The End Of Knowledge - Vault 4/61 - AI For Good - 2021 - Exploring the promise of AI for civil society participation

graph LR classDef main fill:#f9f9f9, font-weight:bold, font-size:14px classDef hosts fill:#ffcc99, font-weight:bold, font-size:14px classDef democracy fill:#ccff99, font-weight:bold, font-size:14px classDef research fill:#99ccff, font-weight:bold, font-size:14px classDef voting fill:#ff99cc, font-weight:bold, font-size:14px classDef control fill:#ccccff, font-weight:bold, font-size:14px classDef political fill:#ffccff, font-weight:bold, font-size:14px classDef debate fill:#cc99ff, font-weight:bold, font-size:14px classDef data fill:#f9cc99, font-weight:bold, font-size:14px classDef discrimination fill:#99ccff, font-weight:bold, font-size:14px classDef transparency fill:#99ffcc, font-weight:bold, font-size:14px classDef blockchain fill:#ff99ff, font-weight:bold, font-size:14px classDef brazil fill:#ccffcc, font-weight:bold, font-size:14px classDef future fill:#99ff99, font-weight:bold, font-size:14px A[Exploring the promise
of AI for
civil society participation] --> B[Juan hosts
AI for
Good series. 1] A --> C[Jerome Duberry discusses
AI for
democracy. 2] A --> D[Research and AI] D --> E[Jerome researched tech
for civil
participation. 3] D --> F[Recent work: AIs promise
for
democracy. 4] D --> G[MIT project inspired
AIs democratic
impact. 5] A --> H[Voting and Control] H --> I[AI replacing human votes
seemed
disruptive. 6] H --> J[AI re-examines voting
importance beyond
efficiency. 7] H --> K[Digital twin voting
solves representation
issues. 8] H --> L[People control AI votes,
maintain
agency. 9] A --> M[Political Use and Debate] M --> N[AI used in politics
and
government. 10] M --> O[Debate: tech threatens democracy,
voting
tech. 11] A --> P[Data and Discrimination] P --> Q[Tech safeguards judicial process,
data
validity. 12] P --> R[Digital twins need
vast, private
data. 13] P --> S[Data available from brokers,
education
needed. 14] P --> T[UNESCO report: AI
discriminates against
women. 15] A --> U[Transparency and Blockchain] U --> V[Tech harms/benefits align
socioeconomically.
16] U --> W[Transparency needed between
citizen input,
policies. 17] U --> X[Blockchain/AI nudge sustainable
behaviors,
rewards. 18] A --> Y[Bias and Applications] Y --> Z[AI overcomes biases
by flagging
info. 19] Y --> AA[Brazil uses AI for
political reimbursement.
20] Y --> AB[AI tests urban planning,
analyzes sentiment.
21] Y --> AC[Future: AI tests policy
with digital
twins. 22] A --> AD[Future Vision] AD --> AE[Uncertain timeline for
AIs ideal
vision. 23] AD --> AF[Co-design AI
due to
uncertainty. 24] AD --> AG[AI for Good
explores AI
for SDGs. 25] class A,B,C hosts class D,E,F,G research class H,I,J,K,L voting class M,N,O political class P,Q,R,S,T data class U,V,W,X transparency class Y,Z,AA,AB,AC blockchain class AD,AE,AF,AG future

Resume:

1.- Juan from ITU hosts AI for Good Let's Talk series discussing AI, the human experience, and building a better world.

2.- Today's guest is Jerome Duberry from the Albert Hirschmann Center on Democracy discussing AI's promise for civil society participation.

3.- Jerome researched how digital technology fosters civil society participation in global environmental governance and biodiversity conservation.

4.- More recent work explores AI's promises for democracy, in addition to the well-researched challenges of AI for liberal democracy.

5.- MIT project inspired thinking about AI's future impact on democracy while there is still time to shape its development.

6.- Replacing human votes with AI digital twins seemed disruptive and impossible to colleagues at first.

7.- AI forces re-examining fundamental questions like what voting means and why it's important beyond just efficiency.

8.- Digital twin voting could enable everyone to participate, solving lack of representation for less wealthy and educated populations.

9.- People could unplug or correct their digital twin's vote, maintaining agency while letting AI vote on less critical decisions.

10.- AI is already used in politics and by governments. Ongoing research aims to raise awareness of this.

11.- Existential threat of technology to democracy debated. Implementing existing tech for voting questioned.

12.- Technology's role in safeguarding judicial process important. Ensuring data validity of digital twins a challenge.

13.- Vast data needed for everyone to have a digital twin. Privacy, ethics, consent issues remain.

14.- Wealth of data likely available from data brokers, platforms. Education needed so people understand data use implications.

15.- UNESCO report highlights AI discriminates, especially against women. Co-designing tech with citizens could help.

16.- Citizens harmed vs benefiting from tech often fall along same socioeconomic lines. Participatory action research could empower citizens.

17.- Lack of transparency linking citizen input to policy reports an issue. Co-designing AI with citizens could build trust.

18.- Blockchain and AI nudge citizens toward sustainable behaviors through incentives like cryptocurrency rewards for recycling.

19.- AI could help overcome cognitive biases by flagging additional information we may disregard due to confirmation bias.

20.- Brazil's congress successfully uses AI to verify politicians' reimbursement claims, changing behavior through consistent oversight.

21.- AI enables testing urban planning scenarios to optimize conditions. Analyzes public sentiment from civic tech platform comments.

22.- Future governments could rapidly test policy ideas with citizens' digital twins before implementing.

23.- Uncertainty remains on timeline for achieving ideal AI vision. Impact often overestimated short-term but underestimated long-term.

24.- Co-designing approach recommended given uncertainty. Progress being made step-by-step.

25.- AI for Good is a leading platform for exploring AI to achieve SDGs. Upcoming events announced.

Knowledge Vault built byDavid Vivancos 2024