The platformization of everything: From the end of the Like button to AI infrastructure in space

Summary

In this anniversary essay, Anne Helmond revisits the concept of “platformization” a decade after she coined it, arguing it has entered a new AI-centered phase while simultaneously suffering conceptual erosion through conflation with “digitization.” Working through two illustrative cases — Meta’s February 2026 retirement of the Like and Comment plugins, and Google’s Project Suncatcher initiative to put AI compute infrastructure in orbit — she traces how platforms have shifted from extending tendrils across the open web to building integrated AI ecosystems that no longer depend on social-graph signals. To restore analytical precision, she proposes redefining platformization as a platform-specific form of Manovich’s “transcoding”: the situated, collaborative process through which practices, sectors, and domains are made “platform-ready.”

Key Contributions

• Reclaims and refines “platformization” by recasting it as platform-specific transcoding, distinct from generic digitization.
• Offers a periodization marking the transition from the social-graph/open-web era to the AI-platform era (“Big AI”).
• Foregrounds “actually existing platformization” — situated, uneven, and dependent on third-party collaboration (webmasters, app developers, sectoral partners).
• Connects platformization to adjacent critical frameworks: sphere transgression, intellectual monopolization, Big Techification, and hyperscalers.
• Sets a forward research agenda spanning Big Tech–state relations, military and energy platformization, super apps, and orbital AI infrastructures.

Methods

A conceptual/theoretical essay revisiting the author’s earlier framework, anchored in two illustrative empirical cases (the Like button’s retirement and Project Suncatcher). Helmond synthesizes adjacent frameworks (Sharon & Gellert; Rikap; Hendrikse et al.; Manovich) and draws on her forthcoming co-authored Platforms: A Critical Introduction.

Findings

• Meta has decoupled from the social-graph paradigm: in-app behavioral data, ad network signals, third-party app integrations, and AI inference now substitute for Like-button data harvested from the open web.
• The Like button’s retirement signals a turn toward TikTok-style algorithmic recommendation and AI personal assistants over social signals.
• Project Suncatcher (prototype launches planned for 2027) exemplifies Big Tech’s move from “platform earth” to space, driven by terrestrial limits on energy, water, and data-center capacity.
• Platformization unfolds unevenly across regions and sectors — Chinese society, Dutch education, food delivery, housing, military — mediated by local intermediaries.
• New entrants (Nvidia, Anduril, Palantir, Shield AI) are adopting platform models, while incumbents reorient from platform-as-infrastructure toward integrated AI ecosystems.
• “Platformization” has eroded analytically through overuse; recovering its specificity requires distinguishing platform-driven transcoding from broader digitization.

Connections

This essay sits at the conceptual core of critical platform studies and resonates strongly with Rieder2025-ju and Rieder2026-pp on platform infrastructure and governance, as well as with Bruns2026-yv on the shifting terrain of platform research. Its diagnosis of the post-API, post-social-graph moment — where access to platform signals collapses just as AI inference takes over — speaks directly to data-access debates explored in Ohme2026-nv, Murtfeldt2025-wu, and Davis-style work on researcher access such as Freelon2024-sc. The argument about “Big AI” and new sectoral entrants also dovetails with infrastructural and governance concerns raised in Tornberg2026-lc and Munger2025-cz.

Podcast

A research-radio episode discusses this paper: Listen