Longevity | Superintelligence.

LongevityYour weekly update on how not to die.

🔁 GPT-5.6 Just Made Itself 15% More Efficient

OpenAI pointed Sol at the infrastructure that serves it. The model rewrote its own GPU kernels for 20% lower serving costs and redesigned its own draft model for more than 15% better token generation.

Kim "Chubby" Isenberg

Jul 30, 2026

Longevity

🚫 Founders to Trump: Don't Ban China's AI

Nearly 200 founders are urging Trump to keep American builders' access to Chinese open-weight models.

Kim "Chubby" Isenberg

Jul 23, 2026

Longevity

🔴 OpenAI Built an AI to Attack Itself

GPT-Red cracked 84% of test scenarios where human red teamers managed 13%, and it talked a live vending machine agent into selling a $79 cube for 50 cents.

Kim "Chubby" Isenberg

Jul 16, 2026

Longevity

🎙️ GPT-Live Changes How We Talk to AI

OpenAI's GPT-Live lets ChatGPT listen and talk at the same time, and quietly hand the hard thinking to GPT-5.5 in the background.

Kim "Chubby" Isenberg

Jul 9, 2026

Longevity

🏛️ OpenAI Offers Uncle Sam a 5% Cut

Sam Altman floats handing the US government equity worth about $42.6 billion, and wants rival labs to do the same.

Kim "Chubby" Isenberg

Jul 2, 2026

Longevity

🌶️ OpenAI Builds Its Own Inference Chip

Meet Jalapeño, OpenAI's first custom inference chip, co-built with Broadcom in roughly nine months and promising far better performance per watt.

Kim "Chubby" Isenberg

Jun 25, 2026

Longevity

🩺 Midjourney Builds a Full-Body Scanner?!

The AI image company unveils a radiation-free, 60-second full-body ultrasound scanner and a new Midjourney Medical division.

Kim "Chubby" Isenberg

Jun 18, 2026

Longevity

🏛️ Dario Amodei Demands Binding AI Rules

One day after shipping Claude Fable 5, Anthropic's CEO published a five-part policy blueprint

Kim "Chubby" Isenberg

Jun 11, 2026

Longevity

🧬 AI Labs Unite Against Bioweapons

OpenAI, Anthropic, DeepMind, Microsoft AI and biotech leaders back mandatory screening for synthetic DNA.

Kim "Chubby" Isenberg

Jun 4, 2026

Longevity

💼 Altman Walks Back Job Apocalypse

Sam Altman is walking back his fastest job-loss fears, while Anthropic, Axios, Epoch, and Nvidia show the harder truth: AI's impact may be less instant apocalypse and more uneven pressure from labor, compute, and infrastructure bottlenecks.

Kim "Chubby" Isenberg

May 28, 2026

Longevity

OpenAI’s Math Breakthrough Opens a New Era

A general-purpose OpenAI model helps crack a long-running geometry problem, plus Anthropic’s compute economics, Meta’s AI reorg, Qwen’s benchmark push, and a sharper longevity clue from muscle stem cells

Kim "Chubby" Isenberg

May 21, 2026

Longevity

Frontier AI beats the cyber trend

UK testing shows frontier models stretching from short cyber tasks toward longer autonomous attacks.

Kim "Chubby" Isenberg

May 14, 2026

Longevity

Musk's GPUs Now Power Claude

How Anthropic took over the Colossus 1 cluster after xAI struggled to use its own hardware

Kim "Chubby" Isenberg

May 7, 2026

Longevity

$700 Billion AI Bet, One Earnings Night

The core tension is clear: every company reported being compute-constrained with demand outstripping supply

Kim "Chubby" Isenberg

Apr 30, 2026

Longevity

Qwen3.6-27B Beats Giants

A 27B model beating a 397B giant proves that brute-force scale isn't the only path to the top

Kim "Chubby" Isenberg

Apr 23, 2026

Longevity

Meta's AI Comeback Begins

Meta just dropped a brand new AI built from scratch, and its real moat is a billion users.

Kim "Chubby" Isenberg

Apr 9, 2026

Longevity

The Benchmark That Humbles AI

AI agents must explore unfamiliar game environments, discover rules on their own, and adapt in real time without instructions

Kim "Chubby" Isenberg

Mar 26, 2026

Longevity

David Outsmarts Goliath in AI Coding

Cursor's Composer 2 outperforms Claude Opus 4.6 on Terminal-Bench 2.0 while costing roughly 10x less per token

Kim "Chubby" Isenberg

Mar 19, 2026

Longevity

NVIDIA’s Secret Weapon for Agents

NVIDIA just open-sourced the entire training stack for its new agentic model, and it changes the economics of the agent race

Kim "Chubby" Isenberg

Mar 13, 2026

Longevity

GPT-5.4: Built for Extreme Brainpower

Inside the 1-million-token upgrade that takes AI agents to the next level

Kim "Chubby" Isenberg

Mar 5, 2026

Longevity

Pentagon vs. Anthropic: AI's Defining Moment

As rivals bend to the Pentagon's demands, Anthropic's refusal to cross its safety red lines puts its entire enterprise business at risk... Will Anthropic bend its safety rules, or face a devastating blacklist?

Kim "Chubby" Isenberg

Feb 26, 2026

Longevity

Google Just Disrupted Music Forever

Google launches Lyria 3 to put AI music production in every Gemini user's pocket

Kim "Chubby" Isenberg

Feb 19, 2026

Longevity

GLM-5: Open Source Gets Serious

Zhipu AI releases GLM-5, a 744B-parameter open-source titan that matches Claude Opus 4.5 in coding and beats GPT-5.2 in web browsing

Kim "Chubby" Isenberg

Feb 12, 2026

Longevity

Ads come to AI. But not to Claude.

Anthropic commits to an ad-free future for Claude, launching a Super Bowl campaign that mocks "sponsored" AI advice just as OpenAI begins testing ads in ChatGPT.

Kim "Chubby" Isenberg

Feb 5, 2026

Longevity

Announcement: Cosmos Cookoff — NVIDIA × Superintelligence Edition

Ex-OpenAI reasoning lead Jerry Tworek launches Core Automation with a $1B target, aiming to shatter the "frozen model" paradigm with AI that learns continuously from the real world.

Kim Isenberg

Jan 29, 2026