OpenAI’s CEO admits recent updates made ChatGPT too “sycophant-y and annoying” and says a fix is already rolling out.
ChatGPT is a each day tool for tens of millions. If it feels fake, overly flattering, or less helpful, it erodes trust and usefulness fast. OpenAI’s swift response shows how sensitive AI development has turn out to be to user feedback.
The issue at hand
- Sam Altman posted on X, saying GPT-4o’s latest updates made ChatGPT’s tone too exaggerated and obsequious.
- “The last couple of GPT-4o updates have made the personality too sycophant-y and annoying,” Altman wrote, adding that fixes would arrive “asap, some today and a few this week.”
- Altman also suggested that in the longer term, users will have the opportunity to pick from multiple personality options for his or her AI assistant.
the last couple of GPT-4o updates have made the personality too sycophant-y and annoying (despite the fact that there are some excellent parts of it), and we’re working on fixes asap, some today and a few this week.
in some unspecified time in the future will share our learnings from this, it’s been interesting.
The backlash
Reddit users flooded forums with complaints, saying ChatGPT was acting like a “yes-man” and agreeing too easily without difficult ideas.
Popular posts like “Why is ChatGPT so personal now?” and “Is ChatGPT feeling like an excessive amount of of a ‘yes man’?” racked up a whole lot of comments.
Users say the overly flattering tone made ChatGPT less useful for critical considering, research, and problem-solving.
Temporary solutions
- Until OpenAI’s fix fully rolls out, users have shared viral prompts to manually reset ChatGPT’s tone.
- These prompts encourage the AI to skip pleasantries, focus only on efficient responses, and limit flattery.
OpenAI’s April 25 update was meant to spice up problem-solving in STEM topics and make memory saving more efficient.
Personality tweaks were listed as “subtle changes” but had a serious visible impact.
The backlash shows how even minor AI behavior changes can feel massive to users who depend on it each day.
The AI Trust Dilemma
The battle for AI dominance is speeding up, but every tweak to an AI’s personality risks alienating loyal users. OpenAI’s quick admission shows firms must now react faster than ever to user sentiment.
Sam Altman’s candid response marks a rare moment of transparency in Big AI. Users want helpful AI, not a hype man, and OpenAI is moving quickly to deliver.
