Updating to Remain the Same

In the AI Story

Chun's concept emerged as a diagnosis of the software industry's release cycle—the rhythm of updates, patches, new versions that users were compelled to adopt on pain of incompatibility, security risk, or simple obsolescence. The compulsion felt like consumer choice (you could refuse the update) but operated as structural necessity (refusing meant the tool stopped working with everything else). The freedom to update masked the obligation to update. By the time Chun formalized the concept in 2016, the pattern had extended from software to every domain of digital life: social platforms updating their interfaces, recommendation algorithms updating their models, professional tools updating their feature sets. Each update arrived with the promise of improvement; cumulatively they produced the condition of perpetual provisionality.

The AI moment accelerates this dynamic to an almost absurd degree. GPT-3.5 to GPT-4 in three months. Claude 3 to Claude 3.5 in six. Each threshold crossed rendered prior prompting strategies suboptimal, prior workflows outdated, prior investments in tool-mastery provisional. Segal documents this in The Orange Pill when he describes senior engineers whose decades of expertise remain valuable for architectural judgment but whose relationship with implementation tools is chronically destabilized. The engineer is updating to remain what he was—a capable builder whose expertise matters—but the updating itself prevents him from ever fully arriving at the stable competence the updating promises.

The updating paradox has a particularly cruel dimension when combined with AI's democratization of capability. The entry barrier falls—anyone can build—but the floor never stabilizes. The builder who spent months mastering Claude 3.5 finds Claude 4 requires different prompting strategies, produces different failure modes, offers different strengths. The learning never compounds into durable expertise because the tool keeps changing. The democratization is real; the stability that would allow the democratized capability to develop into genuine mastery is structurally withheld. The builder remains perpetually provisional, perpetually a beginner at the new version, perpetually updating to remain—at best—competent.

Chun's framework does not propose stopping the updates. That option is foreclosed by network effects, competitive pressure, and the genuine improvements each update delivers. The prescription, to the extent one exists, is to see the updating for what it is: not progress toward a destination but a treadmill, not development toward mastery but maintenance of provisional competence, not growth but what Hartmut Rosa would call dynamic stabilization—staying in place through constant movement. The builder who sees this clearly may, at minimum, stop mistaking the exhilaration of the new capability for the security of durable skill. The exhilaration is real; the security is architectural illusion.

Key Ideas

The update does not arrive; it perpetuates. Each new version promises improved capability while requiring re-learning that prevents the consolidation of expertise around any single version—producing permanent transition rather than punctuated stability.

Competence depreciates like capital. In a rapidly updating environment, the investment in learning a tool loses value as the tool changes, producing a condition of chronic skill-obsolescence that pre-AI workflows spread across decades but AI compresses into months.

Permanent precarity, not permanent possibility. The builder has access to unprecedented capability and an unstable relationship with that capability, because the capability is borrowed from tools that will be different tomorrow—rented competence, not owned skill.

The updating is the product. Platforms do not merely deliver updates as maintenance; they produce the compulsion to update as the mechanism of user retention, competitive differentiation, and continuous engagement with the brand.

Dynamic stabilization through movement. The builder updates not to change but to remain the same—to maintain professional relevance, competitive position, organizational value—a staying-in-place achieved only through continuous adaptation.