Tech Brew Ride Home

Anthropic backtracked on secretly degrading Fable 5 for AI researchers after fierce backlash. OpenAI considers drastic token price cuts anticipating war with Anthropic. Dario Amodei calls for FAA-style AI regulation, the FBI seized fake Chinese consulting domains, and DoorDash launches AI ordering.


Anthropic backtracks on its decision to quietly limit Fable 5's ability to develop LLMs, saying "requests will visibly fall back to Opus 4.8", after backlash (Wired)


Sources: OpenAI is considering drastically lowering the prices it charges users for tokens in anticipation of similar cuts the startup expects Anthropic to make (WSJ)


Dario Amodei outlines policy responses to AI's exponential progress across regulation and public safety, macroeconomics and taxes, science, geopolitics, more (Dario Amodei)


The FBI seizes 13 domains allegedly tied to fake consulting firms that sought information from US government and military employees for suspected Chinese agents (Reuters)


YouTube rolls out a new in-app messaging system for sharing videos and having 1:1 conversations; it discontinued its previous Messages feature in September 2019 (9to5Google)


DoorDash launches an in-app AI chatbot to let users order food and groceries and make reservations with photos and prompts (CNBC)

Learn more about your ad choices. Visit megaphone.fm/adchoices

The CFTC proposed prediction market rules banning bets on war and assassinations. SpaceX drew $250B+ in IPO demand far beyond its $75B target. Microsoft patched a record ~200 security flaws, Tesla has just 59 robotaxis, and Siri AI gets its first hands-on.


Sources: the CFTC will propose new prediction market rules, banning bets it finds aren't in the public interest or that seem highly susceptible to manipulation (WSJ)


Microsoft's June 2026 Patch Tuesday fixes a record nearly 200 security flaws, 33 rated Critical, as it and security researchers increasingly use AI to find bugs (Krebs on Security)


Sources: SpaceX has drawn $250B+ of investor demand for what is poised to be the largest IPO to date, far beyond the $75B that SpaceX is aiming to raise (Reuters)


Tesla has just 59 vehicles in its robotaxi fleet and is limited to three Texas cities after almost a year, nowhere close to Elon Musk's big promises (Bloomberg)


Hands-on with Siri AI: successfully executed multistep prompts, understood context well, has strong guardrails, and seems a bit more dispassionate than Gemini (The Verge)

Learn more about your ad choices. Visit megaphone.fm/adchoices

Anthropic released Claude Fable 5, a guardrailed Mythos-class model, to the public and Mythos 5 to trusted partners. OpenAI confidentially filed for an IPO. Hackers injected credential-stealing malware into 70+ Microsoft GitHub repos, and Apple details its new Gemini-based foundation models.


Anthropic releases Claude Fable 5, a "safe" Mythos-class model it says can't be used for cyberattacks, to the public, and Claude Mythos 5 to trusted orgs (Wired)


OpenAI confidentially files for an IPO, says it has "not decided on timing yet", as "there are things we want to do that are likely easier as a private company" (CNBC)


Microsoft disabled 70+ of its repos on GitHub, including Azure-related tools like azure-functions-host, after hackers added credential-stealing malware to them (TechCrunch)


MG Siegler: after being left for dead in AI, Apple is set to win at the consumer level — the power of the default, superior product instincts, and no real competition (Spyglass)


Ben Thompson: the iPhone is the true core of Siri AI, and Apple is the only company positioned to work across apps with personal context — as long as it's not vaporware (Stratechery)

Learn more about your ad choices. Visit megaphone.fm/adchoices

Apple unveiled Siri AI at WWDC, built on Google's Gemini, launching in beta with no firm date. Google signed a $920M/month SpaceX compute deal, Trump floats government equity stakes in AI labs, and OpenAI plans to turn ChatGPT into a superapp.


Apple announces Siri AI, with on-screen awareness, personal context understanding, a Dynamic Island-based interface, and more, along with a new Siri app (9to5Mac)


Filing: Google has agreed to pay SpaceX $920M per month for access to Nvidia chips as part of a cloud-services deal that runs through mid-2029 (Bloomberg)


President Trump says he is weighing proposals for US government to hold equity stakes in leading AI labs, and will soon discuss the idea with their executives (Bloomberg)


OpenAI plans to overhaul ChatGPT in the coming weeks, turning it into a superapp with coding tools and AI agents to serve as a gateway to higher-margin products (FT)

Learn more about your ad choices. Visit megaphone.fm/adchoices

Bots passed human web traffic for the first time, per Cloudflare's CEO. The S&P 500 rejected fast-entry for mega-cap IPOs like SpaceX. Anthropic embedded engineers at the NSA, Meta hid face-recognition code in its app, and Cambridge trialed the first AI-designed vaccine.


Cloudflare CEO Matthew Prince says agentic traffic is "growing so fast that bots have now passed human traffic online for the first time" (Tom's Hardware)


S&P Dow Jones rejects proposals to expedite S&P 500 eligibility for mega-cap IPOs such as SpaceX's; companies remain ineligible until one year after their IPOs (Bloomberg)


Sources: Anthropic has embedded around half a dozen forward-deployed engineers within the NSA to help the agency deploy Mythos for offensive cyber operations (FT)


Analysis: Meta discreetly added code for an unreleased "NameTag" face-recognition system for its AI smart glasses over multiple Meta AI app updates in 2026 (Wired)


University of Cambridge researchers say they have developed the first vaccine with a key component entirely designed by AI and subsequently trialed it in humans (BBC)

Longreads


A preview of what to expect from WWDC on Monday, including iOS 27, a revamped Siri, macOS 27 Liquid Glass refinements, and more (Bloomberg)


Ted Chiang argues LLM conversations are cleverly disguised sentence continuation, not consciousness, and that no intrinsic property of neural networks suggests otherwise (The Atlantic)

Learn more about your ad choices. Visit megaphone.fm/adchoices