Of Hackers and Yearners: Constructionist Learning's Debt to AI and Cybernetics

• In Event: Artificial Intelligence and Cognition as Social Praxis

Abstract

The learning theory “constructionism,” pioneered by late MIT professor Seymour Papert, has enjoyed some fifty years of popularity in the technology world despite evidence that its most unique and provocative claims do not work. The computer-enabled learning that constructionism promotes shares with situated learning a deep distrust of the institution of school and focuses on providing tools that inspire children to joyfully teach themselves foundational concepts in mathematics and computer science. It diverges, however, in method: constructionism focuses on computers and software like the Logo programming language (or, more recently, Scratch) as the tools of inspiration, and tends to neglect the contextual factors that preoccupy situated learning scholars.

Through an analysis of primary and secondary sources, this paper shows that constructionism’s preoccupation with computer-based learning instead comes from MIT’s artificial intelligence labs of the 1970s and the ‘hacker’ community they fostered. This work traces this history and explores the reasons behind the continued popularity of both constructionism and the ethos of this community, linking both to the cybernetic worldview that humans are, as AI pioneer and MIT professor Marvin Minsky liked to put it, “meat machines,” and that “thinking like a machine” is a powerfully liberating experience. This utopian worldview has had an enduring influence on the technology world more broadly, allowing constructionism to ride along on its coattails. In deconstructing constructionist learning, this paper provides a framework for understanding the ideological histories of other inexplicably tenacious technology projects – and for contextualizing contemporary promises made about artificial intelligence.

Author

• Morgan G. Ames, CSTMS, UC Berkeley