MSCD thesis repository. Looser spillage on TiddlyWiki.
- What does tracing the flow of data involved in CAPTCHA work reveal?
- How does a CAPTCHA organize different human computation markets; how do they interact?
- How do the aesthetic and affective experience of CAPTCHA work predispose users towards groveling?
The focus of this thesis is on the nature of CAPTCHA as a human computation (crowdsourcing) technology. Older paradigms of CAPTCHA work assume some intrinsic difference between machine and human perception. The current predominant (tracking-based) schemes do away with this and instead evaluate users in relation to other users’ performance: the metaphysical question of “are you a human” is set aside in favor of a regime of data collection that combines browser telemetry and user action (e.g. cursor movement) analysis to determine one’s validity as a user.
3 axes: labor, normativity, affect
Latest: scope reconsideration
Latest: Extension exists as v0 MVP; scraper in development.
Started as a personal phenomenological diary of CAPTCHA encounters (Glanceback-inspired) that has since morphed into a data collection instrument. The browser extension captures CAPTCHA imagery and metadata during the user’s everyday browsing. The scraper automates mass collection of CAPTCHA imagery from known sites. Both now primarily serve RP3.
Pending; upstream dependent on scraper producing corpus.
A theoretical and visual anatomy of CAPTCHAs as both human interactive proofs and human computation systems, built from the corpus generated by RP2. Combines critical STS reading with direct embodied research through making.
High-level logistics discussed; timeline is late-semester (mid March–April)
A Fluxus-inspired happening that invites people to collectively reflect on their relationship with CAPTCHAs.