Author: Luiza Jarovsky Source: https://www.luizasnewsletter.com/p/how-ai-is-shaping-us Published: April 24, 2026 (edition #289)
"How AI Is Shaping Us" is a prescriptive essay by Luiza Jarovsky, published April 24, 2026 as edition #289 of her newsletter. It argues that generative AI is the first technology to automate general cognitive processes at scale, and that this creates a skill-atrophy problem analogous to the physical decline produced by sedentary work. Its central prescription is cognitive friction: deliberately scheduled unaided practice sessions intended to maintain skill in any domain a person considers part of their core professional expertise. Full framework coverage is at Cognitive Friction.
Summary of argument
Jarovsky frames people as continually shaped by the environments, people, and technologies that surround them. On this view, generative AI is the first technology that automates general cognitive processes en masse, which she presents as a friction-loss problem analogous to the sedentary-lifestyle physical-friction problem.
The argument proceeds by analogy. Knowledge workers are sedentary, and human bodies were not designed for screens; the essay describes physical friction — walking, running, cycling, weights, the gym — as the deliberate compensation that prevents physical degradation. Generative AI, in this framing, presents the cognitive analog. Because cheap, widely available LLMs perform many cognitive tasks faster than humans, Jarovsky argues that, without cognitive friction, people will become noticeably impaired in any domain AI performs better than they do, comparing this to the loss of basic mental arithmetic and unaided geolocation that followed calculators and navigation apps.
Key claims
Jarovsky supports the argument with several documented effects drawn from recent literature:
- AI does not reduce work but intensifies it (Harvard Business Review, February 2026), via workload creep leading to cognitive fatigue, burnout, and weakened decision-making.
- Skill formation degradation in junior workers (arXiv 2601.20245): "the aggressive incorporation of AI into the workplace can have negative impacts on the professional development of workers if they do not remain cognitively engaged."
- The LLM fallacy (arXiv 2604.14807), defined as "a cognitive attribution error in which individuals misinterpret LLM-assisted outputs as evidence of their own independent competence, producing a systematic divergence between perceived and actual capability."
- Cognitive debt (Kosmyna et al., June 2025).
On choice and literacy, Jarovsky argues that how to use AI, and how much cognitive impairment to accept, is each person's decision, but that without AI literacy and awareness of AI's cognitive, psychological, and physical effects there can be no sound choices, and most people will be swayed by the prevailing direction of AI adoption. She connects this to her broader literacy divide thesis.
The essay illustrates these points with examples including children using AI for homework, people in romantic relationships with AI chatbots, companies assessing employee performance based on AI use, and social-media posts in which users claim to have become "much smarter" since starting to use AI — to which Jarovsky responds, "Well, maybe it is because the AI system is actually doing the writing."
Jarovsky positions cognitive friction as a concrete, individually actionable practice modeled on physical exercise, and presents the LLM fallacy as the demand-side mechanism — a documented academic finding — that, in her account, makes such a practice necessary. She frames individual cognitive degradation as connected to broader societal divides, with AI literacy as a differentiator, and presents the essay as a counter-narrative to the framing, common on social media, that AI use makes its users smarter.
Provenance
The essay appeared in Jarovsky's newsletter on April 24, 2026 (edition #289). It advances an original prescriptive framework rather than reporting news, and is treated here as a foundational source stating a position.
Relationships
- authored-by: Luiza Jarovsky
- introduces: Cognitive Friction, LLM Fallacy (cited from arXiv but established in this essay's framing)
- related: AI Divides (Literacy / Occupational / Ethico-Philosophical), AI Acceleration Paradox, AI Mental Health and Psychological Harm