The Algorithm Is Now a Formative Institution
Recommendation systems do not only distribute content. At scale, they teach people what to notice, imitate, desire, and believe.
Articles
A civic briefing room for foundations, context, consequences, what we protect, and action.
Recommendation systems do not only distribute content. At scale, they teach people what to notice, imitate, desire, and believe.
If junior tasks disappear, so does the pathway by which people become senior, skilled, trusted, and economically independent.
The simplest test for aligned technology is whether the user leaves with more ability to choose, act, and understand.
A product-manager playbook for building technology that protects liberty, strengthens agency, and refuses captive interfaces.
A policy framework for recommender systems: disclose the objective, return control to users, and make duty of care enforceable.
A single path through the full article library, from foundations to action.
The quiet civic feeling that ordinary choice has less force than it used to.
Economic liberty, information liberty, and inner liberty.
A product is code plus defaults, incentives, and a theory of the user.
The human ground to protect is embodiment, relationship, judgment, and meaning.
What happens when tools begin to substitute the person.
AI may not give institutions the same grace of time.
Exponential technology punishes reactive habits.
Attention capture becomes intelligence at scale.
A market story and a power story.
Machine capability crossed into territory once assumed uniquely human.
The real question is who directs attention while people are there.
Powerful systems shape children without relationship or responsibility.
Recommendation systems teach people what to notice, imitate, desire, and believe.
Adolescent anxiety and depression rose as smartphone platforms became childhood infrastructure.
If junior tasks disappear, so does the pathway to responsibility.
A society can become richer while becoming less whole.
Engagement systems weaken the virtues that help people stop.
Humility matters when human power is being amplified.
A people who cannot govern appetite and attention will struggle to govern a republic.
The user should leave with more ability to choose, act, and understand.
The application decides whether capability becomes extraction or genuine help.
The same technology can extract from people or genuinely help them.
Users need to see and choose the frameworks shaping outputs.
Moral claims must become incentives, defaults, standards, and enforcement.
A product-manager playbook for technology that protects liberty.
The goal is a stronger child, formed by people and practices that love them.
Product choices are not morally empty.
Capital shapes timelines, incentives, governance, and products.
The AI moment is also a formation crisis.
The deeper issue is agency: understand, refuse, appeal, and control.
Disclosure, user control, and duty of care for recommender systems.