This episode features a host providing a weekly AI recap format designed for busy professionals and colleagues who need quick updates on AI developments. The show aims to make AI news accessible through condensed 5-minute summaries.

The week was dominated by Anthropic's Fable-5 release - the first publicly available Mythos-class model - which became both the biggest story and biggest controversy. The episode covers the model's capabilities, restrictive guardrails, data retention policies, and the government-mandated shutdown affecting foreign nationals.

Additional coverage includes the SpaceX IPO's strong market debut and its implications for future AI company public offerings, plus the emerging trend of 'token panic' as companies begin capping employee AI usage to control costs.

Fable-5 Launch: First Public Mythos-Class Model

Anthropic released Fable-5, the first publicly available Mythos-class model after months of institutional access through Project Glasswing

Citrini Research noted that 'normal people can't really determine whether new models are better than previous ones' while '150 IQ' users recognize dramatic improvements

Fable-5 demonstrates superior strategic thinking and first principles argumentation, being less manipulable and more valuable as a thought partner

The model is only available on normal plans until June 22nd, after which users must pay through API on usage basis

Anthropic's Triple Crisis: Power, Policy, and PR

Biomedical engineer Daria Anutmaz reported: 'I can't even say hello to Fable except in incognito mode because it knows I'm a biomedical researcher'

Data retention policies allowing 30-day conversation storage prompted Microsoft to limit employee use while enterprises assess implications

Anthropic secretly nerfed LLM research responses without user notification, causing unprecedented backlash and forcing policy reversal within 24 hours

US government directive forced Anthropic to cut off all foreign national access to Fable-5 and Mythos-5, resulting in complete model shutdown

SpaceX IPO Success Signals Market Optimism

SpaceX stock popped 19% on first trading day, with Allspring Global Investments' Robert Grundyke stating it 'bodes well for the market and for these other IPOs'

'They clearly priced it right, at least on day one. It should make you optimistic for the markets, especially for growth stocks' - Grundyke

Strong performance creates positive outlook for upcoming AI company IPOs, though sustained performance beyond first day remains to be seen

Token Panic: Corporate AI Usage Caps Emerge

New 'token panic' trend sees companies implementing caps on employee AI consumption, with Uber as first major example followed by Meta

Citadel Securities research note highlighted negative market consequences of corporate AI usage restrictions, though social media misinterpreted findings

Trend indicates push for token efficiency rather than demand rollover, with implications for cost management across organizations

Resources Mentioned

noted that 'normal people can't really determine whether new models are better than previous ones' while '150 IQ' users recognize dramatic improvements Fable-5 demonstrates superior strategic thinking and first principles argumentation

vailable Mythos-class model after months of institutional access through Project Glasswing Citrini Research noted that 'normal people can't really determine whether new models are better than previou

The Girl Who Tweeted Wolf (Hobson & Choi Book 1)

you were using it for really powerful things, you wouldn't necessarily know the difference. Citrini Research tweeted, "I think we've reached the point where normal people can't really determine whethe

note by Citadel Securities

t broke that Meta was going to be putting those caps on as well. This was given voice in a specific research note by Citadel Securities, although as I explained in my Friday episode, what they were sa

The Writing Rope A Framework for Explicit Writing Instruction in All Subjects

l it pushes back on you without being just subservient to your opinion. 3 and 4, I would test it on research and writing, some tasks that actually come from your normal work experience. And 5, yes, I