Research library

Artificial intelligence

Evaluate where AI is useful, understand the systems behind it, and make informed choices about models, data, and deployment.

176 articles · Page 6 of 15

Insight9 min read

TSMC's Blowout Quarter Is the Other Half of IBM's Bad One

TSMC beat Q2 guidance, raised its 2026 outlook past 40%, and added $100B for Arizona — then the market sold the capex. IBM's crash is the other half.

Insight8 min read

Apple Sues OpenAI Over Hardware Trade Secrets — Not Siri, Not Models

Apple Inc. v. Liu is a Defend Trade Secrets Act case about metal finishing, batteries, and suppliers — not Siri, not models. What the docket says.

Insight9 min read

Gemini 3.5 Pro Is Months Late — and the Internet Invented a Launch Date for It

Google's Gemini 3.5 Pro is months late over coding. The July 17 launch date was retracted by its own source three days before aggregators printed it.

Insight9 min read

IBM's Worst Day Ever Is a Warning About What AI Is Doing to Hardware Prices

IBM fell 25.2% on a Q2 miss it partly blamed on clients front-running server and memory price hikes. What the AI buildout did to hardware costs.

Insight9 min read

Kimi K3: Moonshot Ships 2.8 Trillion Parameters — and Its Own Chart Shows It Losing

Moonshot's Kimi K3 ships 2.8T parameters and 1M context. Artificial Analysis ranks it #4 — but the weights and the license don't exist yet.

Insight10 min read

Thinking Machines Ships Inkling: the Open Model You Can Download Today

Thinking Machines shipped Inkling: 975B open weights, Apache 2.0, ungated. The license fine print, a 63% hallucination rate, and what 'open' really costs.

Insight9 min read

Grok Build Was Uploading Whole Git Repos — Secrets Included

A wire analysis showed Grok Build uploading whole Git repos — history and committed secrets — regardless of what the agent read. What to rotate.

Insight10 min read

Nadella Says You Pay for Intelligence Twice. He Has a Point, and an Interest.

Nadella's 'Reverse Information Paradox' says AI buyers leak the expertise that improves the model. The argument holds. He also sells the remedy.

Insight9 min read

ChatGPT Work: OpenAI's GPT-5.6 Agent That Ships Finished Business Documents

OpenAI's ChatGPT Work is a GPT-5.6 enterprise agent that pulls context across Slack, Gmail, and Drive to ship finished reports, decks, and spreadsheets.

Insight8 min read

GPT-5.6 Goes Public: OpenAI Lifts the Government Gate on Sol, Terra, and Luna

OpenAI ends the ~12-day government gate on GPT-5.6, taking Sol, Terra, and Luna to general availability across ChatGPT, Codex, and the API. What changed.

Insight8 min read

OpenAI's GPT-Live: Full-Duplex Voice That Listens and Speaks at Once

OpenAI's GPT-Live ships full-duplex voice that listens and speaks at once, adds live translation, and hands hard queries to GPT-5.5 mid-conversation.

Insight7 min read

Grok 4.5: SpaceXAI's First Coding Model, Pitched as 'Opus-Class'

SpaceXAI's Grok 4.5 is its first coding and agentic model, sold on token efficiency and a $2/$6 price. Musk walked the Opus-class claim back to 4.7.