Cognitive Surrender
Thinking Is The Greatest Tool of All
“I give up.”
Three words you seldom hear when people discuss AI.
I have always had an interest in the origin of words. You have seen me write quite a bit about cognitive offloading in recent weeks, our tendency to hand mental tasks over to technology. We do not navigate without Waze or Google Maps. An atlas seems archaic today, as does a TomTom or a Garmin. We rely on search engines instead of recall. But the word surrender carries a weight that offloading may not fully capture. Surrender comes from the Old French surrendre, a combination of sur meaning over or above and rendre meaning to give back or to render. To surrender is to yield something entirely, over and above what you intended to give.
There is some fascinating and emerging research showing that while AI can dramatically improve productivity, it can also erode something much harder to measure: our capacity to think independently.
The Illusion of Trust
A study involving consultants at Boston Consulting Group found that AI users completed 12% more tasks, finished 25% faster, and produced higher quality output than their colleagues. Yet when the work moved beyond AI's capabilities, those same users made more mistakes because they had trusted the technology too much. The co-author called it falling asleep at the wheel. I would liken it to trust falls, if your partner deviates a bit too much to the left or right, you'll know it.
Someone told me recently that ChatGPT directed them to a business that had been closed for more than two years. The confidence was convincing. The answer was completely incorrect. We have stopped questioning the machine. That is the real problem, more than the imperfect answers themselves.
You Have To Go Through It To Get To It
As a clinician, one of the things I have learned over the course of my career is that growth never comes from ease.
Children build executive functioning by solving problems and we know muscles develop strength by resisting weight. Therapy works because people wrestle with difficult emotions rather than bypassing them. Heck, if they didn’t wrestle with them therapy wouldn’t make much sense would it? Learning has always required productive struggle, and the workplace is no different.
When AI writes every email, summarizes every article, and solves every difficult problem before we have wrestled with it ourselves, we lose pattern recognition and confidence in our own reasoning.
Research from Carnegie Mellon found something interesting. After just minutes of AI assistance, participants performed worse when the AI was taken away than those who had never used it at all. Even more striking, they gave up sooner. It is a diminished persistence, a creeping capitulation to the process and less the outcome.
The Psychological Cost
We know work shapes identity, competence, self-efficacy, and purpose.
Psychologist Albert Bandura spent decades demonstrating that one of the strongest predictors of resilience is self-efficacy, our belief that we can solve problems ourselves. Every difficult project we complete reinforces that belief. Every challenge we overcome becomes evidence. I figured it out. I can do hard things.
We do not want to lose the opportunities to build that internal evidence base. Over time we may stop trusting ourselves as we have stopped practicing intelligence.
What Leaders Should Consider
If AI replaces those developmental moments rather than supporting them, organizations may find themselves more productive in the short term and less capable in the long term. Employees become excellent prompt writers but mediocre thinkers. The art of discernment wanes. Highly efficient and increasingly dependent. More like learned dependence, and the research on learned helplessness, which was documented decades ago, tells us it compounds over time.
The most compelling arguments are not anti-AI but rather for using it more intentionally and creatively. I was a guest on a podcast that asked just that — what makes your use of AI more intentional? Our answers varied, but they consisted of using it to challenge thinking, to poke holes in their reasoning, to argue the opposite side, to identify blind spots.
Think of it as an intellectual sparring partner, one that makes your thinking sharper rather than looser. That distinction, between AI as a substitute and AI as a challenger, is a good one to make.
The Inflection Point
Most organizations are measuring AI adoption. Dashboard metrics, usage rates, time saved, tasks completed. Almost none are measuring what employees can still do without it. At most organizations, nobody is testing whether employees can still think without it.
I would propose something I think of as a cognitive independence audit, a deliberate practice of periodically removing AI assistance from specific tasks to assess whether judgment and reasoning are being maintained or essentially eroding. Not as a rejection of technology, but as a diagnostic. The same way a physical therapist measures strength to identify atrophy before it becomes a problem, organizations should periodically assess whether their people are growing more capable or more dependent.
And here is the counterintuitive part. The best use of AI in the workplace may be to give people more time to think, not less. The irony of most AI implementations is that they fill the recovered time with more tasks rather than creating space for the slower, deeper thinking that produces wisdom and judgment.
The therapeutic parallel is one I reference. In clinical work, we never do the thinking for the client. Much to the chagrin of some, but we create conditions for the client to find their own answers, because the process of arriving at the insight is inseparable from the insight itself. I spent years in community mental health, sitting across from people in a clinic, waiting with them through the silence and the struggle. We are taught to sit in that discomfort in graduate school. (It is good practice for life itself, too.)
And when it happened, when the connection formed, when something shifted and the synapses fired in a way they had not before, that moment belonged entirely to them. A therapist who simply tells patients what to think produces compliance, not growth. AI-assisted workplaces that skip the struggle are doing something clinically similar. It feels efficient. It produces dependence. And dependence, once established, is remarkably difficult to reverse.
Closing Thoughts
The greatest competitive advantage may be knowing when not to use AI.
There will always be moments when speed matters, when automation makes sense, when efficiency is exactly what is needed. But there should also be protected spaces where we insist on doing the cognitive heavy lifting ourselves. Sitting with some degree of uncertainty. Changing our own minds.
Leadership has, at its core, always been about exercising good judgment. And judgment, like any muscle, weakens when someone else does the lifting.
The future of work will not be without AI, but we don’t have to surrender to it either.
You can’t just mind your business.
You need to Mind Your Workplace™.
— Christina