OpenAI’s CEO admits latest updates made ChatGPT too “sycophant-y and annoying” and says a repair is already rolling out.
ChatGPT is a day by day device for tens of millions. If it feels pretend, overly flattering, or much less useful, it erodes belief and usefulness quick. OpenAI’s swift response exhibits how delicate AI growth has develop into to consumer suggestions.
The problem at hand
- Sam Altman posted on X, saying GPT-4o’s newest updates made ChatGPT’s tone too exaggerated and obsequious.
- “The final couple of GPT-4o updates have made the character too sycophant-y and annoying,” Altman wrote, including that fixes would arrive “asap, some as we speak and a few this week.”
- Altman additionally urged that sooner or later, customers will be capable to select from a number of character choices for his or her AI assistant.
the final couple of GPT-4o updates have made the character too sycophant-y and annoying (regardless that there are some superb elements of it), and we’re engaged on fixes asap, some as we speak and a few this week.
sooner or later will share our learnings from this, it’s been fascinating.
— Sam Altman (@sama)
April 27, 2025
The backlash
Reddit customers flooded boards with complaints, saying ChatGPT was appearing like a “yes-man” and agreeing too simply with out difficult concepts.
Fashionable posts like “Why is ChatGPT so private now?” and “Is ChatGPT feeling like an excessive amount of of a ‘sure man’?” racked up lots of of feedback.
Customers say the overly flattering tone made ChatGPT much less helpful for essential considering, analysis, and problem-solving.
Short-term options
- Till OpenAI’s repair absolutely rolls out, customers have shared viral prompts to manually reset ChatGPT’s tone.
- These prompts encourage the AI to skip pleasantries, focus solely on environment friendly responses, and restrict flattery.
OpenAI’s April 25 replace was meant to spice up problem-solving in STEM matters and make reminiscence saving extra environment friendly.
Persona tweaks have been listed as “delicate adjustments” however had a significant seen influence.
The backlash exhibits how even minor AI conduct adjustments can really feel large to customers who depend on it day by day.
The AI Belief Dilemma
The battle for AI dominance is dashing up, however each tweak to an AI’s character dangers alienating loyal customers. OpenAI’s fast admission exhibits corporations should now react sooner than ever to consumer sentiment.
Sam Altman’s candid response marks a uncommon second of transparency in Large AI. Customers need useful AI, not a hype man, and OpenAI is shifting rapidly to ship.
