A question worth sitting with before your next prompt: if the AI went away tomorrow, would your judgment still work?
Sarker's point, in short: we are building machines that process vast information, generate arguments, and imitate reasoning — while many societies rushing to adopt them struggle to teach the skills needed to question their answers: literacy, numeracy, critical thinking, media literacy, civic reasoning.
The divide that matters, the piece argues, isn't access to the tool. It's whether the habit of independent judgment survives the tool.
Researcher Michael Gerlich at SBS Swiss Business School surveyed 666 UK respondents and conducted 50 in-depth interviews. The findings, in plain language:
Honest limits: this is one study. It used self-report measures, and a correlation is not proof of causation — it could be that weaker critical thinkers reach for AI more, not (only) that AI weakens thinking. Treat it as evidence, not a verdict.
Scammers now write with the same fluent, confident AI you do. The phishing email, the fake invoice, the "urgent" message from your "boss," the deepfaked voice of your grandchild — they all read perfectly. Fluency is no longer a signal of truth.
That means the defense against AI-powered deception is exactly the habit Sarker describes: independent verification. Cross-check. Look for the source. Ask for the opposite argument. The person who verifies survives the fluent lie; the person who outsources judgment clicks it.
Critical thinking isn't just an academic virtue anymore. It's the firewall.
Answer honestly. "Yes" means you actually do the habit — not that you intend to.
Researcher Michael Gerlich at SBS Swiss Business School surveyed 666 UK respondents and found AI use and critical thinking moved in opposite directions (correlation r = -0.68). Mediation analysis suggested roughly 60% of the negative link ran through cognitive offloading. The study is correlational: it describes a pattern, not proven causation.
Cognitive offloading is handing memory and problem-solving over to devices instead of doing the work yourself. In the study it was strongly linked to AI use (r = 0.72) and inversely linked to critical thinking (r = -0.75).
Scammers now write with the same fluent, confident AI you do - phishing emails, fake invoices, and deepfaked voices all read perfectly. Fluency is no longer a signal of truth, so the defense is independent verification: cross-check, look for the source, ask for the opposite argument.
Ages 17-25 used AI the most and scored lowest on critical thinking; those 46+ used it least and scored highest. In interviews, respondents with advanced degrees tended to cross-check AI answers while less-schooled respondents tended to follow the tool.
It tests verification habits: whether you check important AI answers against another source, look for sources yourself, test the opposite side, and can explain how you verified the last important AI answer you used.