What our radars are built on
Our personal radars are not an internet quiz. Every area we measure comes from peer-reviewed research — the exact citations and links are below so you can check them yourself.
Instrument version 1.0 · reviewed 2026-08
Why we built the radars
AI changes not only what we do, but how we think. We hand over calculations, wording and decisions — so smoothly that nobody notices the drop in their own effort. Without measurement, neither a person nor a company can tell when helpful support has quietly become dependency.
The radar is therefore a self-assessment tool with a stable structure: fill it in today, again in three months, and watch the trend. The value is not in a single number but in the movement over time and the conversation it starts — with yourself, your team, or your cohort.
Six areas and the research behind them
Each area has five questions. For every one we state what it measures, why it matters, and which research it draws on.
Cognitive offloading and dependency
What it measures: How often we hand thinking work over to AI, and how we cope when AI is not available.
Scientific relevance: The more effort we delegate to a tool, the less we invest ourselves. A survey of 319 knowledge workers (Carnegie Mellon and Microsoft Research) shows that higher confidence in AI goes together with a self-reported reduction of one's own critical effort.
- Lee, H.-P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers. CHI 2025. Open source →
Critical thinking
What it measures: Whether we verify AI output, weigh it against our own judgement, and look for counter-arguments.
Scientific relevance: Gerlich's study (n = 666) found a negative relationship between frequent AI tool use and critical thinking ability, mediated precisely by cognitive offloading. The risk is not using AI, but accepting its output passively.
- Gerlich, M. (2025). AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking. Societies, 15(1), 6. Open source →
Memory and retention
What it measures: How much of AI-assisted work we actually remember, and how far we consider the result our own.
Scientific relevance: In an MIT Media Lab experiment (EEG, 54 participants), the group writing with an LLM showed the weakest brain connectivity, poorer recall of their own text, and a lower sense of ownership. The authors call this “cognitive debt” — effort saved today, retention weakened later.
- Kosmyna, N., Hauptmann, E., Yuan, Y. T., Situ, J., Liao, X.-H., Beresnitzky, A. V., Braunstein, I., & Maes, P. (2025). Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task. MIT Media Lab, arXiv:2506.08872. Open source →
Digital hygiene and burnout
What it measures: Screen time, breaks, sleep, and the ability to disconnect.
Scientific relevance: A validated digital burnout scale for the AI era (Frontiers in Psychology) confirms that digital exhaustion is a measurable, multi-factor construct — not a vague feeling but a state that can be tracked over time.
- Zhao, L., Zhao, J., Cao, E. (Y.), Li, K. (K.), Pan, L., Zou, Y., & Sun, X. (2026). Development and validation of a digital burnout scale in artificial intelligence era. Frontiers in Psychology, 16, 1580422. Open source →
Attention and focus
What it measures: Attention fragmentation, task switching, and the capacity for deep work.
Scientific relevance: The classic Stanford study showed that heavy media multitaskers are worse at filtering irrelevant stimuli. Longitudinal data in JAMA Network Open additionally link increasing social media use with a rise in depressive symptoms.
- Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. PNAS, 106(37), 15583–15587. Open source →
- Nagata, J. M., et al. (2025). Social Media Use and Depressive Symptoms During Early Adolescence. JAMA Network Open, 8(5), e2511704. Open source →
Work–life balance
What it measures: Whether AI actually removes work, or merely expands its scope and shifts the load.
Scientific relevance: A study by the Upwork Research Institute (2,500 respondents) describes the productivity paradox: 96% of leaders expect AI to raise productivity, yet 77% of employees report AI has added to their workload.
- Upwork Research Institute (2024). From Burnout to Balance: AI-Enhanced Work Models for the Future. Open source →
Two personal radars, two perspectives
Both draw on the same research base but look from opposite directions. Take one, ideally both.
Individual Cognitive Readiness
It asks how consciously you work with AI: whether you verify output, understand the tool's limits, can frame a task, and keep your own judgement. It measures competence and approach.
When to pick it: When you want to know how well you use AI and where to improve.
StartAI Cognitive Health Radar
It asks about impact: AI dependency, critical thinking, memory, digital hygiene, attention, and work–life balance. It measures state, not skill.
When to pick it: When you feel scattered, exhausted, or that you retain less of your work.
StartHow we read the results
The 0–100 score is the average of the six areas. A higher number means healthier functioning.
Your score indicates you're maintaining independent thinking and a healthy approach to AI. Continue with your current practices and monitor regularly.
We noticed some signs of increased AI dependency or reduced digital hygiene. Consider more breaks, limiting screen time, and practicing independent thinking.
Your score indicates a decline in critical cognitive functions. We strongly recommend digital detox, reducing AI dependency, and establishing clear work boundaries.
Your responses indicate serious concerns. Please consider consulting a doctor or digital wellness expert and implement significant changes in technology usage.
What the radar is not
- It is a self-assessment tool, not a diagnosis. It does not replace a physician, psychologist, or clinical assessment.
- We use the data anonymised. In a corporate setting we only show an aggregate from three respondents up, so no individual can be identified.
- A single run is a snapshot of one day. What matters is repeated measurement and the trend over time.
How to cite us
Using the results in your own report or deck? Copy this block — and please always add the completion date, the instrument version, and a note that this is self-assessment.
AI Cognitive Health Radar (v1.0, beCOHORTs.com, 2026). Built on: - Lee et al. (2025), CHI 2025 — generative AI and critical thinking - Gerlich (2025), Societies 15(1):6 — cognitive offloading - Kosmyna et al. (2025), MIT Media Lab, arXiv:2506.08872 — cognitive debt - Zhao et al. (2026), Frontiers in Psychology 16:1580422 — digital burnout scale - Ophir, Nass & Wagner (2009), PNAS 106(37) — media multitasking - Nagata et al. (2025), JAMA Network Open 8(5):e2511704 — social media and mood - Upwork Research Institute (2024) — AI productivity paradox
What about the corporate radars?
AI Quick Check and AI Deep Dive measure something else: organisational practice — how employees perceive AI adoption, transparency, education and ethics in the company. They do not measure personal cognitive state, so they rest on a different methodological base. If the corporate version interests you, get in touch.