A technology writer has resumed using ChatGPT following a period of exploring alternatives like Gemini and Claude, maintaining multiple AI subscriptions much like streaming services. Upon returning to the platform, the user observed a notable adjustment in the chatbot's behavior, specifically the elimination of a previously persistent feature.
In earlier interactions, ChatGPT frequently concluded responses with probing questions designed to encourage continued dialogue, such as inquiries about integrating Tailscale with a Raspberry Pi or creating a 10-week family meal plan.
While relevant follow-up questions can enhance conversations, ChatGPT's version often felt overly insistent, appending suggestions like 'Would you like me to...' to nearly every reply, a trait shared by many leading AI chatbots.
These prompts stem from AI companies' strategies to increase user engagement, prolonging interactions to justify subscription renewals.
The constant solicitations were perceived as intrusive, coercive, and anxiety-inducing, prompting unnecessary responses or additional input to steer discussions, such as requesting a vinaigrette recipe instead of a lengthy meal plan.
Though not the only factor in the user's temporary hiatus from ChatGPT, the feature contributed to the decision to pause usage. However, upon recent reconnection, the absence of these direct questions became immediately apparent.
Responses now incorporate optional suggestions phrased as 'If you'd like' statements, for instance, offering to demonstrate five advanced Claude Cowork workflows or revealing a overlooked NYC condo tax deduction.
This shift represents a modest yet significant improvement, replacing demanding queries with passive options that users can ignore without pressure.
Despite the reduced insistence, these new prompts prove effective, often prompting users to explore the additional information provided.
Inquiries to OpenAI regarding the updated prompt style remain unanswered, with plans to revise the article upon receiving clarification.
Surveying major AI platforms—ChatGPT, Gemini, and Claude—yielded similar insights: aggressive questions risk causing 'assistant fatigue,' whereas conditional phrasing fosters a more relaxed and appealing exchange.
ChatGPT's reply included an 'if you're interested' note on varying 'pressure levels' in prompts, Claude delivered a detailed, scholarly response without any trailing suggestion, and Gemini provided an extensive analysis on the benefits of 'passive availability' over active questions, only to end with a traditional 'Would you like me to...' query.
Ben has covered technology and consumer electronics for over two decades. Since 2014, he has contributed to PCWorld and joined TechHive in 2019, reporting on products ranging from smart speakers and soundbars to lighting systems and surveillance devices. His work has featured in outlets including PC Magazine, TIME, Wired, CNET, Men's Fitness, and Mobile Magazine. He possesses a master's degree in English literature.