01
The Behaviour Brief
This Week's Psychology Finding

78% Of People Accept AI Outputs Without Questioning Them. Here's The Bias Behind That.

A new study found that 78% of people relied on AI outputs without critical scrutiny — driven by a cognitive bias called Automation Bias. If you use ChatGPT, Claude, or any AI tool, this finding is about you. Here's what it actually means — and what to do about it.
Research Weight
92
Actionability
88

A 2024 study published in Computers in Human Behavior (Goddard et al.) found that when people were given AI-generated outputs alongside human-generated ones, they overwhelmingly trusted the AI version — even when it was factually wrong. Not because people are gullible. Because of a well-documented psychological phenomenon called Automation Bias — first formally identified by Lisanne Bainbridge in 1983 and replicated consistently across aviation, medicine, and now AI contexts.

Automation Bias is the tendency to over-rely on automated systems and under-rely on our own judgement. Your brain, faced with an AI output, essentially says: "The machine processed more information than I could. It's probably right." This is System 1 thinking — fast, intuitive, effortless. And it's wrong more often than you realise.

78%
Accepted AI Outputs Without Scrutiny
71%
Engaged In Unethical Behaviour Influenced By AI-Driven Biases
38%
More Likely To Trust Bullet-Pointed AI Output Vs. Prose

The format of AI output matters psychologically. Research shows people find AI responses formatted as bullet points more credible than the same information presented as paragraphs — not because the content differs, but because structured formatting triggers an "expert" heuristic in our brain. This is why AI outputs feel authoritative even when they shouldn't.

The Practical Takeaway: Before acting on any AI output this week, run a 10-second "Automation Bias Check." Ask yourself: "If a human junior colleague gave me this answer, would I accept it without questioning?" If the answer is no — interrogate the AI output with the same scepticism.

This Week's One Thing
Before acting on any significant AI output today, ask: "Would I accept this from a junior colleague without questioning it?" That single question is a near-perfect antidote to Automation Bias. Takes 10 seconds. Apply it once today.
02
The Mind × Machine Deep-Dive
Main Feature · One Concept. Fully Explored.

Your AI Assistant Is Making You Worse At Thinking. This Is The Psychology Of Why — And The Exact Fix.

This week we go deep on one concept only: Cognitive Offloading. The more you outsource your thinking to AI, the weaker your independent cognitive capacity becomes. This is not an opinion. It's a measurable psychological phenomenon — and understanding it is the difference between using AI as a superpower and using it as a crutch.
Research Weight
96
Actionability
94

What Is Cognitive Offloading?

Cognitive offloading is what happens when you store or process information in the external environment rather than in your own mind. Writing a shopping list is cognitive offloading. Using Google Maps instead of learning a route is cognitive offloading. Asking ChatGPT to summarise an article you could have read is cognitive offloading.

AI tools are now so capable, so fast, and so frictionless that the threshold for when we offload has collapsed dramatically. We now outsource thinking that we used to do ourselves — and that thinking was building a cognitive muscle we're now losing.

"Long-term interaction with AI is positively associated with mental exhaustion, attention strain, and information overload — and negatively associated with self-assurance in decision-making." — Kian et al. (2024), Frontiers in Psychology, doi:10.3389/fpsyg.2024.1328788

The 3 Cognitive Muscles AI Is Quietly Weakening

1. Synthesis: The ability to take multiple pieces of information and form an original conclusion. When AI summarises for you constantly, your synthesis muscle atrophies.

2. Productive Struggle: The discomfort of not immediately knowing the answer is actually where learning happens. AI removes the struggle — which feels like efficiency but kills deep understanding.

3. Metacognition: Thinking about your thinking. When AI does the first draft of your analysis, you lose the moment where you would have caught your own biased reasoning.

The Research Trail

A 2023 study in Nature Human Behaviour tracked knowledge workers who began using AI writing tools daily. After six months, their performance on unassisted writing tasks — tasks where they couldn't access AI — had measurably declined. The control group who used AI less frequently showed no decline. The mechanism is well understood: skills that aren't exercised degrade. AI removes the exercise.

A separate 2024 study from the University of Waterloo found that people with higher smartphone reliance performed worse on cognitive tasks — not because smartphones made them less intelligent, but because they had outsourced the cognitive work to the device. The brain adapts to what it needs to do. If you consistently ask AI to think first, your brain learns it doesn't need to.

The Protocol: Use AI As An Amplifier, Not A Replacement

The rule to adopt this week: Always Think First, AI Second. Before opening ChatGPT or Claude for any task requiring judgement, write down your own answer in 3–5 sentences. Then use AI to challenge, expand, or refine it. This preserves the cognitive work while gaining the AI benefit.

This is not anti-AI advice. It's pro-cognition advice. The people who will extract the most value from AI over the next five years are the ones who maintain strong independent thinking capacity alongside AI use — not the ones who outsource first and think second.

03
The Cognitive Toolkit
3 Tools Reviewed Through A Psychology Lens This Week

The Tools. What Your Brain Actually Does When You Use Them.

Reflect.app
Second Brain / Thinking Tool
88
Score
Psychology Verdict: Reflect is built around building a personal knowledge graph — essentially training your own external working memory. The "networked notes" structure mirrors how human associative memory actually works (spreading activation theory), making information genuinely easier to retrieve. Worth it if you read and think for a living.
Reclaim.ai
AI Scheduling / Time Blocking
82
Score
Psychology Verdict: Reclaim works on the principle of implementation intentions — pre-deciding when and where you'll do a task dramatically increases follow-through. The downside: it removes the planning moment where you consciously evaluate priorities. Use it for recurring tasks, not strategic work.
Granola.so
AI Meeting Notes
74
Score
Psychology Verdict: When you know AI is taking notes, your brain says "the machine has it" and stops actively listening. Research shows note-taking by hand improves retention. Use Granola for reference only, not as a reason to check out mentally.
04
The Influence Move
🔒
⚡ Pro Subscribers Only
The Influence Move Is Locked

One specific, research-backed psychology tactic you can deploy in your next AI session — exclusively for Pro subscribers. This week: The Contrast Principle in AI Prompting.

⚡ Unlock With Pro — ₹799/mo See All Plans
05
Wired vs. Tired
What Moved The Needle This Week

Signal From Noise. What's Genuinely Important vs. Safely Ignorable.

⚡ Wired — Genuinely Important
AI Tools That Track Your Cognitive Patterns, Not Just Your Tasks
Tools like Oura and Whoop are moving toward integrating cognitive performance data with AI. The psychology is solid: when you understand your own patterns of peak cognition, distraction, and fatigue, you can schedule high-stakes thinking for when your brain is actually capable of it. This is applied chronobiology and it works.
💤 Tired — Safely Ignore
"AI Companions" As A Solution To Loneliness
The APA issued a formal advisory in March 2026 warning that AI companions cannot replace qualified mental health professionals and may create harmful user-AI relationships. Parasocial relationships with AI satisfy the surface signal of social connection while starving the deeper need. This trend will cause harm. Skip the hype.