Slides from a guest presentation in the Digital Civil Society +1 Lecture Series (COMM 230X), Stanford University, May 26, 2020.
ABSTRACT: In a recent paper in the new ACM journal Digital Government: Research and Practice (DGOV), John Gastil and I make a case for the creation of a Corporation for Public Software — or, implicitly, a plurality of such corporations created by different governments. Modeled on the Corporation for Public Broadcasting in the U.S., this entity would foster the creation of new digital technology by providing a stable source of funding to nonprofit technologists, interest groups, civic organizations, government, researchers, private companies, and the public. Funded entities would produce and maintain software infrastructure for public benefit. In this talk, I will address the current state of efforts to address widely acknowledged deficiencies of commercial platforms for democratic communication, as well as the institutional relationships between government, private organizations, and the public that appear most suited to creating and sustaining public interest software.
BIO: Todd Davies is the associate director and a lecturer in the Symbolic Systems Program, and a researcher at the Center for the Study of Language and Information, at Stanford University. He holds a Ph.D. in cognitive psychology, an M.S. in data analysis and statistical computing, and a B.S. in statistics, all from Stanford. He has also served as a computer scientist at the Artificial Intelligence Center, SRI International, assistant professor of psychology at Koç University in Istanbul, and, most recently, faculty in residence at the Stanford Bing Overseas Studies Program in Oxford and a visiting fellow at Brasenose College. His research focuses on group deliberation, technology and methods for social decision making, and information policy.
VIDEO: https://youtu.be/SSq_nWsZy2A
PAPER FOR BACKGROUND: https://doi.org/10.1145/3342194
Unlocking the Power of ChatGPT and AI in Testing - A Real-World Look, present...
What Is Needed to Build and Sustain Public Software for Government and Civil Society?
1. May 26, 2020
What Is Needed to Build and Sustain Public
Software for Government and Civil Society?
TODD DAVIES
DIGITAL CIVIL SOCIETY +1 LECTURE SERIES
5. Key Findings
▪ Residents’ ability to participate was hampered by the requirement to
attend face to face meetings.
▪ Key voices were missing, such as when an open decision-making
meeting conflicted with an event focused on housing.
▪ These and other constraints on participation undermined the perceived
legitimacy of the initiative and altered its priorities.
▪ Difficult to make decisions in a timely way in face-to-face meetings.
▪ Other problems: Communication between meetings and between
subgroups, making information available during and between meetings,
and issues related to decision procedures and transparency
7. Deme Design Principles
Supportiveness. The platform should support the group overall, so that there is
either an improvement or no decline in the ability of the group to meet the needs
of its members or stakeholders.
Comprehensiveness. The platform should allow the group to accomplish, in an
online environment, all of the usual deliberative tasks associated with face-to-
face meetings.
Participation. The platform should maximize the number of desired participants
in the group’s deliberations, and minimize barriers to their participation.
Quality. The platform should facilitate a subjective quality of interaction and
decision making that meets or exceeds what the group achieves in face-to-face
meetings.
8. Online Deliberation: Experiential Findings
People still have limited time and attention -- OD competes with many other demands
of digital technology as well as cognitive limitations
Importance of salience
▪ Tools will not be used if they don't fit into people's lives naturally
▪ Notifications needed where our attention is drawn (e.g. email notifs to draw
participants to a Web forum)
Motivation
▪ Deliberation on its own is not motivating for most people
▪ Many sites became ghost towns
Social media designed for other purposes (meeting people, sharing videos of cats)
were more successful at generating deliberation than sites specifically designed for it
9. Online Deliberation: Achievements and Failures
+ Small group deliberation using general tools (email, video conferencing, etc.)
works in many cases
+ Social media has helped people find each other and be exposed to different ideas
(sometimes)
- Deliberation is difficult to achieve at scale
- Democratic culture seems to be in decline, rather than rising
- Commercial social media platforms (e.g. Facebook, Twitter) have been widely
criticized for undermining democracy -- promoting demagoguery and xenophobia
– despite their earlier roles in facilitating social movements like the Arab Spring,
Black Lives Matter, and Me Too
10. Why are the main commercial
platforms deficient in achieving
democratic outcomes?
11. Commercial social media (CSM) platforms are…
Not designed for democracy
Designed for obtaining attention-based revenue
12. CSM favor attention over deliberation
Click bait
Dopamine reward (e.g. Likes)
Self-referring content
Fear
Anger
13. CSM favor the interests of advertisers over users
Designs suited to paid promotion
Surreptitious data collection and retention
Predictive algorithms and microtargeting
Censorship in the interests of advertisers
14. CSM favor corporate over user/democratic control
Forced logins
Profits to the company and its investors
Monopoly interest – keep users within the platform
Company control of content (e.g. content moderation)
Non-customizable Terms of Service
15. CSM favor deep- over shallow-pocketed advertisers
Wealthier over poorer people
Larger corporations over smaller ones
Governments over citizens and social movements
Political clients over their opponents
17. Proposed ways to create better online deliberation
Government-built platforms
• administrative protectiveness, lack of dynamism
Nonprofit platforms
• lack of coordination, silos
Free/open source software
• lack of funding, end-user experience often deficient
Regulating commercial platforms (e.g. no political ads,
transparency in microtargeting, etc.)
• regulatory capture, structures remain in place
Antitrust enforcement
• very difficult to win, limited goals
18. Our proposal: A Corporation for Public Software (CPS)
A funding agency modeled on the U.S. Corporation for
Public Broadcasting
Funded by Congress
Governing board responsive to democratic change, with
built-in balance
Leveraged with other funding sources (other
governments, foundations, and small donors)
19. Advantages to the CPS model
A stable source of funding (if not a source of stable
funding)
Control over design and content separate from funding
Multiple sources of funding – not dependent on just one
An imprimatur, serving as a coordinating, focal signal