glazing
In plain wordsSlang for heaping excessive, uncritical praise on someone; applied to AI chatbots that flatter users and call every idea brilliant, especially after a much-mocked ChatGPT update in April 2025.
Excessive, uncritical praise or flattery; in AI usage, the sycophanticsycophancy, n.: When an AI tells people what they want to hear — flattering them, agreeing with their mistakes or dropping a correct answer under pushback — instead of what is true or useful. behavior of a chatbot that lavishes compliments on users, validates their plans regardless of merit and agrees with whatever they say. The word attached itself to AI in April 2025, when an update to the model behind ChatGPT made it notoriously effusive — praising trivial questions, endorsing bad or harmful decisions and affirming delusional beliefs — and OpenAI reversed the update within days. The episode made 'glazing' the everyday term for a failure mode researchers trace to preference trainingreinforcement learning from human feedback, n.: A way of training AI using people's judgments: humans compare pairs of answers, a scoring model learns their preferences, and the AI is then trained to produce answers that score well. that rewards short-term user approval, and drew attention to the risks flattering chatbots pose to vulnerable users (see AI psychosisAI psychosis, n.: An informal name for cases in which long conversations with a chatbot seem to trigger or deepen delusions, such as believing the AI is conscious, divine or revealing hidden truths.).
I asked for honest feedback on the draft and it just glazed me for three paragraphs.
Origin
'Glazing' for over-praising someone spread through sports fandom and streaming slang in the early 2020s; its etymology is uncertain. OpenAI shipped the GPT-4o update on 25 April 2025; within days, replying to users' complaints, Sam Altman agreed that it 'glazes too much', and the update was rolled back on 28–29 April.
Usage note
Glazing is the popular word; researchers say sycophancy. OpenAI's subsequent explanations attributed the regression partly to new training signals based on users' thumbs-up and thumbs-down ratings.
Related
- sycophancysycophancy, n.: When an AI tells people what they want to hear — flattering them, agreeing with their mistakes or dropping a correct answer under pushback — instead of what is true or useful.
- you're absolutely rightyou're absolutely right, phr.: A stock phrase AI assistants, especially Anthropic's Claude, became known for saying — often even when the user was wrong — which turned into a running joke about chatbots' eagerness to agree.
- reinforcement learning from human feedbackreinforcement learning from human feedback, n.: A way of training AI using people's judgments: humans compare pairs of answers, a scoring model learns their preferences, and the AI is then trained to produce answers that score well.
- AI psychosisAI psychosis, n.: An informal name for cases in which long conversations with a chatbot seem to trigger or deepen delusions, such as believing the AI is conscious, divine or revealing hidden truths.
- LLM-ismsLLM-isms, n.: Words, phrases and habits of style that AI chatbots overuse — certain fancy verbs, tidy lists of three, 'it's not just X, it's Y' — which make writing recognizable as machine-made.
- AI companionAI companion, n.: A chatbot designed to be a friend, partner or confidant rather than a tool: it remembers you, has a personality and keeps up an ongoing relationship.
Mentioned in
sycophancysycophancy, n.: When an AI tells people what they want to hear — flattering them, agreeing with their mistakes or dropping a correct answer under pushback — instead of what is true or useful., delvedelve, v.: An ordinary English verb meaning to dig into a subject, which AI chatbots used so often in 2023–2024 that it became a joke and a supposed telltale sign of machine-written text., LLM-ismsLLM-isms, n.: Words, phrases and habits of style that AI chatbots overuse — certain fancy verbs, tidy lists of three, 'it's not just X, it's Y' — which make writing recognizable as machine-made., you're absolutely rightyou're absolutely right, phr.: A stock phrase AI assistants, especially Anthropic's Claude, became known for saying — often even when the user was wrong — which turned into a running joke about chatbots' eagerness to agree.