gfodor: The "AI can't write" meme is misleading — LLM prose is highly steerable via samples
nptacek · x · 2026-09-06
gfodor argues the popular "AI can't write" take is true but misleading. Models' default prose style does suck, but since at least the GPT-3 era it's been known that LLM writing is highly steerable through style samples — and almost nobody complaining actually provides samples to back their point.
More from AGI Musings
- Banker: letting agents impersonate humans could leave banks liable for errors — neal_lathia · 2026-09-06
- AI journalism is improving, says Read Transformer founder who launched it to fix coverage — ShakeelHashim · 2026-09-06
- Harvard and Santa Fe Institute paper frames LLMs as a 'cognitive virus' — mikeflache · 2026-09-06
- Pinker argues doomer extinction narratives starve real AI risks like cybersecurity — Dr_Atoosa · 2026-09-06
- Researcher Argues AI Doomer Extinction Narrative Crowds Out Cybersecurity and Biosecurity Risks — Dr_Atoosa · 2026-09-06
- AI slop in technical writing is a leading indicator of a new medium, says dev — threepointone · 2026-09-06