The coming product retention crisis
AI will collapse consumer app retention because apps will become ephemeral content—built and consumed like memes—while only "adversarial creativity" (novelty that works against training data) wins attention in an infinite supply world.
Read Original Summary used for search
TLDR
• AI crushes work drudgery (verifiable, pattern-based) but generates slop for consumer apps (novelty-seeking, adversarial creativity required)
• Apps will be "poasted" like content—quantity beats quality, retention metrics collapse, D30 >20% becomes a dream
• "Adversarial creativity" = being fresh is literally adversarial against the model's training data; what's novel is what ISN'T in the dataset
• Future is PvP for attention: win by staying ahead of the cultural OODA loop or do expensive in-person signaling with "hot people"
• One AI tell (emdashes, "load bearing") breaks parasocial trust instantly—you can't fake authenticity with copy-paste AI
In Detail
The piece argues that AI's workplace success will paradoxically create a consumer product crisis. AI excels at verifiable, pattern-based work (coding, forms, boilerplate) but fails at consumer apps because humans are "novelty-seeking monkeys" who detect and reject AI slop. The key framework is verifiable vs adversarial: AI automates verifiable outcomes brilliantly, but consumer products require "adversarial creativity"—being novel in ways that actively work against what's in the training data. This is why AI-generated marketing emails and social content feel stale; they pattern-match to the past while humans crave what's culturally fresh.
As AI makes app creation as easy as editing a YouTube video, apps will become ephemeral content. We'll see "app poasting"—quantity over quality, riding viral spikes instead of building retention. Traditional metrics like D30 retention >20% will collapse because apps will be consumed like memes: huge spike, then move to the next thing. The question shifts from "how sticky is this app?" to "how many apps hit how many eyeballs?" Success becomes about the system for generating hits and the brand/IP built over time, not individual app retention.
The competitive landscape becomes zero-sum PvP for attention in a world of infinite supply. Winning requires staying slightly ahead of the cultural OODA loop—fast enough to be fresh, slow enough to not alienate. The author predicts only four job types survive: product generalists (high-velocity tool users), infrastructure people (stitching together the explosion of stuff), "adults in the room" (governors on acceleration), and "hot people" (forward-deployed in-person roles doing expensive signaling). The barbell strategy: either win the digital attention game or do costly in-person work to prove customers are special. The danger: one AI tell breaks parasocial trust instantly—you can't automate authenticity.