AI News: Fable Banned, New Open-Source Leader, Midjourney Shocker
This AI news roundup covers the US government forcing Anthropic to shut down its Fable 5 and Mythos 5 models due to a security vulnerability jailbreak, the release of a competitive open-source model GLM 5.2 by ZAI, and MidJourney's surprising pivot into medical imaging technology with a new ultrasound-based body scanner.
Summary
The video opens with what the host calls the biggest AI story of the week: the US government forced Anthropic to shut down its Fable 5 and Mythos 5 models for all users globally on June 12th. The shutdown stemmed from a jailbreak vulnerability discovered by a trusted testing partner — later revealed to be Amazon CEO Andy Jasse — who reported it to senior Trump administration officials. Because Anthropic could not selectively block foreign nationals from accessing the models, they had to disable them entirely. The host notes the irony that Anthropic CEO Dario Amodei had spent months warning about the dangers of these models and had even published an essay arguing the government should have the power to block or reverse AI model releases — yet when the government did exactly that, Anthropic pushed back, calling the jailbreak minor. David Sacks is cited as saying the administration acted reluctantly and found Anthropic's dismissiveness 'bewildering.' The host speculates the conflict is less about security and more about a broader antagonistic relationship between Anthropic and the US government.
The second major story covers ZAI's release of GLM 5.2, a 753-billion parameter open-weight model with an MIT license and a 1-million token context window. On coding benchmarks like Swebench Pro and the web dev arena, it outperforms GPT 5.5 and comes close to Claude Opus 4.8, while costing significantly less per token. The host tests it live, finding it capable of generating slides and eventually producing a working (if imperfect) game clone after multiple prompts, concluding it does not yet match Fable 5 for one-shot coding tasks.
MidJourney's surprise announcement of 'MidJourney Medical' is discussed at length. The company revealed an ultrasound-based whole-body scanner that submerges users in water and uses ~9,000 transducers to image internal structures, claiming it operates at a fraction of the cost and time of an MRI. Rather than selling to hospitals, MidJourney plans to open 'MidJourney Spas' starting in San Francisco in 2027, combining body scanning with sauna and wellness amenities. The host notes MidJourney has no outside investors and is self-funding this venture. However, Hank Green's critique is referenced — arguing that ultrasound cannot replicate what MRIs or CT scans do, particularly for lung or colon imaging, and that the marketing overstates the technology's capabilities.
In the rapid-fire section, several additional stories are covered: OpenAI's Codex 'record and replay' feature that learns repetitive computer tasks from screen recordings; Claude Design's new canvas editing and cross-tool integration features; Perplexity's 'Brain' self-improving memory system for its cloud-based computer agent; a Claude-integrated video editor from Palmir; Adobe adding AI assistants to Premiere, Illustrator, and InDesign; Google's 'Ask Ad Manager' chatbot for ad optimization; and a Pew Research survey showing ~50% of US adults now use AI chatbots but remain broadly skeptical of AI's societal impact. The video ends with a brief mention of a self-driving robotic toilet designed for people with mobility issues.
Key Insights
- The host argues that Anthropic essentially invited government intervention by spending months publicly claiming its models were dangerous enough to 'wreak havoc in the wrong hands' and publishing an essay arguing the government should have FAA-style authority to block or reverse AI model releases — yet then dismissed the jailbreak vulnerability as minor when the government acted on those very principles.
- David Sacks is cited as saying the Trump administration issued the export control reluctantly, and found it 'bewildering' that Anthropic refused to comply with safety requests it had previously claimed were its highest priority — suggesting the shutdown was driven as much by Anthropic's dismissive attitude as by the vulnerability itself.
- Amazon CEO Andy Jasse was identified as the whistleblower who raised security concerns about Anthropic's models to senior Trump administration officials, which the host calls bizarre given that Amazon is one of Anthropic's largest investors and cloud vendors, making the disclosure appear contrary to Amazon's own financial interests.
- ZAI's GLM 5.2 ranks second only to Claude Fable 5 on the web dev arena leaderboard, beating Claude Opus 4.8 and GPT 5.5 by a significant margin, while costing approximately $1.40 per million input tokens compared to Claude Opus 4.8's $525 — representing roughly a 375x cost difference for near-comparable coding performance.
- Hank Green argues that MidJourney's marketing of its ultrasound scanner as an MRI replacement is misleading, because ultrasound cannot image through air-filled organs like lungs and cannot replicate what a CT scan does for detecting lung or colon cancer — meaning the device complements but does not replace existing diagnostic imaging modalities.
Topics
Transcript
[0:05] It was a wild week in the world of AI and I don't want to waste your time. So, let's just get right into it. Starting with the story that pretty much rocked the AI world this week. For the first time ever, the US government forced a publicly released commercial AI model off of the market. Enthropic had to shut down Mythos 5 and Fable 5 for every user on Earth. And in last week's news video, I spent a good chunk of the video talking about how much I loved this model. But unfortunately, we only got it for less than a week. On Friday evening, [0:36] June 12th, it was gone. So, here's what happened. The…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Matt Wolfe
AI News: Claude's Massive Leap & Siri Gets Good!?
This AI news roundup covers the release of Claude Fable 5 (a Mythos-tier model from Anthropic) and its controversial safety restrictions, Apple's WWDC AI announcements including a major Siri overhaul, and updates from Google including NotebookLM upgrades and real-time translation. Additional rapid-fire items include OpenAI and SpaceX IPO filings, ChatGPT email sending, and a teased Midjourney hardware device.
Shopping Online Is About To Change Forever
The video introduces 'agentic commerce,' a new AI-driven shopping paradigm where AI agents proactively match users to products before they search. The platform Glance is highlighted as a leading example, using selfies and personal data to generate personalized outfit recommendations with direct purchase links. The creator frames this as a major evolution in e-commerce beyond chat-based AI search.
Microsoft Build Recap in 82 seconds
Microsoft Build in San Francisco featured seven new in-house AI models, including a flagship reasoning model, a coding model, a transcription model, and a voice generation model. Microsoft also entered the AI agent space with Microsoft Scout, giving OpenAI direct access to Microsoft products and Windows management. The announcements signal a major push by Microsoft into competitive AI across multiple domains.
How Big Tech Lies About AI Layoffs
The transcript argues that major tech companies are using AI as a cover story for layoffs that are actually driven by post-COVID over-hiring and a desire to cut costs. Industry leaders like Nvidia's Jensen Huang and DeepMind's Demis Hassabis have publicly called out this practice as lazy and irresponsible.
AI News: Microsoft Finally Reveals They're Plan!
The video covers major announcements from Microsoft Build and Nvidia's Computex event, including Microsoft's seven new in-house AI models, the Microsoft Scout autonomous agent, and Nvidia's RTX Spark chip. The host also recaps new image generation models, voice models, and updates from OpenAI, and includes an interview with Microsoft AI CEO Mustafa Suleyman about AI self-sufficiency, healthcare AI, and the philosophy behind building superintelligence.