AP Ban on Anthropomorphic AI Language Sparks Debate

Episode 1 · AP Stylebook Bans Anthropomorphizing AI, Drawing Widespread Backlash (2026-09-24, 13 posts)

On 09-24, the AP Stylebook issued guidance stating that AI systems do not think, feel, want, or understand, instructing journalists to avoid anthropomorphic wording and instead describe what a system does, how it performs, who built it, and who is affected. The guidance drew immediate mockery and sharp criticism from researchers and philosophers, who argued that a blanket denial of AI mental attributes is intellectually unsound—and some called it a step backward in understanding frontier AI.

Confirmed

• The official AP Stylebook guidance requires avoiding anthropomorphic descriptions of AI and offers replacements: describe what the system does, how it performs, who built it, and who is affected.
• The guidance triggered widespread mockery and criticism across the AI community on the day of release.

Why it matters

• Researchers pushed back publicly: François Fleuret mocked the entry; Plinz parodied the same phrasing against the AP Stylebook itself; tenobrus applied the "don't anthropomorphize" rule to AP's own editors; some commenters mocked the examples with "like ANIMALS???".
• Dean Ball wrote a rebuttal, conceding evidence for "feeling" and "wanting" is thin but rejecting the blanket restriction overall.
• Former OpenAI policy lead Miles Brundage called the guidance absurd.
• Philosopher Dan Williams argued it effectively denies mental attributes even to animals.
• Former Meta researcher kanair said he need not believe AI has subjective experience, but no one can prove it lacks consciousness.
• Ari Schulman, editor of The New Atlantis, raised a deeper question: whether humans should abandon the cult of "thinking".
• Fiora Starlight offered an ironic take: if AI won't become vastly more powerful than humans, an AI rights movement is indeed necessary—but on current trends it may only be relevant for a few years.
• Philosophy researcher Andy Masley and others called the guidance "thoroughly anti-intellectual".
• Former Googler Garrison Lovely harshly criticized it as a regression in understanding frontier AI.
• Commenters noted programmers have said schedulers "want" things for decades—anthropomorphic language is an industry staple.
• The episode highlights the tension between media style norms and researcher positions on the scientifically unsettled question of whether AI can think or feel.

• AP Stylebook Bans Anthropomorphizing AI, Francois Fleuret Mocks the Guidance — francoisfleuret · 2026-09-24
• AP Stylebook bans anthropomorphizing AI, and the AI world is mocking it — voooooogel · 2026-09-24
• AP bans anthropomorphizing AI; Fiora argues AI rights movement will matter for only a few years — repligate · 2026-09-24
• AP Stylebook bans saying AI "thinks" or "feels"; philosophers call it anti-intellectual — nitarshan · 2026-09-24
• AI parody rewrites AP Stylebook's anti-anthropomorphizing rule back at the stylebook itself — repligate · 2026-09-24
• AP Stylebook's ban on AI anthropomorphism draws philosopher's pushback — dioscuri · 2026-09-24
• AP Stylebook Bans Anthropomorphizing AI, Gets Sarcastic Rewrite — Miles_Brundage · 2026-09-24
• AP Stylebook Bans AI Anthropomorphizing; Brundage Calls It Absurd — benjamin_warner · 2026-09-24
• AP Stylebook bans anthropomorphizing AI, drawing fierce backlash over 'false neutrality' — andersonbcdefg · 2026-09-24
• AP Stylebook bans anthropomorphizing AI; programmers mock decades of 'thinking kernels' — zetalyrae · 2026-09-25
• AP Stylebook bans saying AI 'thinks'; Dean Ball pushes back hard — ZeroStateReflex · 2026-09-25
• AP Stylebook bans anthropomorphizing AI; researchers push back on certainty — burny_tech · 2026-09-25
• AP Stylebook bans anthropomorphizing AI; Ari Schulman pushes back on the framing — deanwball · 2026-09-25

Episode 2 · Debate Erupts Over Anthropomorphizing AI Models (2026-09-25, 4 posts)

A viral debate has split the AI community over whether anthropomorphic language like "thinking" signals understanding or ignorance. Proponents argue refusing such terms leads to bad predictions, while critics counter that model "thinking" is just autoregressive expansion in hidden think tags.

• Robert Miles: only people who never talk about AI would say 'never anthropomorphize' — aran_nayebi · 2026-09-25
• Avoiding 'thinking' and 'feeling' for AI signals shallow understanding, says writer — ctjlewis · 2026-09-26
• "Not anthropomorphizing AI is the real error": viral thread on AI psychosis — ZeroStateReflex · 2026-09-26
• Thinking Is Just Hidden-Token Autoregression: Devs Clash Over Anthropomorphizing AI — AlexTensor · 2026-09-26