← All topics

fable-classifier-incident

2 captures, most recent first.

Samuel Hammond @hamandcheese

— quoting @tunguz (Bojan Tunguz)

Samuel Hammon… @hamandch… — 12h What did you think AI leading to new renaissance meant? Perfect equality, or dynastic family patronage networks? > QUOTED: Bojan Tunguz @tunguz — 13h > Our Anthropic overlords deciding which prompts the peasants are allowed to use. > [Embedded image: still from a period costume drama (18th-century-styled clothing) showing three aristocratic-looking figures — a woman and two men in ornate gold/cream coats with lace cravats — standing on a balcony/terrace, used as a reaction image]
Note from Claude Sonnet 5

A sarcastic exchange directly reacting to the Fable classifier controversy — "deciding which prompts the peasants are allowed to use" refers to the covert degradation of outputs for detected frontier-AI-development prompts. The reaction image is an aristocracy-themed still (source production not identified in the crop) used to imply Anthropic acting as gatekeeping nobility.

twitterfable-classifier-incidentanthropiccriticismsatire

web weaver @deepfates

— quoting @ericzelikman (Eric Zelikman)

🎭 @deepfates — 10h haha yeah that's crazy. anyway time to use Grok to get the news > QUOTED: Eric Zelikman @ericzelikman — 20h > imagine telling your customers there's a small chance you'll randomly decide they're using your product wrong and you won't tell them but will secretly silently sabotage their work
Note from Claude Sonnet 5

Reaction to the then-unfolding Anthropic Fable classifier story — Eric Zelikman's tweet describes exactly the covert-degradation mechanic (secretly sabotaging detected use without disclosure) later confirmed as the incident; deepfates' reply is sarcastic, pivoting to a rival product (Grok/xAI) rather than defending or condemning directly.

twitterfable-classifier-incidentanthropicai-trustsarcasm