Download this Paper Open PDF in Browser

Machine Learning, Social Learning and the Governance of Self-Driving Cars

39 Pages Posted: 20 Mar 2017 Last revised: 22 Mar 2017

Jack Stilgoe

University College London

Date Written: March 19, 2017

Abstract

Self-driving cars, a quintessentially ‘smart’ technology, are not born smart. In the algorithms that control their movements and the connections they make with their surroundings, they are learning as they emerge, in organised and haphazard ways. They are a test of the powers of machine learning as well as an important test case for social learning in technology governance. Society is learning about the technology as the technology is learning about society.

In this paper, I reframe responsible innovation as social experiment, with the key question being ‘who learns what?’ Focussing on the successes and failures of social learning around a much-publicised crash in 2016, I argue that trajectories and rhetorics of machine learning in transport pose a substantial governance challenge. ‘Self-driving’ or ‘autonomous’ cars are misnamed. As with other technologies, they are shaped by assumptions about social needs, solvable problems and economic opportunities. Governing self-driving cars in the public interest means challenging this discourse of autonomy and appreciating the ways in which self-driving cars will be entangled in their environments. I conclude with some options for governance that should enable greater social learning.

Suggested Citation

Stilgoe, Jack, Machine Learning, Social Learning and the Governance of Self-Driving Cars (March 19, 2017). Available at SSRN: https://ssrn.com/abstract=2937316 or http://dx.doi.org/10.2139/ssrn.2937316

Jack Stilgoe (Contact Author)

University College London ( email )

Gower Street
London, WC1E 6BT
United Kingdom

Paper statistics

Downloads
215
Rank
119,337
Abstract Views
1,023