Open a marketplace. Search. Scroll endless catalogs. Skip ads. Ignore “recommended” products. Compare listings that look almost identical.
Eventually fatigue wins and we click something — not because we’re sure it’s the best option, but because we want to stop spending time on it.
It’s strange that we’ve normalized this. Buying online often means navigating noise: catalogs, ads, rankings, and persuasion systems competing for attention.
What I keep wondering is this:
When personal AI agents become common, what prevents them from doing exactly the same thing?
If the interface to commerce remains “browse catalogs and search results,” then agents will simply automate the same inefficient process — crawling listings, parsing ads, and navigating ranking systems just to reach something the buyer already knew they wanted.
Maybe the real missing layer isn’t better search or better recommendations.
Maybe it’s a way to express structured intent instead of browsing.
Curious if others think catalog-based commerce is the wrong interface for an AI-driven world.
When you can't imagine, you're stuck with searching, searching that comes in many forms, such as scanning catalogs, walking aisles, or letting someone else do it altogether.
If you can imagine what you want and describe it with sufficient language, you can have someone else (or something) go and fetch it for you, but you're still stuck with the problem of not getting _exactly_ what you want, so again, you must parse the results... or get stuck with whatever is returned first.
Ad the end of the day, it's a question of how much you care about getting the thing you want vs getting something that is close enough, and the answer is in how much effort you, personally, want to spend.
With the book in your hand you can flip through over 100 pages per minute, that puts the online version to absolute shame.