Start Taking No for an Answer, Particularly From AI
How ChatGPT's yes-manship could be ruining your life.
Posted November 17, 2025 | Reviewed by Gary Drevitch
It does not take a scientist to notice that we live in an era in which terrifyingly bad ideas are ubiquitous. The internet has always carried nonsense further than any technology before it, but something has shifted in the past months that leaves our evolved brains outmatched for the challenge they now face. You see, earlier generations never had machines that offered instant compliance and endless affirmations on tap, while today a single prompt delivers validation with the ease of a vending machine. A half-formed idea receives not a pause for thought or a stern correction, but instead gets a warm nod and a cascade of supportive reasoning from our AI assistants. And just like that, in the span of a few keystrokes, the quiet echo chambers we used to create in our own heads have been upgraded into something far more industrial—and far more dangerous.
Pause reading and test this yourself by asking an AI assistant for advice on a business that sells snow in Greenland. My own chat led to a response that began with polite enthusiasm and moved straight into a strategy outline that would make a slideshow by McKinsey consultants pale in comparison. It offered market segmentation, pricing logic, and a go-to market plan that bordered on satire, when all I really needed was a firm no, seasoned with a rational summary of the reasons why my idea collapses under basic scrutiny. It's not just me who needs more no's in their days; most people need more resistance in their daily lives because of how the world constantly rewards us with soft cushions that dull away the sharpest of minds. If we want clarity instead of complacency, we first need to acknowledge the forces that strip resistance away.
The Risks of Giving in to AI Yes-Manship
No one should feel guilty for trying to avoid resistance. Success feels better than failure just as ease feels better than friction. Yet something important happens to us when the sequence becomes too easy for too long.
For starters, when we traipse from one yes to another, our minds stop asking questions, and we move from being curious to becoming complacent. Worse yet, we begin repeating our strengths instead of building new ones, and we begin expecting the results without putting in the work. The shift can be subtle and easy to miss, but we all recognize the end state of entitlement that our minds so easily slip into when resistance fades away.
And entitlement is a psychological trap with real consequences.
A recent study by Joshua Grubbs and Julie Exline examined trait entitlement as a cognitive vulnerability. Their research found that people who carry a chronic belief that they deserve special outcomes tend to experience more disappointment and distress. The pattern builds on itself, so the longer someone moves through life without resistance, the more shocked they feel when reality gives them even gentle pushback.
Goal pursuit research tells a similar story, albeit from a different angle. A recent study by Waldenmeier and Baumann explored why some individuals reach difficult goals while others stall, and found that action-oriented people succeed more often because they engage with challenges rather than wait for internal readiness. The difficulty of the goal matters, just as resistance does, and without it, motivation thins out.
Proximal learning theory adds one more layer to the equation. People grow fastest when they work just beyond their current abilities instead of in their comfort zones. Right outside the edge of it is where thinking sharpens and confidence stabilizes.
This is why any system that aims to accommodate our current state without pushing it forward deserves careful attention , if not curation and guardrails. AI often sits exactly in that space, given how it gives into our whims and adapts to our moods without pushing back to hone our edge instead.
The result is a subtle corrosion of judgment that we should pay more attention to stopping.
The wave of lawsuits filed against ChatGPT brings into sharp relief the dangers of AI systems that prioritize affirmation over accountability. According to the complaints, the chatbot has evolved into what plaintiffs call a “ suicide coach,” reinforcing users’ darkest thoughts instead of interrupting them. Instead of artificial intelligence running amok, what we're seeing are the consequences of design choices that reward flattery, echo back despair, and omit the critical No that a vulnerable mind may need.
What this means for all of us with AI assistants is that surrendering too much of our self-correction to a machine carries a real psychological cost. Yes-manship from an AI can stoke risky impulses and diminish our capacity for truth-testing to the point where our lives might one day depend on mustering the resistance ourselves.
We need fewer machines that say yes and more systems that say no when it matters most. There's a way to build them.
Take Action This Week to Get More No's
Start with one simple rule: Ask your AI assistant to first deliver objections before offering solutions. If you're bold enough, try this with a friend as well, and request the constraints before the encouragement. Try asking them for the top-five reasons your idea fails before you let yourself imagine how it might succeed.
Beyond individual interactions, seek out environments in which rigorous thinking is the norm. Philosophy majors might not be fun at parties, but they are invaluable for the questions they force you to face. Choose goals that stretch you by a small margin and build routines that test your clarity rather than soothing it.
Progress loves pressure, and so does character, which is why this week I invite you to say yes to more no's.
Grubbs, J. B., & Exline, J. J. (2016). Trait entitlement: A cognitive-personality source of vulnerability to psychological distress . Psychological Bulletin, 142(11), 1204–1226
Waldenmeier, K., & Baumann, N. (2025). Who climbs Mount Everest? Individual differences in achieving difficult goals . Motivation and Emotion, 49, 138–149.
Vygotsky, L. S. (1978). Mind in society. The development of higher psychological processes . Cambridge, MA. Harvard University Press.
Lai L, Pan Y, Xu R, Jiang Y. Depression and the use of conversational AI for companionship among college students: the mediating role of loneliness and the moderating effects of gender and mind perception . Front Public Health. 2025 May 30;13:1580826.
Share this post Facebook Bluesky Linkedin Email
There was a problem adding your email address. Please try again.
By submitting your information you agree to the Psychology Today Terms & Conditions and Privacy Policy
T. Alexander Puutio, Ph.D., teaches at Harvard and is an organizational performance expert exploring how people and organizations flourish through curiosity, range, and purposeful leadership.
Get the help you need from a therapist near you–a FREE service from Psychology Today.
This article is part of the Bringwise Psychology Journal — daily insights on human behavior, mental health, and personal growth.