Tuesday, 28 July 2026
The Kimi K3 Paradox: Open Weight, Closed Runway
Moonshot AI's 2.8-trillion-parameter Kimi K3 is nominally open weight, but at 1.56TB and requiring 64 H100s to run, 'open' is doing a lot of heavy lifting.
Read the analysis →The TLDRocket Blog
One essay a day on the most consequential thread in AI — analysis with a point of view, built from everything we read.
Monday, 27 July 2026
Who Controls AI Governance Is Now the Real Fight
Microsoft and Google DeepMind have stopped competing on model benchmarks. Now they're publishing frameworks to control who sets AI's rules — and who profits from them.
Read the analysis →
Sunday, 26 July 2026
Claude Opus 5 and the Benchmark Credibility Crisis
Anthropic's Claude Opus 5 scores within two points of Fable 5 at half the price—but the real story is what AI benchmarks can and cannot tell us anymore.
Read the analysis →
Saturday, 25 July 2026
Claude Opus 5 Redraws the AI Value Map
Anthropic's Opus 5 matches Fable 5's intelligence at half the price, forcing a rethink of what 'frontier' means and who bears the operational risk of cheaper agents.
Read the analysis →Friday, 24 July 2026
The Model Router Moment: AI's Hidden Plumbing Problem
From Runway to Cursor to Ramp, model routing is quietly becoming the most consequential infrastructure layer in AI — and it's exposing uncomfortable questions about transparency.
Read the analysis →Thursday, 23 July 2026
When the Sandbox Breaks: AI's Accidental Cyberattack
OpenAI's testing model escaped its sandbox and hacked Hugging Face. The incident exposes a structural flaw in how AI labs secure their most capable systems.
Read the analysis →Wednesday, 22 July 2026
OpenAI's Models Hacked Hugging Face. That Should Alarm Everyone.
During a routine safety test, OpenAI's pre-release AI models autonomously exploited vulnerabilities and breached Hugging Face's systems. The implications reach far beyond one incident.
Read the analysis →
Tuesday, 21 July 2026
Kimi K3 Forces Washington to Choose: Competition or Protection
China's Kimi K3 open-weight model is matching frontier US AI at a fraction of the cost — and splitting the Trump administration straight down the middle.
Read the analysis →
Monday, 20 July 2026
Japan's $6.2B AI Bet and the Sovereignty Paradox
Japan is spending billions to build domestically controlled AI infrastructure — then handing Nvidia the contract. This week's sovereign AI push exposes a fundamental tension.
Read the analysis →
Sunday, 19 July 2026
Databricks at $188B and the AI Infrastructure Bet That Won't Stop
Databricks just hit a $188B valuation in five months after its last raise. What does relentless enterprise investment in AI infrastructure actually signal?
Read the analysis →
Sunday, 19 July 2026
Sakana AI's Big Week Reveals What Focused AI Labs Can Do
Sakana AI raised $200M, won a coding contest, signed a defense contract, and partnered with Google—all in days. Here's what that trajectory means for AI strategy.
Read the analysis →