AI summary by TwinMind
6 big ethical questions about the future of AI | Genevieve Bell
TED
14m 48s
Introduction: AI in Everyday Systems
- AI already present in lift systems at 1 Bligh Street, Sydney
- Proto-artificial intelligence algorithms anticipate user needs
- System rearranges carriages before button pressed
- Manages energy efficiency and traffic flow
- AI ubiquitous in built environment, not just science fiction robots
- 200+ years industrialization suggests AI will achieve systems-level scale easily
- Traditional narrative: find technology → achieve scale → revolution follows
- Scaling technology requires more than just technological change
3Ai Institute Foundation and Mission
- Founded at Australian National University, September 2017
- Mission: establish new branch of engineering for AI scaling safely, sustainably, responsibly
- Three-pronged approach:
- Teaching into existence through experimental education
- Researching at diverse locations: Shakespeare's birthplace, Great Barrier Reef, Australia's largest autonomous mine
- Theorizing through cybernetic systems complexity analysis
- Goal: create next generation critical thinkers and doers
- Focus on AI's many pasts/stories, collaborative work, question framing vs problem solving
- Actively decolonizing imaginations, building inclusive curriculum/pedagogy
Historical Grounding: Brewarrina Fish Weirs
- December 2018 visit to Brewarrina, NSW-Queensland border
- Aboriginal meeting place for ceremonies, gatherings
- Fish weir system on Barwon River:
- 1.8 kilometers stone walls shaped like fishnets
- "Us" pointing downstream for fish trapping
- Variable height walls for different water levels
- Fish holding pens with cool, clear running water storage
- System integrated technical, cultural, ecological knowledge
- Age: possibly 40,000 years old (archaeological estimates)
- Last recorded use: 1910s
- Remarkable longevity and scale
- Photo displayed at Institute as inspiration/reminder
- Legacy for building sustained systems in place with 60,000+ year continuous occupation
Six Framework Questions for AI Systems
1. Autonomy
- Is system autonomous (can act without being told)?
- Lift example: autonomous within constraints (up/down only, can't leave building)
2. Agency
- Does system have agency?
- Current lifts: red keyslot for emergency services override
- AI systems: where does override key live (physical/digital)?
- Who gets access? How is override manifested to users?
3. Assurance
- Components: safety, security, trust, risk, liability, manageability, explicability, ethics, public policy, law, regulation
- How to communicate system safety and functionality to users?
4. Interfaces
- How will humans interact with AI-driven systems?
- Will we talk to them?
- Challenge: familiar technologies (lifts, cars, electrical grid, traffic lights, home devices) behaving differently
5. Indicators
- Traditional industrial measures: productivity, efficiency
- 21st century expansion: sustainability, safety, responsibility
- Who judges these metrics?
- User need to understand regulation, management, construction
6. Intentionality
- What is system designed to do?
- Who decided this was good idea?
- What world is system building?
- How is that world imagined?
- Relationship to current world?
- Who participates in conversation?
Systems Thinking Framework: Cybernetics Heritage
- Influenced by 1940s work
- 1944: Gregory Bateson, Margaret Mead, Norbert Wiener convened conversations
- Macy Conferences on Cybernetics (1946-1953, 10 conferences total)
- Wiener's definition: "develop language and techniques for control and communication in advanced computing technologies"
- Holistic approach: humans, computers, broader ecological world as integrated system
- Conference focus: mind function, intelligence, learning, technology's future role
- Often forgotten in AI discussions
- Core to Institute's work: system accommodating culture, technology, environment
Institute Team and Diversity
- Staff includes:
- Anthropologists
- Systems and environmental engineers
- Computer scientists
- Nuclear physicist
- Award-winning photojournalist
- Policy and standards expert
- Diversity requires constant conversation facilitation
- Willingness to dwell with conflict
- Commitment to bringing others on journey
Education Program
- First master's program launched 2018
- Two cohorts master's students, one cohort PhD students
- Student origins: Australia, New Zealand, Nigeria, Nepal, Mexico, India, United States
- Age range: 23-60 years
- Backgrounds: maths, music, policy, performance, systems, standards, architecture, arts
- Pre-Institute experience: company runners, government workers, army service, high school teachers, arts organization managers
- Adventurous, committed to each other and building something new
Collective Approach Philosophy
- Rejects "lone inventor" Silicon Valley narrative
- Reality: collections of people with shared purpose change world
- Emphasis on starting where you stand
- Acknowledgment of traditional land owners: Ngunnawal and Ngambri people
- Land never ceded, always sacred
- Respects to elders past and present
- Australia continuously occupied 60,000+ years
- Aboriginal people built worlds, social systems, technologies, management systems
- Privilege and responsibility from this history
- Rich resources, worldviews, knowledge should inform all work
- Future and scale achieved collectively
- "Humans in it together" enables responsible, safe, sustainable outcomes
