Echo Wang / Reuters: Source: SpaceX plans to raise $75B in its IPO by selling 555.6M shares at $135 each, an unusual move since most companies set a price range before the roadshow  —  SpaceX plans to raise $75 billion in its initial public offering by selling 555.6 million shares at a target price of $135 per share …

Jason Koebler / 404 Media: Source: Google asked Google Play app devs to join a “confidential content offer pilot”, offering to pay for access to codebases to use them to train AI tools  —  Google is trying to buy code from some Android developers as part of a “confidential” program.

Discussion | Link

Scott Pelley spent over 30 years at CBS News before his exit this week. Frazer Harrison/WireImage "60 Minutes" correspondent Scott Pelley is out after over three decades at CBS News. Pelley reportedly had harsh words for Bari Weiss during a heated staff meeting on Monday. Executive producer Nick Bilton said Pelley was fired, citing the journalist's "misconduct." Longtime "60 Minutes" correspondent Scott Pelley is out of the famed news program. Pelley, who spent over 30 years at CBS News, was fired on Tuesday from the company after a clash with new "60 Minutes" top producer Nick">Nick Bilton. "Your antipathy to the future of the show has come through loud and clear," Bilton wrote in a memo to Pelley that was seen by Business Insider. "I therefore write on behalf of CBS News, Inc. to inform you that your employment with CBS is terminated for cause effective immediately." A CBS News spokesperson declined to comment. Pelley's departure follows a testy staff meeting with Bilton on Monday, an audio recording of which was obtained by Status and The New York Times. During one exchange, Pelley reportedly said CBS News editor in chief Bari Weiss "does not love this place," and is "murdering" the "60 Minutes" newsmagazine. In Bilton's memo, the executive wrote that he had tried to reach out to Pelley but was instead rejected and met with an "ambush." "Despite yesterday's misconduct, I had hoped that in sitting down with you today we could find a path forward together," Bilton wrote. "You made clear that you are not interested in such a path." In a statement after his termination, Pelley slammed "60 Minutes" management, saying he had been directed, but refused, to "inject falsehoods and bias into a politically sensitive story." "The collapse of values at the top has become untenable. The leadership of 60 Minutes is no longer recognizable. The principles I hold dear are gone, and so I must leave as well," Pelley wrote. CBS News under Weiss Weiss has characterized her overhaul of CBS News — including "60 Minutes" — as part of a mission to remake the organization for the digital era. Her critics say she's trying to make the broadcast network more friendly to President Donald Trump, whose administration is tasked with approving or rejecting CBS-parent Paramount Skydance's plan to buy Warner Bros. Discovery. Weiss has pushed back against criticism and said her actions are not political. During the Monday staff meeting, Pelley told Bilton he had "slender qualifications" for the top job at "60 Minutes," according to The New York Times. Bilton, formerly a New York Times tech reporter who's also produced documentaries for HBO and Netflix, doesn't have a TV news background. Pelley also grilled Bilton on the departures of correspondents">correspondents Cecilia Vega and Sharyn Alfonsi from "60 Minutes" last week. In an exit memo, Vega said that she and her colleagues battled "efforts to insert political bias into our stories" that she characterized as "censorship, both imposed and self-driven." Alfonsi referenced "an intense editorial dispute" with Weiss, who delayed her December story about the Trump administration's migrant deportation tactics involving El Salvador's CECOT prison. This decision by Weiss "was a deliberate choice to penalize a journalist for refusing to sanitize factually accurate reporting, and it sends a chilling message to the entire newsroom," Alfonsi wrote. Weiss told staffers that her decision to delay the story wasn't politically motivated. Read the original article on Business Insider

Ben Thompson / Stratechery: Google selling $10B of equity to Berkshire is fascinating and functions as a signal that AI demand is high, in a world where the most cash buys the most compute  —  What does the most beautiful business model of all time look like?  —  First, imagine that your supply is free.

Bloomberg: Bitcoin fell as much as 7% over 24 hours to below $67K on Tuesday for the first time since April, following Strategy's first BTC sale since 2022; MSTR fell 9%  —  Souring Bitcoin sentiment has triggered almost $1.5 billion in crypto liquidations over the past 24 hours, as the largest digital asset sank back to a two-month low.

Google is trying to buy code from some Android developers as part of a "confidential" program.

Daphné Leprince-Ringuet / Sifted: Oxford Quantum Circuits, a UK-based startup building and deploying quantum computers, has raised a $350M Series C led by Bullhound Capital  —  UK company has deployed quantum systems to partner data centres across Europe, the US and Asia  —  Oxford Quantum Circuits (OQC) …

When the puck drops tonight in Game One of the NHL Finals between the Vegas Golden Nights and the Carolina Hurricanes, it will mark the 134th year for the Stanley […] The post NHL Scores a Hat Trick with VAST Data appeared first on BigDATAwire.

Even in a future of robot taxis, someone still has to return the things passengers leave behind.

Mo Gawdat advises job seekers to focus on human skills to thrive in the AI-era job market. Kate Green/Getty Images for BoF Mo Gawdat advises job seekers to focus on human skills to thrive in the AI-era job market. Gawdat suggests "human-centric jobs" like nursing or counseling. Tech leaders have highlighted communication and judgment as vital skills AI struggles to replicate. An ex-Google executive has tips for surviving the AI era. Mo Gawdat, the former chief business officer for Google, predicts that 30% of certain job sectors will be gone by 2028, but people can still "double down on human skills" to overcome the turbulence. "We have an entire generation that is out of college today that will struggle, unfortunately," Gawdat said on an episode of Steven Bartlett's podcast "The Diary Of A CEO" that aired on Sunday. "And my advice to them is learn the tool and focus on human-centric jobs." "A lot of people can make a living by being a nurse or by being a counselor or by being, you know, anything that connects to humans," Gawdat added. His comments come as tech leaders increasingly point to communication, relationship-building, and having good judgement as skills that remain difficult to automate. OpenAI president Greg Brockman recently said that "taste" is "a new core skill." He said that while AI can create things quickly, it takes judgment to decide what to keep. Duolingo CEO Luis von Ahn said in a May podcast interview that AI still can't match the work of the company's artists and designers, and that he won't let AI erode the quality of his app. And Salesforce CEO Marc Benioff said in an earnings call in May that while the company's engineering head count has remained "mostly flat" at around 15,000 employees over the past two years, it is still adding workers in the sales department. "Because, I think we all realize the one thing that we are doing here with you, selling and communicating, the agents are not exactly doing that," Benioff said during the earnings call. Gawdat also said on "The Diary Of A CEO" that even though AI may bring about uncertainty, AI is not the enemy and can be harnessed to create opportunities. "By definition, the better you are at using an AI to do your job, the more likely you are to be successful," said Gawdat. "Learn how to interact with AI. Welcome AI into your hybrid world of work." Read the original article on Business Insider

"It’s a reminder of how human activity is changing the natural world in unanticipated ways.”

A government committee says that the country’s growing dependence on the data analytics company is a serious liability.

One hardware announcement and several software highlights from Microsoft Build.

The cybersecurity company is nearing a $300 million round led by Evolution Equity Partners.

Bloomberg: Sources: a Broadcom-backed portion of Apollo and Blackstone's $36B debt deal to buy TPUs for Anthropic to lease may yield ~5.75% vs. 8%-9% for a riskier portion  —  Broadcom Inc.'s promise to backstop the largest portion of a record-setting microchip financing deal for Anthropic is keeping yields …

<p>Microsoft <a href="https://microsoft.ai/news/building-a-hillclimbing-machine-launching-seven-new-mai-models/">announced two new text LLMs</a> this morning - <strong><a href="https://microsoft.ai/news/introducing-mai-thinking-1/">MAI-Thinking-1</a></strong> (reasoning, 1T parameters, 35B active, available to "select early partners") and <strong><a href="https://microsoft.ai/news/introducingmai-code-1-flash/">MAI-Code-1-Flash</a></strong> (137B Parameters, 5B active, "purpose-built for GitHub Copilot and VS Code to deliver high performance and lower cost [...] rolling out to GitHub Copilot individual users in Visual Studio Code"). I've not been able to try either of them just yet.</p> <p><strike>It's very interesting to see Microsoft releasing models with such low parameter counts, especially given how expensive larger models are to access right now. They claim MAI-Thinking-1 "is preferred to Sonnet 4.6 in our blind human side-by-side evaluations", which is impressive for a 35B model seeing as I frequently run models larger than that on my own laptop.</strike> (UPDATE: I got this entirely wrong, see note below.)</p> <p>Also <a href="https://microsoft.ai/news/introducing-mai-thinking-1/">of note</a>:</p> <blockquote> <p>We trained [MAI-Thinking-1] from the ground up on enterprise grade, clean and commercially licensed data, without distillation from third-party models.</p> </blockquote> <p>And for <a href="https://microsoft.ai/news/introducingmai-code-1-flash/">MAI-Code-1-Flash</a> as well:</p> <blockquote> <p>It is built end-to-end by Microsoft using clean and appropriately licensed data.</p> </blockquote> <p>I would <em>very much</em> like to learn more about this "appropriately licensed" data! Could these be the first generally useful code-specialist models that didn't train on an unlicensed dump of the web? (<strong>Update</strong>: the answer is no, see note below.)</p> <p><strong>Update</strong>: My initial published notes got the size of the models wrong. I misread Microsoft's announcements and interpreted the MoE active parameter count as the total parameter count, but the <a href="https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF">model card for MAI-Code-1-Flash</a> lists it as 137B with 5B active and the <a href="https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf">MAI-Thinking-1 technical paper</a> reveals it to be a 1T model with 35B active.</p> <p>I deeply regret this error.</p> <p><strong>Update 2</strong>: That technical paper describes the training data in some detail from page 80 onwards. It has the same licensing problems as all of the other major LLMs: it's trained on a crawl of the public web:</p> <blockquote> <p>The majority of our web HTML corpus comes from a proprietary crawl. After initial page discovery and selection, approximately 1.2 trillion pages are crawled and parsed. [...] In addition to Microsoft standard policy Sec. 2.4, we apply UT1 block list (Prigent, 2026) to remove adult content and piracy-related domains. In all, this filtering reduces the corpus from 1.2 trillion pages to 794 billion pages. Given the prevalence of AI-generated content on the web, we also score pages with a proprietary AI-content detection model and use manual inspection to identify domains with extensive AI-generated content; those domains are filtered out of the training corpus.</p> <p>[...]</p> <p>We process Common Crawl with the same pipeline. [...] After filtering, deduplication, merging with the proprietary web corpus, and a final round of exact-URL and content-level fuzzy deduplication, the Common Crawl portion contains 24.2 billion pages.</p> </blockquote> <p>I did not cover this one at all well, which is somewhat ironic since I was at the Microsoft Build conference when I wrote this up! I'm sorry for not digging deeper before publishing my initial notes.</p> <p>Tags: <a href="https://simonwillison.net/tags/llm-release">llm-release</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/microsoft">microsoft</a>, <a href="https://simonwillison.net/tags/llms">llms</a>, <a href="https://simonwillison.net/tags/training-data">training-data</a></p>

The vendor aims to demonstrate that enterprises can use its AI models and agent-building tools on its cloud ecosystem.

Samantha Subin / CNBC: Palo Alto Networks reports Q3 revenue up 31% YoY to $3B, including $388M from CyberArk and Chronosphere, vs. $2.94B est., and forecasts Q4 revenue above est.  —  - Palo Alto Networks topped Wall Street's third-quarter estimates, citing AI advancements