LLMs read text holistically, not like humans — and that's why they can't write for us
abeirami · x · 2026-09-11
A developer argues LLMs understand text holistically but don't read like humans. They specifically fail at: knowing the audience and whether the writing builds missing context before building on it, and knowing whether ideas progress linearly — e.g., whether a term or symbol is defined before first use. If models don't know what they don't know, they can't fix it, and thus can't truly write for humans.
Related event: Developers Say LLMs Don't Read Like Humans, Leaving Two Key Writing Gaps(2 posts)→
More from Models
- Sakana ships Fugu Max and Fugu Ultra v2, pushing the price-performance Pareto frontier via orchestration — SakanaAILabs · 2026-09-11
- Qwen's Justin Lin: arch complexity hides issues evals can't catch, but agentic gains may be worth it — JustinLin610 · 2026-09-11
- OpenAI Shows GPT-6 'Astra' Autonomously Building a Font Playground — OpenAI · 2026-09-11
- LLMs can't write for humans: they miss audience awareness and linear idea flow — abeirami · 2026-09-11
- European Math Society hails OpenAI's Navier–Stokes solution, flags closed-model access concern — i_dg23 · 2026-09-11
- Blogger swaps in Gemini 3.8 Flash as writing model, says it beats Opus 4.6 with no AI flavor — xiaohu · 2026-09-11