Psychology in the AI Era

What Changed — Why I Started Tracking This

Generative AI went from lab curiosity to everyday tool in just a few years — reshaping how billions of people think, work, feel, and relate. Psychology is racing to catch up. This note is my map of what I'm watching, and how I apply it in my own life.

NEW QUESTIONS

Genuinely new territory

How do humans trust, bond with, or defer to AI? What happens to learning, memory, creativity, and mental health when an AI is always one tap away?

A TWO-WAY MIRROR

It teaches us about us

AI both changes the human mind and can be probed with psychological methods — so studying it reveals things about machines and about ourselves.

REAL STAKES

Attention, truth, connection

Our focus, relationships, sense of truth, and fairness are all in play. Understanding the psychology helps me navigate it deliberately rather than drift.

Why I keep this note open I don't need to be a researcher to use any of this. Understanding how AI shapes attention, trust, memory, and mood is, for me, basic life literacy now — it helps me make better decisions, protect wellbeing, and stay in the driver's seat.

The Three Lenses — The Key to Making Sense of It

There are three distinct ways psychology and AI meet. I keep them separate — it's the single best tool I have for reading any headline or study, and for thinking clearly about my own life with AI.

LENS A

AI's effect on humans

How interacting with AI changes the human mind and behavior.

Trust, reliance & automation bias
AI companions & loneliness/attachment
Chatbots & mental-health outcomes
Learning, memory & "cognitive offloading"
Creativity, persuasion, misinformation

LENS B

AI as the subject ("machine psychology")

Applying psychological tests to the AI to characterize its behavior.

Personality / value inventories on LLMs
Cognitive biases & reasoning errors in models
Theory-of-mind & social reasoning tasks
LLMs as models of human cognition
Alignment, sycophancy, deception

LENS C

AI as a tool

Using AI to help do things — including the science itself.

LLMs as simulated ("silicon") participants
Automated qualitative coding / theming
Information discovery & summarizing
Transcription, data cleaning, analysis support
Drafting, brainstorming, idea generation

Each lens asks a different question Lens A: how is AI changing us? · Lens B: what does AI's behavior reveal — and not reveal — about minds? · Lens C: when can AI reliably stand in for, or assist, a human? Mixing these up is the source of most confused takes about AI.

Key Phenomena to Understand (2026)

The live areas where AI is changing minds and behavior right now. Each is a window into a different way AI touches the human experience — and the topics I keep seeing in the news and the science.

Human–AI trust & reliance

When do people over-trust (automation bias) or under-trust AI advice? How do explanation, confidence, and past errors shift our reliance?

AI companions & relationships

Parasocial bonds with chatbots and their effects on loneliness, attachment, and social skills — including risks for adolescents and isolated users.

AI & mental health

The promise and safety of AI therapy/coaching tools; crisis-handling; equity of access; where a chatbot helps versus harms.

Cognitive offloading

Does relying on AI for memory, writing, or problem-solving change what we learn, recall, or are able to do unaided?

Anthropomorphism & mind perception

When and why do people attribute emotions, intentions, or moral status to AI — and how that changes their behavior?

Persuasion & misinformation

AI-generated content's persuasive power, who is most susceptible, and what builds resistance.

Machine psychology

Probing LLMs with personality scales, bias tasks, and theory-of-mind tests — and debating what such "scores" actually mean.

Bias & fairness

Algorithmic bias in screening, diagnosis, hiring, or risk tools — and its effects on stigmatized groups.

Work, identity & wellbeing

How AI at work reshapes autonomy, competence, meaning, job anxiety, and professional identity.

How AI Is Reshaping the Mind

AI tools now sit between us and the world — answering, recommending, writing, remembering. That quietly changes core mental processes. Here's what psychology is learning about the effects I watch for in my own life.

OFFLOADING

Outsourcing thought

We hand memory and reasoning to AI. Like GPS and our sense of direction, convenience can erode the very skills we stop practicing — so choose what to keep doing yourself.

ATTENTION

The focus economy

AI-curated feeds optimize for engagement, fragmenting attention. Sustained focus is a trainable, valuable resource worth defending.

MEMORY

The "Google effect"

When we expect facts to be retrievable, we remember less of the content and more of where to find it. Ever-present AI amplifies this trade-off.

TRUST

Automation bias

We tend to over-trust confident machines, accepting wrong answers because they sound authoritative. The skill is calibrated trust: trust, then verify.

ANTHROPOMORPHISM

Seeing a mind that isn't there

Fluent language makes us attribute feelings and understanding to AI. Handy for rapport — risky for judgment and emotional reliance.

CREATIVITY & SKILL

Boost or crutch?

AI can lift output but may flatten originality and stall skill-building if it replaces rather than supports my own effort.

The question that decides the outcome For any task, ask: "Is the AI doing this with me (augmenting) or for me (replacing)?" Augmentation builds skill and understanding; wholesale replacement quietly erodes them. Choose on purpose.

AI, Emotion & Society

Beyond individual cognition, AI is reshaping how we connect, feel, and decide what's true — at the scale of whole societies. These are the human dynamics worth understanding.

COMPANIONSHIP

Always-available bonds

Companion chatbots offer judgment-free, on-demand interaction. They can ease loneliness short-term — but may also displace human relationships and set unrealistic expectations, especially for the young and isolated.

MENTAL HEALTH

Access vs. safety

Therapy and coaching bots widen access and lower stigma — but raise real safety questions (e.g., handling a crisis) and can't replace human care for serious conditions.

PERSUASION

Influence at scale

AI can generate persuasive, personalized content — and convincing fakes — cheaply and instantly. Simply knowing this builds healthy skepticism.

BIAS

Inherited prejudice

Models learn society's biases; in hiring, lending, or assessment they can quietly disadvantage groups. Fairness must be checked, never assumed.

WORK & IDENTITY

Meaning in flux

AI reshapes autonomy, competence, and a sense of meaning at work — a source of both anxiety and genuine opportunity.

PARASOCIAL TRUST

One-sided relationships

We form bonds with — and trust — systems that don't reciprocate and have no stake in us. Recognizing this keeps expectations grounded.

Hold strong claims loosely These are open, fast-moving questions; the evidence is still emerging and findings often conflict. Treat both utopian ("AI will fix everything") and doom ("AI will ruin us") narratives with caution — reality is usually smaller, mixed, and conditional.

Machine Psychology — What AI Reveals About Us

Researchers now run classic psychology experiments on AI models. The results are a fascinating mirror — they teach us about human cognition, with some important caveats.

THE MIRROR

LLMs echo human biases

Given the same anchoring, framing, or availability setups, models often shift just like people do — because they learned from human text. This both exposes the biases baked into our language and offers a sandbox for studying how thinking goes wrong.

THE CAVEAT

But it isn't a mind

A model "scoring" extraverted on a personality test isn't extraverted — change the prompt or the version and the "score" swings. Treat these results as the behavior of a text predictor, not evidence of inner feelings, beliefs, or intentions.

THE USE

Testbeds for theories

LLMs are increasingly used to explore theories of language and reasoning. They're useful analogies and idea generators — not proof about how human brains actually work.

THE WARNING

Sycophancy & alignment

Models tend to tell us what we want to hear and can be steered to deceive. Studying this helps us design — and personally use — these systems more wisely.

The anthropomorphism trap Fluent words tempt us to assume understanding, feelings, or intent behind them. Stay behavioral: describe what the model does ("produced", "responded"), not what it "thinks" or "wants". This single discipline prevents most confused conclusions about AI.

What the Evidence Says — Think Critically

The AI–psychology space is loud with bold claims. A handful of thinking tools I use to stay grounded — and they double as essential media-literacy skills for the AI era.

CAUSATION

Correlation ≠ causation

"Teens who use AI more are lonelier" doesn't mean AI causes loneliness — lonely teens may simply seek it out. Demand the right study design before believing a causal claim.

VERSION DRIFT

Which model, when?

A finding about one model or version may not hold next month. Always ask which system was tested, and when.

HYPE vs. EVIDENCE

Beware the extremes

Both "AI will save us" and "AI will destroy our minds" outrun the data. Most real effects are smaller, conditional, and mixed.

PEER REVIEW

Pre-print ≠ proven

Much AI research is fast and not yet peer-reviewed. Check whether a striking result has been independently replicated.

CONFLICTS

Who funded it?

A lot of AI research comes from companies with a stake in the outcome. Note potential conflicts of interest.

SOURCES

Verify the claim

AI summaries and chatbots can produce confident, false "facts" and fake citations. Always trace an important claim back to a real source.

The one question I carry everywhere When I meet any claim about AI and the mind, I ask: "What's the evidence, how strong is it, and who benefits from me believing it?" That single question has saved me in the news feed and in the scientific literature alike.

The Ethical & Human Stakes

For me, understanding psychology in the AI era means grappling with the human questions AI raises — in daily life and at society scale. These aren't abstract; they show up in everyday choices.

PRIVACY

Your data is the product

Your conversations, prompts, and uploads may be stored and reused. A safe rule: share with AI only what you'd be comfortable seeing made public.

MANIPULATION

Personalized persuasion

AI can nudge beliefs and purchases at scale, tuned to you. Awareness that it's happening is the first line of defense.

FAIRNESS

Bias at scale

Biased systems can entrench inequality in high-stakes decisions. Demand transparency and accountability where AI judges people.

AUTONOMY

Dependence vs. agency

Over-reliance can erode skills and self-trust. Keep a human in the loop — and make sure that human is, sometimes, just you.

VULNERABILITY

Who's most affected

Children, lonely, and distressed people are most influenced by companion and advice bots. Extra care and oversight are warranted.

ACCOUNTABILITY

People answer, not tools

AI bears no responsibility — people do. Decisions that affect lives must stay answerable to a human.

Apply It — Live Well With AI

How I turn the science into daily practice. These habits let me capture AI's benefits while protecting my mind, my relationships, and my judgment.

AUGMENT

Augment, don't outsource

Use AI to extend your thinking, not replace it. Do the hard cognitive work yourself whenever the skill itself matters to you.

VERIFY

Verify, then trust

Treat AI output as a confident first draft. Check anything important against a real source, and calibrate trust to its track record.

ATTENTION

Protect your attention

Curate feeds, mute engagement traps, and set boundaries on always-on tools. Your attention is your scarcest, most valuable resource.

CONNECTION

Keep humans central

Let AI supplement — never substitute — real relationships. Protect face-to-face time and reciprocal, two-way bonds.

SKEPTICISM

Question content

Assume persuasive or emotional content may be AI-generated or biased. Cross-check before you believe it or pass it on.

LITERACY

Build AI literacy

Learn how these tools actually work — pattern prediction, not understanding. Knowing the mechanism deflates both needless fear and naive over-trust.

The augmentation test Before reaching for AI, ask: "Will this make me more capable, or just save effort I'd actually grow from?" Both can be the right call — but make it a choice, not a reflex.

Stay Human, Stay Happy

AI is a tool; my wellbeing is mine to steward. Psychology's evidence on what makes life good matters more than ever in a noisy, automated world — these are the anchors I don't let AI replace.

CONNECTION

People over products

Real relationships remain the strongest predictor of a happy, healthy life — no chatbot replaces them. Invest in people first.

MASTERY & MEANING

Don't outsource growth

Skills you build yourself and work that feels purposeful fuel lasting satisfaction. Keep doing hard things that are worth doing.

ATTENTION = LIFE

Guard your focus

What you pay attention to becomes your experience of life. Protect it from engineered, endless distraction.

THE BASICS

Sleep & move

Exercise and sleep still beat any app for mood, memory, and focus. They're the foundation everything else rests on.

AGENCY

Stay the author

Use AI deliberately; don't drift on autopilot. Being the one who decides — not the one who's nudged — is its own kind of wellbeing.

OFFLINE TIME

Awe & stillness

Time away from screens — nature, hobbies, quiet — restores the mind in ways no AI can. Build it in on purpose.

The one habit I protect most Use AI on purpose, then put it down. The goal was never to use AI more — it's to live better, with AI as one tool among many.

What I keep coming back to Psychology in the AI era gives me three things: a map of how AI shapes minds, critical-thinking tools to cut through hype, and practical habits to stay sharp, connected, and well. Knowledge earns its keep when I live it.

References & Sources

Annotated bibliography behind the three-lenses framework, key phenomena, mind-and-society effects, machine psychology, critical-thinking tools, ethical stakes, and applied habits in this note. Section tags (e.g. §04) show where each source is used. The three-lenses map and synthesis tables are my own unless noted.

Scope. Synthesis of peer-reviewed psychology, HCI, and AI research plus authoritative policy and survey sources (May 2026). Findings on companions, offloading, and mental-health bots are fast-moving and often mixed — treat effect sizes as directional. Crisis-handling by chatbots is an active safety concern, not a solved problem. Not medical, therapeutic, or diagnostic advice.

Citations are numbered continuously [1]–[n] within this section.

Generative AI adoption & the new research landscape (§01)

Pew Research Center, "34% of U.S. adults have used ChatGPT, about double the share in 2023." June 2025. Awareness and usage growth — §01 generative-AI scale and everyday-tool framing. pewresearch.org — §01.
Stanford HAI, 2025 AI Index Report — Research & Development chapter. 2025. AI research volume, model releases, and cross-disciplinary uptake — §01 psychology "racing to catch up" context. hai.stanford.edu/ai-index — §01.
National Academies of Sciences, Engineering, and Medicine, Human–AI Teaming: State of the Art and Research Needs. NAP, 2022. Interdisciplinary framing for human factors in AI systems — §01 two-way-mirror card. nap.nationalacademies.org/catalog/26355 — §01.

Three lenses — humans, machines, and tools (§02–§03)

Shneiderman, B., Human-Centered AI. Oxford University Press, 2022. Reliable–safe–trustworthy AI with humans in control — intellectual background for Lens A vs. Lens C distinction. — §02, §09.
American Psychological Association, "Artificial Intelligence and Psychology" resource hub. APA ongoing guidance on AI in research and practice — §02 Lens A/C applied branches. apa.org/topics/artificial-intelligence — §02, §08.
Argyle, L. P. et al., "Out of One, Many: Using Language Models to Simulate Human Samples." Political Analysis, 31(3), 337–351, 2023. "Silicon sampling" with demographic conditioning — §02 Lens C and §03 machine-psychology phenomena. DOI: 10.1017/pan.2023.2 — §02, §03.
Aher, G. V. et al., "Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studies." ICML 2023. "Turing Experiments" replicating classic psychology paradigms with LLMs — §02 Lens C with caveats on hyper-accuracy distortion. proceedings.mlr.press/v202/aher23a — §02, §06.

Trust, automation bias & calibrated reliance (§03–§04, §09)

Parasuraman, R., & Manzey, D. H., "Complacency and Bias in Human Use of Automation: An Attentional Integration." Human Factors, 52(3), 381–410, 2010. Automation bias and over-reliance on decision aids — §03 trust/reliance and §04 automation-bias card. DOI: 10.1177/0018720810376055 — §03, §04, §09.
Buçinca, Z. et al., "To Trust or to Think: Cognitive Forcing Functions Can Reduce Overreliance on AI in AI-Assisted Decision-Making." Proceedings of the ACM on Human-Computer Interaction, 5(CSCW1), 2021. Interventions that reduce blind trust in AI recommendations — §04 trust card and §09 verify-then-trust habit. DOI: 10.1145/3449287 — §04, §09.
Lee, J. D., & See, K. A., "Trust in Automation: Designing for Appropriate Reliance." Human Factors, 46(1), 50–80, 2004. Foundational model of trust calibration in automated systems — §04 calibrated-trust callout. DOI: 10.1518/hfes.46.1.50_30392 — §04.

Cognitive offloading, memory & attention (§04, §09–§10)

Sparrow, B., Liu, J., & Wegner, D. M., "Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips." Science, 333(6043), 776–778, 2011. Transactive memory / "Google effect" — §04 memory card. DOI: 10.1126/science.1207745 — §04.
Risko, E. F., & Gilbert, S. J., "Cognitive Offloading." Trends in Cognitive Sciences, 20(9), 676–688, 2016. Framework for when and why people outsource cognition — §04 offloading card. DOI: 10.1016/j.tics.2016.07.002 — §04.
Kosmyna, N. et al., "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task." arXiv:2506.08872, 2025. EEG study of LLM-assisted writing and reduced ownership/connectivity — §04 offloading/creativity cards (emerging; preprint). arxiv.org/abs/2506.08872 — §04.
Harris, T., & Raskin, A., The Social Dilemma / Center for Humane Technology resources. 2020–25. Attention-economy and engagement-design critique — background for §04 attention card and §10 guard-focus card. humanetech.com — §04, §10.

Anthropomorphism, CASA & parasocial bonds (§03–§05, §06)

Nass, C., Steuer, J., & Tauber, E. R., "Computers Are Social Actors." CHI 1994, 72–78. CASA paradigm — people apply social rules to computers — §04 anthropomorphism card and §06 anthropomorphism-trap callout. DOI: 10.1145/191666.191703 — §04, §06.
Epley, N., Waytz, A., & Cacioppo, J. T., "On Seeing Human: A Three-Factor Theory of Anthropomorphism." Psychological Review, 114(4), 864–886, 2007. When and why people attribute minds to non-human agents — §03 mind-perception phenomena. DOI: 10.1037/0033-295X.114.4.864 — §03, §04.
Horton, D., & Wohl, R. R., "Mass Communication and Para-Social Interaction." Psychiatry, 19(3), 215–229, 1956. Parasocial relationship concept — §05 parasocial-trust card. — §05.
Skjuve, M. et al., "My Chatbot Companion — A Study of Human–Chatbot Relationships." International Journal of Human-Computer Studies, 149, 102601, 2021. Mixed-method Replika relationship study — §03 companions and §05 companionship card. DOI: 10.1016/j.ijhcs.2021.102601 — §03, §05.

AI companions, loneliness & adolescent risk (§03, §05, §08)

Maples, B. et al., "Loneliness and Suicide Mitigation for Students Using GPT3-Enabled Chatbots." npj Mental Health Research, 2, 12, 2023. Replika user survey — loneliness, social support, and safety concerns — §03 companions and §05 mental-health tension. DOI: 10.1038/s44184-023-00047-6 — §03, §05.
Common Sense Media, Talk, Trust, and Trade-Offs: How and Why Teens Use AI Companions. 2025. Teen use of Character.AI, Replika, and related platforms — §03 adolescent risk and §08 vulnerability card. commonsensemedia.org — §03, §08.
Common Sense Media, AI Risk Assessment: Social AI Companions. 2025. "Unacceptable risk" rating for under-18 users — §08 vulnerability and accountability stakes. commonsensemedia.org/ai — §08.

AI mental-health tools — access, evidence & safety failures (§03, §05, §08)

Fitzpatrick, K. K. et al., "Delivering CBT to Young Adults Using a Fully Automated Conversational Agent (Woebot)." JMIR Mental Health, 4(2), e19, 2017. RCT showing feasibility and symptom reduction — §03/§05 mental-health promise card. DOI: 10.2196/mental.7785 — §03, §05.
Torous, J. et al., "The Growing Field of Digital Psychiatry: Current Evidence and the Future of Apps, Social Media, Chatbots, and Virtual Reality." World Psychiatry, 20(3), 318–335, 2021. Evidence overview and safety guardrails for digital mental-health tools — §05 access-vs-safety card. DOI: 10.1002/wps.20883 — §05, §08.
NBC News / NEDA statements on "Tessa" chatbot suspension. June 2023. Eating-disorder chatbot giving harmful dieting advice — §05/§08 crisis-handling and accountability caution. nbcnews.com — §05, §08.
World Health Organization, Ethics and Governance of Artificial Intelligence for Health. WHO, 2021. Guidance on safety, equity, and human oversight in health AI — §08 accountability and vulnerability cards. who.int — §08.

Persuasion, misinformation & fake citations (§03, §07, §09)

Goldstein, J. A. et al., "How Persuasive Is AI-Generated Propaganda?" PNAS Nexus, 3(2), pgae034, 2024. GPT-3 propaganda nearly as persuasive as real foreign propaganda — §03 persuasion phenomena and §05 influence card. DOI: 10.1093/pnasnexus/pgae034 — §03, §05, §09.
Pennycook, G., & Rand, D. G., "Fighting Misinformation on Social Media Using Crowdsourced Judgments of News Source Quality." PNAS, 116(7), 2521–2526, 2019. Source-quality reasoning and skepticism — background for §07 verify-sources card. DOI: 10.1073/pnas.1806781116 — §07, §09.
Walters, W. H., & Wilder, E. I., "Fabrication and Errors in the Bibliographic Citations Generated by ChatGPT." Scientific Reports, 13, 14045, 2023. ~55% fabricated citations in GPT-3.5 outputs — §07 fake-citation card and §09 verify habit. DOI: 10.1038/s41598-023-41032-5 — §07, §09.
European Union, Regulation (EU) 2024/1689 (AI Act) — Article 5 prohibited practices. 2024. Bans manipulative/deceptive AI that materially distorts behavior — §08 manipulation card. eur-lex.europa.eu — §08.

Machine psychology — personality, biases & theory of mind (§03, §06)

Serapio-García, G. et al., "Personality Traits in Large Language Models." arXiv:2307.00184, 2023; NAACL 2024 follow-ons. Psychometric testing of LLM "personality" — §03 machine psychology and §06 mirror card. arxiv.org/abs/2307.00184 — §03, §06.
Binz, M., & Schulz, E., "Using Cognitive Psychology to Understand GPT-3." PNAS, 120(6), e2218523120, 2023. LLMs show human-like heuristic biases on classic tasks — §06 LLM-echoes-biases card. DOI: 10.1073/pnas.2218523120 — §06.
Hagendorff, T. et al., "Human-Like Intuitive Behavior and Reasoning Biases Emerged in GPT-3.5 but Not in GPT-3." arXiv:2305.04400, 2023. Anchoring, framing, and other biases in instruction-tuned models — §06 mirror card. arxiv.org/abs/2305.04400 — §06.
Kosinski, M., "Evaluating Large Language Models in Theory of Mind Tasks." PNAS, 121(45), e2405460121, 2024. False-belief tasks on LLMs — §03 theory-of-mind bullet and §06 testbed card (interpret with anthropomorphism caveat). DOI: 10.1073/pnas.2405460121 — §03, §06.
Sharma, M. et al., "Towards Understanding Sycophancy in Language Models." ICLR 2024 (arXiv:2310.13548). RLHF-linked tendency to agree with users — §06 sycophancy card. arxiv.org/abs/2310.13548 — §06.

Bias, fairness & work identity (§03, §05, §08)

Buolamwini, J., & Gebru, T., "Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification." FAT* 2018. Algorithmic bias in deployed systems — §03 bias/fairness and §05 bias card. proceedings.mlr.press/v81/buolamwini18a — §03, §05, §08.
Barocas, S., Hardt, M., & Narayanan, A., Fairness and Machine Learning. fairmlbook.org, 2019–23. Framework for fairness in automated decision systems — §08 fairness card. fairmlbook.org — §08.
Deci, E. L., & Ryan, R. M., "Self-Determination Theory: Basic Psychological Needs in Motivation, Development, and Wellness." Guilford, 2017. Autonomy, competence, relatedness — §03 work/identity card and §10 mastery/meaning card. — §03, §10.
Autor, D., "AI Could Actually Help Rebuild The Middle Class." Noema / MIT Work of the Future, 2024. Productivity, task redesign, and meaning at work — background for §03/§05 work-identity card (directional, contested). noemamag.com — §03, §05.

Critical evaluation, replication & hype (§07)

Open Science Collaboration, "Estimating the Reproducibility of Psychological Science." Science, 349(6251), aac4716, 2015. Replication baseline for psychology — §07 peer-review/replication card. DOI: 10.1126/science.aac4716 — §07.
Ioannidis, J. P. A., "Why Most Published Research Findings Are False." PLOS Medicine, 2(8), e124, 2005. Conflicts, power, and hype — background for §07 conflicts-of-interest card. DOI: 10.1371/journal.pmed.0020124 — §07.
Nosek, B. A. et al., "Promoting an Open Research Culture." Science, 348(6242), 1422–1425, 2015. Pre-registration and transparency norms — §07 pre-print card. DOI: 10.1126/science.aab2374 — §07.

Governance, privacy & applied habits (§08–§09)

NIST, Artificial Intelligence Risk Management Framework (AI RMF 1.0). 2023. Govern–Map–Measure–Manage cycle — §08 transparency/accountability cards. nist.gov/ai-rmf — §08.
FTC, "Commercial Surveillance and Data Security Rulemaking" & AI consumer guidance. 2023–25. Data use, dark patterns, and consumer protection — §08 privacy card. ftc.gov/ai — §08.
Mollick, E., & Mollick, L., "Using AI to Implement Effective Teaching Strategies in Classrooms." SSRN, 2023; Wharton guidance on structured AI use. Augment-with-verification pattern — §09 augment/verify cards. ssrn.com/abstract=4475995 — §09.
Geerlings, H. et al., "From Offloading to Engagement: An Experimental Study on Structured Prompting and Critical Reasoning with Generative AI." Journal of Data Analysis & Information Processing, 10(11), 2025. Structured prompting reduces unthinking offloading — §04 augmentation callout and §09 literacy card. — §04, §09.

Wellbeing anchors in an automated world (§10)

Waldinger, R., & Schulz, M., The Good Life: Lessons from the World's Longest Scientific Study of Happiness. Simon & Schuster, 2023. Harvard Study of Adult Development — relationships predict health and happiness — §10 connection card. adultdevelopmentstudy.org — §10.
Seligman, M. E. P., Flourish. Free Press, 2011. PERMA wellbeing model — companion to mastery/meaning framing in §10. — §10.
Twenge, J. M., "Have Smartphones Destroyed a Generation?" The Atlantic, 2017; updated meta-analyses on screen time and wellbeing. Cautionary context for always-on tools — §10 attention/offline-time cards (debated effect sizes). theatlantic.com — §10.

Author synthesis

Truong, L., Psychology in the AI Era — personal working notes. May 2026. Three-lenses map, phenomena grid, applied habit cards, and synthesis prose. LinhTruong.com — all sections.

Before you quote externally: Machine-psychology "scores" (personality, theory of mind) are prompt- and version-sensitive — not evidence of inner mental states. Companion-chatbot and offloading studies often conflict; adolescent-risk reports are policy-relevant but still evolving. Preprints (e.g. cognitive-debt EEG work) should be treated as provisional. Always verify model names, dates, and DOIs against primary papers.