I subscribe to a service that sends me coffee. Every two weeks, a bag arrives. I didn't choose the beans. An algorithm did, based on preferences I indicated once, years ago. I could intervene, the site lets me swap selections, but I haven't in eighteen months. The coffee is fine. The decision is handled. I don't think about it.
The conversation about agents replacing human choice often assumes the human objects. The person defending their preferences against optimizing machines. The consumer who values the act of selection. But many people experience choice as burden. They don't want to research coffee beans. They don't want to compare detergents. They don't want to decide which gas station, which brand of batteries, which flight among seventeen options departing within the same hour.
They want the outcome without the process.
The Weight of Options
Barry Schwartz wrote about the paradox of choice twenty years ago. More options should make us happier, more likely to find what we truly want. Instead, abundant choice produces anxiety, regret, and paralysis. We spend more time deciding. We feel worse about what we chose. We wonder about the options we didn't take.
The research findings have replicated in different contexts. Retirement plan participation drops when employees face too many fund options. Jam purchases decline when shoppers encounter too many varieties. The effect is robust: past some threshold, additional options subtract rather than add.
The consumer economy responded by adding options anyway. Product variety expanded. Service customization proliferated. The assumption that choice is always good proved more durable than the evidence against it.
Delegation to systems offers an exit. The coffee subscription means I never stand in a store aisle comparing roasts. The auto-replenishment means I never run out of detergent and never consciously purchase it either. Each delegation removes decisions from my day. The cognitive overhead shrinks.
What We're Really Tired Of
The exhaustion isn't just volume. It's the knowledge that much of what we're choosing between doesn't matter.
Detergents are largely equivalent. Airlines operating the same route provide similar service. Generic medications contain identical active ingredients. Many product categories differentiate on brand and marketing more than substance. The consumer who researches carefully often discovers that the choice didn't warrant the research.
The effort is misallocated. Hours spent comparing equivalent options are hours not spent on choices that actually matter. The exhaustion comes partly from the suspicion that the decision architecture was designed to extract attention, not to serve the person making decisions.
Delegating those decisions to agents isn't surrender. It's triage. Focus attention on choices that warrant it. Let systems handle the rest.
The Trust Transfer
Willingness to delegate depends on trust. I trust the coffee service to send acceptable coffee. I trust the auto-replenishment to reorder before I run out. If either failed repeatedly, I'd take the decisions back.
The trust isn't in the technology specifically. It's in the outcome. The system that produces acceptable outcomes earns continued delegation. The system that fails loses it. This is pragmatic rather than principled. I don't believe algorithms should make my decisions. I've observed that, for some decisions, they make acceptable ones with less effort than I'd invest.
Agent-mediated commerce will scale on this trust transfer. Early use cases will be low-stakes decisions where acceptable outcomes are common and failures are cheap. Household supplies. Routine travel. Commodity purchases. The agent that handles these reliably earns delegation of higher-stakes decisions. The agent that fails loses access to any decisions.
What This Reveals About Choice
The willingness to delegate suggests that choice was never as central to the consumer experience as marketing claimed.
Brand narratives assume the consumer who cares. Who researches. Who selects deliberately. Who feels loyalty and exercises preference. The entire apparatus of brand building, awareness, consideration, preference, loyalty, presumes a consumer engaged in the act of choosing.
Many consumers were never that engaged. They bought what was available, what was familiar, what required the least effort. Brand "loyalty" was often just habit. "Preference" was often just recognition. The engaged, deliberating consumer was a smaller segment than the marketing assumed, and that segment was exhausted.
Agent delegation makes visible what was always true: a large portion of consumer behavior was never really choosing. It was satisficing—finding acceptable options fast and moving on. Agents just do that more efficiently.
The Choices That Remain
Not all decisions will be delegated. The categories where choosing itself provides value will resist automation.
Gifts. The value is partly in the selection. An agent-chosen gift is a different object than a personally-chosen one, even if physically identical. The choice carries meaning. Delegation drains the meaning.
Identity purchases. The clothes I wear, the art on my walls, the books on my shelves, these express something about who I am. Delegating them to an agent would feel like outsourcing identity. Even if the agent chose well, the choice wouldn't be mine.
Experiences. A restaurant recommendation is welcome. An agent booking dinner without my input is presumptuous. The anticipation is part of the experience. The act of choosing is part of enjoying.
These categories share a feature: the process of choosing matters, not just the outcome. Delegation would save time but destroy value. The exhausted consumer wants relief from meaningless choices, not all choices.
The Question Underneath
If people are willing to stop choosing, what were we doing when we thought we were choosing?
The discomfort in this question is real. Consumer identity in market economies is bound up with choice. We are what we buy. Our preferences define us. The freedom to choose is linked to other freedoms. Surrendering choice, even mundane choice, feels like surrendering something important.
Maybe we overestimated how much choosing we were really doing. Maybe the choices that matter were always a small subset, buried under a mass of trivial decisions that felt like choices but functioned as chores. Maybe the freedom worth defending isn't the freedom to compare detergents but the freedom to decide when comparison matters.
The consumer who delegates commodity purchases to an agent and reserves attention for choices that warrant it may be exercising more meaningful choice than the consumer who dutifully researches every purchase. The exhausted chooser who stops choosing may be choosing something important: where to invest limited attention in a world designed to extract it.
This is part of a series on the robotic web fromTinyFish, which builds infrastructure for machine operation of the web.