The End Of Knowledge - Vault 4/82 - AI For Good - 2023 - Imagining the future of work in an AI-driven world

graph LR classDef intelligent fill:#f9d4d4, font-weight:bold, font-size:14px classDef economic fill:#d4f9d4, font-weight:bold, font-size:14px classDef future fill:#d4d4f9, font-weight:bold, font-size:14px classDef human_roles fill:#f9f9d4, font-weight:bold, font-size:14px classDef ecological fill:#f9d4f9, font-weight:bold, font-size:14px classDef envisioning fill:#d4f9f9, font-weight:bold, font-size:14px A[Imagining the future
of work in
an AI-driven world] --> B[Intelligent machines: achieving
objectives, creating general AI. 1] A --> C[General AI: lift living
standards, increase GDP. 2] A --> D[AI enables better healthcare,
education, progress. 3] A --> E[Key issue: human roles
in AI-driven world. 4] A --> F[Typical solutions: retraining,
empowering humans incomplete. 5] B --> G[Productivity gains raise labor
demand and wages. 6] B --> H[Thought experiment: AI works
free, humans lose jobs. 7] B --> I[Employment follows inverted U
as technology advances. 8] B --> J[Impacts depend on demand
elasticity, not substitution. 9] B --> K[Focus AI on unmet
needs, grow jobs. 10] C --> L[Long-term: routine labor
automated, find human roles. 11] C --> M[Keynes: challenge of
leisure time well spent. 12] C --> N[Marx: alienated vs unalienated
self-realizing work. 13] C --> O[Optimistic: AI frees humans
for self-realizing activities. 14] C --> P[Probable future: AI causes
unemployment, inequality, exclusion. 15] D --> Q[Expanded useless class: social
implosion, confrontation. 16] D --> R[AI has ecological costs
in energy, minerals. 17] D --> S[Scarcity of minerals may
constrain AI growth. 18] D --> T[Wars over AI-critical materials:
new oil. 19] D --> U[Urgent to control AI
to avoid disaster. 20] E --> V[Find AI applications at start
of job growth. 21] E --> W[Long-term: human roles in
care, creativity. 22] E --> X[Value in care roles needs
advances in psychology. 23] E --> Y[Imagining positive AI future:
challenge assumptions. 24] E --> Z[Critical skills: creativity, critical
thinking, community. 25] F --> AA[Shift to regenerative economic
paradigm. 26] F --> AB[Deep sea, space mining:
ecological, geopolitical risks. 27] F --> AC[Regain public agency over
AI development. 28] F --> AD[Economic system: provide common
goods, not just profits. 29] F --> AE[Cultural shift: from having
to being. 30] G --> AF[Competition: envision AIs
future in short videos. 31] G --> AG[Videos lacked texture of
experiencing AI future. 32] G --> AH[Groups brainstormed AI
work scenarios. 33] G --> AI[Envisioned futures: global collaboration,
UBI, personalized care. 34] G --> AJ[AI struggled imagining positive
futures outside current paradigms. 35] H --> AK[Collective imagination needed for
non-dystopian AI futures. 36] H --> AL[Translating visions to policy
change is difficult. 37] H --> AM[Challenge core assumptions, involve
youth, diversity. 38] H --> AN[AI limited in envisioning
future, lacks human experience. 39] H --> AO[Key themes: unmet needs,
collaboration, leisure time. 40] I --> AP[Emerging roles: life architects
for fulfilling lives. 41] I --> AQ[More youth inclusion in
envisioning AI future. 42] I --> AR[Privileged enable diverse voices,
think intergenerationally. 43] I --> AS[Engage public: relatable,
entertaining participation. 44] I --> AT[Inspire participation: connect issues
to lives. 45] J --> AU[Tangible AI impacts: engage experts,
policymakers, public. 46] J --> AV[Envision best future self before
AIs role. 47] J --> AW[Digital divide, uneven AI
skills access barriers. 48] J --> AX[Incentivize public participation,
enable real say. 49] J --> AY[Military struggles to get
top AI talent. 50] K --> AZ[AI community open, corporate
labs publish results. 51] K --> BA[Big tech democratizes AI
usage for non-coders. 52] K --> BB[Journey to envision positive
AI futures begun. 53] K --> BC[Examples translating visions
to policy exist. 54] K --> BD[AI limited in imagining,
collaborative envisioning key. 55] L --> BE[Accessible AI education lowers
participation barriers. 56] L --> BF[Diversity, youth, Global South
crucial for envisioning. 57] L --> BG[Tap personal aspirations before
AIs role. 58] L --> BH[AI development open, collaborative,
military lags. 59] L --> BI[Iterative imagination, connecting aspirations
to policy. 60] class A intelligent class B,G economic class C,D,E,F future class H,I,J human_roles class K,L,M ecological class N,O,P,Q,R,S,T envisioning class U,V,W,X,Y,Z envisioning class AA,AB,AC,AD,AE envisioning class AF,AG,AH,AI,AJ envisioning class AK,AL,AM,AN,AO envisioning class AP,AQ,AR,AS,AT envisioning class AU,AV,AW,AX,AY envisioning class AZ,BA,BB,BC,BD envisioning class BE,BF,BG,BH,BI envisioning

Resume:

1.- AI is making intelligent machines that act to achieve objectives, with the goal of creating general purpose AI exceeding human capabilities.

2.- If we had general purpose AI, we could lift everyone's living standards substantially, increasing global GDP by over 1000%.

3.- AI could enable better healthcare, personalized education, faster scientific progress, and potentially better politics for an improved civilization.

4.- The key issue is what will humans do if AI is delivering this improved civilization - an idea contemplated since Aristotle.

5.- Typical solutions like retraining everyone as data scientists or having AI empower rather than replace humans are incomplete.

6.- Economic principle that productivity gains raise labor demand and wages denies technological unemployment, but AI substitution could change this.

7.- Thought experiment - if AI copies worked for free, humans would lose jobs. Technology can increase or decrease employment based on demand.

8.- In many industries, employment follows an inverted U as technology advances - first increasing, then decreasing as demand is saturated.

9.- Impacts depend on elasticity of demand, not just if technology complements or substitutes. AI effects will be felt across the economy.

10.- Focus AI on unmet needs - delivering healthcare/education in poor countries, cargo inspection, city cleaning - to grow rather than displace jobs.

11.- Long-term, routine physical and mental human labor will likely be automated. We must determine meaningful human roles to avoid dystopia.

12.- Keynes foresaw the challenge of occupying leisure time well as technology reduced work. Veblen noted some would aggregate capital to avoid work.

13.- Marx distinguished alienated work from unalienated self-realizing work. AI may enable the latter if economic structures are transformed.

14.- Most optimistic scenario - AI does alienating work, freeing humans for self-realizing activities. But this requires decoupling survival from work.

15.- Probable future without change - AI causes mass unemployment, inequality, exclusion, poverty, lack of purpose especially for the young.

16.- Expanded useless class could lead to social implosion and geopolitical confrontation. Retraining isn't enough - economic transformation is needed.

17.- AI also has ecological costs in energy and rare mineral use. Projected needs outstrip known supplies. Geopolitical risks loom.

18.- Scarcity of needed rare earths may constrain projected AI growth unless new frontiers in space/oceans are exploited, risking environmental damage.

19.- Next wars may be fought over AI-critical materials which are new 'oil', with production highly concentrated in China currently.

20.- Urgent to control AI deployment to avoid combined ecological and social disaster. Channel it to serve rather than sacrifice humanity.

21.- Find AI applications at start of employment growth curve - meeting unmet needs in less developed countries, inspections, city maintenance.

22.- Long-term, routine physical and mental labor will be automated. Higher value human roles may be in interpersonal care and creativity.

23.- But delivering value in interpersonal roles requires advances in human sciences to understand psychology, not just technical AI capability.

24.- Imagining a positive future with AI is hard but vital. Steps: a) Challenge assumptions, b) Involve diverse voices, youth c) Translate ideas to policy

25.- Critical skills for an AI future: creativity, critical thinking, sense of community and belonging to collective.

26.- Shifting to a regenerative economic paradigm over extractive one is essential for sustainable AI deployment and stable transition.

27.- Deep sea and space mining of rare minerals for AI needs could cause ecological and geopolitical disasters. Governance is urgently needed.

28.- Regaining public agency over AI development is an existential necessity. Redefine relationship to work, consumption and each other.

29.- Economic system must move from maximizing shareholder value to providing common goods. Relationship with environment must turn regenerative.

30.- Cultural shift from having to being is needed, especially in richer nations. But this ambitious transformation is necessary to avoid dystopia.

31.- Competition had people envision the future of work with AI in short videos. Themes: AI aiding sustainability, creativity, collaboration.

32.- But videos lacked texture of experiencing an AI future directly. Breakout groups aimed to create more tangible visions using AI tools.

33.- Groups brainstormed scenarios of AI transforming work - optimistic and concerning. AI struggled to imagine non-work roles and utopian settings.

34.- Envisioned futures: AI enabling global virtual collaboration, machines doing dangerous tasks, personalized AI medical/mental care, universal basic income.

35.- But AI had difficulty imagining positive futures outside current economic paradigms. Fundamental assumptions need challenging to envision real alternatives.

36.- Iterative and collective imagination is needed to envision non-dystopian AI futures. Single Visionaries are insufficient. Younger perspectives are vital.

37.- Translating imaginative visions to policy change is difficult but groups are attempting it, such as on SME financing and creative industries.

38.- Challenging core assumptions is difficult but needed to move beyond incremental changes. Youth and experiential diversity is required in envisioning.

39.- AI is limited in helping envision the future as it lacks human experience. In-person collaborative imagining is most productive format.

40.- Key themes in positive visions: using AI to meet unmet needs, enabling human collaboration, providing more leisure time, social engagement.

41.- Emerging roles: "Life architects" helping shape fulfilling individual lives. But only 0.01% have such roles in some sci-fi utopian visions.

42.- More people, especially youth, need inclusion in envisioning the future. Their alienation with current trajectories must be addressed.

43.- Privileged have agency and responsibility to enable inclusion of diverse voices in future imagination. Think intergenerationally as the "now" generation.

44.- Engaging public requires meeting them where they are at and relating to their concerns. Make it welcoming, unthreatening and entertaining.

45.- Giving inspiring and relatable reasons to participate is key, showing how issues connect to their lives. Provide entry points and examples.

46.- Bringing together experts, policy makers, companies and general public to make AI impacts tangible helps engagement. Make it accessible.

47.- Having people envision their best future self before imagining AI's role connects it to personal aspirations. Provides context for participation.

48.- Digital divide, uneven access to AI skills are barriers to inclusion. Providing accessible AI education in multiple languages helps.

49.- Meaningful paths to impact and agency incentivize public participation. Direct democracy mechanisms can enable having a real say.

50.- Military struggles to get top AI talent due to uncompetitive pay vs tech giants. More concerning than any classified military AI.

51.- AI community still quite open, with major corporate labs quickly publishing results. But some signs of more internal development emerging.

52.- Big tech also launching platforms to democratize AI usage for non-coders. But expanding access is only one piece.

53.- The journey to collectively envision positive AI futures has just begun. Experiential learnings from workshops will shape the process.

54.- There are already some good examples of translating future visions into policy around SME financing, creative industry support.

55.- AI is still limited in imagining futures, constrained by training data. Collaborative human envisioning is key to moving beyond standard narratives.

56.- More accessible AI education efforts like Finland's "Elements of AI" are lowering barriers to public participation in shaping AI futures.

57.- Experiential diversity, youth voices, and inclusion of Global South are crucial for the collective envisioning process, not just experts.

58.- Envisioning exercises should tap personal aspirations first before imagining AI's role - provides more resonant entry points to contribute.

59.- Despite some concerns, AI development is still largely open, fast-paced and collaborative. Classified military AI lags behind commercial efforts.

60.- The journey has only begun - iterative collective imagination, connecting aspirations to policy impacts, is key to realizing positive AI futures.

Knowledge Vault built byDavid Vivancos 2024