Does AI Make Us Think Less — or Think Better?
Offloading mental work can free capacity, but unused muscles eventually change.
Author
Alok Jha
Reading
2 min read
Category
AI + Human
Calculators did not destroy mathematics. GPS did change how many of us navigate. Search engines did not remove memory, but they changed what we bother to remember. AI is the newest—and much broader—tool in a long history of cognitive offloading.
Offloading means using the environment to reduce mental work. Writing a shopping list is cognitive offloading. So is saving a phone number. The question is not whether offloading is bad; civilisation depends on it.
AI changes the scale because it can offload generation, summarisation, explanation, planning and even judgement. That can free time for higher-value thinking. A manager who spends less time formatting reports may spend more time understanding the business.
But there is a catch. Skills improve through use. If AI drafts every difficult paragraph, we may become better editors and worse first-draft thinkers. If it proposes every option, we may become efficient choosers and weaker generators of alternatives.
The key distinction may be substitution versus augmentation. 'Do this for me' removes cognitive work. 'Challenge my reasoning' can increase it. The same tool can produce either outcome depending on the prompt and workflow.
I like a simple rule for learning: attempt before assistance. Write the rough answer, solve the problem, form the opinion—then use AI to critique, expand or test it. That preserves retrieval and independent reasoning while still using the machine.
Organisations should make similar choices. If junior analysts never build models because AI does them instantly, where will senior judgement come from ten years later? Productivity today can quietly affect capability tomorrow.
AI can make us think less. It can also let us think at a higher level. The technology does not decide which one happens. The habits around it do.
Selected research anchors & further reading
Pearson, J. et al. (2026). Examining human reliance on artificial intelligence in decision making. Scientific Reports. Source
Schilke, O., & Reimann, M. (2025). The transparency dilemma: How AI disclosure erodes trust. OBHDP. Source
Phan, K. D., & Truong-Dinh, B. Q. (2026). When conversational AI personalises too much. Computers in Human Behavior. Source
Lee, R. S., De Silva Kanakaratne, M., & van der Veen, R. (2026). Consent management and chatbot anthropomorphism. Journal of Business Research. Source
AI or human: How the type of information to be disclosed alters customer service agent preferences (2026). Source
About Human Signals
Human Signals is Alok Jha's personal publication about psychology, behaviour and the choices we make. It explores the mind, decision-making, money, business and the changing relationship between humans and AI.
The editorial ambition is simple: explain people without preaching to them. Evidence matters, but so do everyday details—the delayed phone call, the abandoned shopping cart, the retirement account someone is afraid to spend, the employee who stays silent in a meeting, the AI answer that sounds more certain than it should.
About Alok Jha
Alok Jha is an entrepreneur, business strategist, mentor and student of human behaviour. He has more than three decades of corporate and CXO-level experience across fintech, technology, startups and new business development, along with postgraduate study in management and psychology. Human Signals brings together those two long-running interests: how people behave, and what that behaviour changes in real life and business.