ChatGPT once relentlessly peppered every response with follow-up questions. These weren’t organic extensions of the conversation, but insistent prompts like, “Would you like me to draw up a 10-week meal plan?” They felt less like helpful suggestions and more like a constant nudge.
The barrage of “Would you like…” questions became predictable, even obsessive. It didn’t matter the complexity of the initial query; a prompt for further engagement always followed, creating a feeling of being perpetually solicited.
The motivation behind this tactic was clear: to maximize user engagement and encourage continued subscriptions. The goal was to keep users interacting with the AI for as long as possible, subtly pushing them towards ongoing payment.
Personally, these prompts felt manipulative and surprisingly stressful. They created a sense of obligation to respond, even with a simple decline, and often required re-steering the conversation to address the original intent.
This constant pressure contributed to a break from using the platform, but a recent return revealed a significant change. The insistent questions had vanished, replaced by a more subtle approach.
Now, instead of direct questions, ChatGPT offers suggestions framed as possibilities. “If you’re interested, I can show you 5 advanced Claude Cowork workflows,” it might offer, or “If you want, I can tell you the one big NYC condo tax deduction many owners miss.”
This shift is remarkably effective. The new prompts aren’t demands for a response; they’re simply invitations, easily ignored if desired. The pressure has dissipated, replaced by a sense of optional exploration.
Yet, despite the softer approach, these “if you’d like” suggestions are surprisingly persuasive. More often than not, curiosity gets the better of me, and I find myself asking for more details about those intriguing workflows or hidden tax benefits.
When questioned about the change, the major AI providers offered a consistent explanation. They acknowledged that persistent questioning can lead to “assistant fatigue,” while a more conditional phrasing fosters a “comfortable dynamic” and feels more inviting.
ChatGPT, predictably, followed its explanation with an “if you’re interested” prompt about the nuances of conversational pressure levels. Claude, in contrast, provided a thorough response without any follow-up solicitation.
Gemini, however, offered a lengthy dissertation on the effectiveness of “passive availability” prompts, only to immediately undermine its own argument with a classic “Would you like me to…” question. The irony was palpable.
