Results
8 articlesSort
31:58EN
Theo - t3․gg· Aug 14, 2026Claude’s AI Watermarks and the Limits of Detecting Generated ContentTheo Browne · t3.ggTheo Browne examines Anthropic’s marking plans, the challenge of text watermarks, and what signed provenance can and cannot establish.
44:29ENTheo - t3․gg· Jul 25, 2026Theo’s Opus 5 Review: Coding Style, Cost and Model TradeoffsTheo Browne · Ping LabsAfter a day of coding with Opus 5, Theo compares it with Fable and GPT-5.6 Sol across plans, benchmarks, costs and real work.
32:53ENTheo - t3․gg· Jun 11, 2026Theo’s Fable 5 Review: Benchmarks, Real Projects, and LimitsTheoTheo tests a newly released model on coding and creative work, highlighting impressive results alongside failures, safeguards and cost.
35:13ENTheo - t3․gg· May 7, 2026Theo on Anthropic’s SpaceX Compute Deal and Claude Usage LimitsTheo Browne, Dario Amodei · Ping LabsThe May 2026 partnership expands capacity for Claude while raising questions about AI infrastructure, training data and competition.
34:39ENTheo - t3․gg· May 1, 2026Claude Code’s Billing False Positive and Anthropic’s Developer Trust ProblemTheo Browne · T3Theo Browne examines a subscription billing false positive, explains prompt caching, and shares T3 Chat’s API costs.
44:14ENTheo - t3․gg· Apr 20, 2026Did Claude Get Worse? Tracing Regressions Across the AI Coding StackTheo Browne · T3Theo Browne examines user reports, benchmarks, agent harnesses, tokenization and past infrastructure bugs to understand changing Claude Code performance.
37:59ENTheo - t3․gg· Apr 17, 2026Testing Claude Opus 4.7: Strong Capabilities, Uneven Coding ResultsTheo Browne · T3Theo Browne tests Anthropic’s new model on a cryptography puzzle, a project migration and a repository-cloning script.
30:07ENTheo - t3․gg· Apr 16, 2026Testing Claude Code Desktop: Promising Features and Interface FrictionTheo Browne · T3Theo Browne explores Anthropic’s desktop coding interface, from split views and worktrees to permissions, diffs and browser preview.