Fable 5 (actually becoming available), GPT-5.6-Sol (+Terra/Luna), Grok 4.5, Muse Spark 1.1, Muse Image, GPT-Live-1, SWE 1.7 actually here and talk of Opus 5, Fable 5.1, GPT-6 on the horizon.
This month is BONKERS for models.
@maggie_hott@OpenAI Kids. The great joys of our lives and bringers of destruction to recommendation algorithms.
NLW, welcome back to Spotify. Have you heard the latest Crazy Frog?
NLW, welcome back to Kindle. Great news! There's a new Sammy Bird book out!
Jon crisply articulates the market part of the thesis that underlies the episode I did last week "Why Only AI Training Can Save the Economy" https://t.co/d80T8NDssE
The entire AI buildout — and the narrative around perceived shortages — hinges on how quickly Anthropic and OpenAI can grow. Their contracts alone account for more than half of the roughly $2 trillion in hyperscaler revenue backlogs. Any change in revenue scaling (+ or -) could completely change the prevailing narrative.
Unlike advertising, reselling compute is highly capital-intensive and must be built well ahead of demand. Pre-committing and building capacity for hundreds of billions in combined Anthropic and OpenAI revenue is a massive bet on continued rapid scaling from both companies. Misforecasting that demand could prove extremely value-destructive, given the core characteristics of the hyperscaler business: fixed capacity built well in advance of variable demand, paired with long-term deflationary pricing pressure.
Neoclouds represent viable emerging competition, Nvidia retains significant leverage (the opposite dynamic from the later x86 era), and both Anthropic and OpenAI have strong incentives to drive token cost deflation.
This is occurring against a backdrop of elevated costs. Price per token is not declining as quickly as many expected, most paid tiers have shifted to token-based pricing, and tight supply in high-bandwidth memory continues to pressure margins. The regulatory and political environment remains uncertain. AI faces growing public skepticism, and the current administration has shown caution toward concentrated market power in frontier AI — particularly with Anthropic, given differing political alignments on safety, energy constraints, and related issues. Why should any single company effectively control the development of digital god at this scale?
I’m reminded of a question an old mentor used to ask: “Are you certain nothing will go wrong over the next 3–5 years?” As@GavinSBaker wisely puts it, the three most important words in investing are “I don’t know.” Humility remains essential — in both directions.
STOP HOLDING BACK WHEN PROMPTING you can literally one shot whatever feature in one prompt just yap for longer. aim to describe every thing you can possibly imagine in ONE prompt
and obviously use voice. i often talk for 15minutes straight