argued that training its legal AI on data scraped from Thomson Reuters was fair use. Now an appeals court has upheld that decision, though it remains sealed.

It’s not just celebrity deepfakes that are plaguing social media. AI hot girl accounts are stealing comedians’ jokes, and painters are seeing AI versions of their process videos rack up millions of views. Mosseri says it’s “not practical” to ban AI, but it’s not sure how to handle content that isn’t copyrighted, like jokes.

During an interview with Bloomberg, Charles Schwab Corp. CEO Rick Wurster called out Meta’s AI agent, Muse, saying:

With public opinion broadly against AI data centers, the Today, Explained podcast from Vox brought on Nicole Thomas, a marketing and PR exec who represents data center companies.