1 of 23

AI Ecosystem�-draft-

Kilnam Chon

전길남

KAIST

2021.6.9rev6.9/6.20/6.25/6.26/6.30/7.15

2 of 23

Table of Contents

  • Introduction
  • Technology
  • Conferences
  • Organizations
  • International Collaboration
  • Interdisciplinary Collaboration
  • Users
  • Public Sectors
  • Commercial Sectors
  • Abuse
  • Risks
  • Issues and Remarks

References

Appendix

2

3 of 23

1. Introduction

  • Ecosystem: A complex network or interconnected system
  • Digital Ecosystem: Internet Ecosystem, AI Ecosystem,…
  • AI Ecosystem; early stage even though AI is as old as other digital technologies
  • AI: pervasive technology, and one of major technologies in the modern history
  • AI: Internet AI, Business AI, Perception AI and Autonomous AI
  • AI and Data: Symbiotic Relationship

3

4 of 23

2. Technology

Narrow AI vs General AI

Deep Learning, machine learning and AI

SuperIntelligence

4

5 of 23

3. Conferences

  • List of conferences on AI and AI governance is in Appendix A.
  • Very active and numerous now with mostly academic conferences
  • Many conferences are of large scale; several thousands or more participants
  • AI governance conferences are in early stage and of small scale.

5

6 of 23

4. Organizations

  • List of AI organizations and AI governance organizations are in Appendix B.
  • Many AI organizations; mostly academic.

6

7 of 23

5. International Collaboration

  • Very extensive international collaborations around the world.
  • USA is the focal point of the international collaboration.
  • Europe has extensive Europe-wide collaborations.
  • International collaboration is weak in the rest of the world: Africa, Asia, and Latin America.

7

8 of 23

6. Interdisciplinary Collaboration

  • Very extensive interdisciplinary collaboration due to nature of AI; AI itself is interdisciplinary, and AI applications can be found in almost all disciplines
  • Data science may be a special case with its symbiotic relationship with AI.
  • Others could be grouped; natural science, engineering, social science, arts, and business

8

9 of 23

7. Users

  • Users shall be the main stakeholder of AI ecosystem.
  • Almost all people in the world are users; one way or the other.
  • Need to pay special attention to handicapped; physiological, economical and geographical as well as gender.

9

10 of 23

8. Public Sectors

  • State government
  • Global organizations; UN,..
  • Others: local government, public institutions

10

11 of 23

9. Commercial Sectors

  • Becoming major area of commercial activities; Toward ~20% of the global GDP
  • Nature of “winner takes all”; both companies and countries
  • AI covers almost all commercial subsectors.

11

12 of 23

10. Abuse

  • Abuse is common in AI like many other digital technologies.
  • Abuse with AI technology could be very damaging; face recognition, censoring,…

12

13 of 23

11. Risk

  • AI would be one of the existential risks along nuclear technology [Cambridge].

13

14 of 23

12. Issues and Remarks

  • Flexibility and adaptability are very important for AI ecosystem to be healthy.
  • Evolution shall be taken into consideration.
  • Changing environments over time and regions need to be considered.
  • Interdisciplinary nature of AI needs to be taken into consideration.
  • Existential risk; how do we incorporate existential risk by AI in AI ecosystem?
  • Asimov’s Three Laws of Robotics to be included in AI ecosystem.

14

15 of 23

References

J. Ding, China AI Newsletter.

Kilnam Chon, Chapters 2 and 3, Asia Internet History, Fourth Decade, 2021.

Kilnam Chon, Chapter 3, Asia Internet History, Fifth Decade, 2022.

Center for Study on Existential Risks, Cambridge.

Future of Life Institute

Stephen Hawkins, Brief answers to big problems, 2018.

Kai-Fu Li, Internet Superpower, 2018.

Nvidia Blog, Deep learning, Machine learning, and AI, 2016.

Stuart Russell, Human Compatible, 2019.

15

16 of 23

Appendix

Appendix A1: List of AI Organizations

Appendix A2: List of AI Governance Organizations

Appendix B1: List of AI Conferences

Appendix B2: List of AI Governance Organizations

Appendix C: List of AI Courses

Appendix D: List of AI Principles

16

17 of 23

Appendix A1: List of AI Organizations

AAAI

ACM

AI Initiative�Alan Turing Institute, UK�Element AI�Future of Life Institute (FLI), USA�Human Centered AI, Stanford, USA�ISO/IEC JTC1 SC42, Standardization on AI�Leverhulme Center for Future of Intelligence, Cambridge, UK�Machine Learning Research Institute (MIRI), Berkeley, USA�Open AI, USA�Partnership on AI, USA�

17

18 of 23

Appendix A2: AI Governance Organizations

AI for Good�AI Governance Online, Beijing�AI Now Institute, New York University, USA�AI Policy Initiative (SAPI), Seoul National University, South Korea �AI Transparency�Berkeley Existential Risk Initiative (BERI), Berkeley, USA�Center for Study of Existential Risk (CSER), Cambridge, UK�Center for Study on Governance of AI (GovAI), Future of Humanity Institute, Oxford, UK�Center for Human-Compatible Artificial Intelligence, Berkeley, USA�DeepMind Ethics & Society�Ethics and Governance of AI Foundation, USA

Ethics and Gov. of AI Initiative, Berkman Klein Center and Media Lab., USA�European AI Alliance, Futurism, EU�Future of Humanity Institute (FHI), Oxford, UK�Future Society (AI Initiative), Harvard University, USA�iCenter, China�IEEE, Global Initiative on Ethical Consideration on AI and AS, USA �JSAI Ethics Committee, Japan Society on Artificial Intelligence, Japan �Research Center for AI Ethics and Sustainable Development, BZAIRI, Beijing�

18

19 of 23

Appendix B1: List of AI Conferences

NeurIPS – Neural Information Processing Systems�ICML – International Conference on Machine Learning�ICLR – International Conference on Learning Representations�AAAI – Association for the Advancement of Artificial Intelligence�CVPR – Computer Vision and Pattern Recognition�ICCV – International Conference on Computer Vision�GECCO – Generic and Evolutionary Computation Conference�COLT – Conference on Learning Theory�IROS – International Conference on Intelligent Robots and Systems�ICIP – International Conference on Image Processing�IJCAI - International Joint Conference on Artificial Intelligence�

19

20 of 23

Appendix B2: List of AI Governance Conferences

AAAI/ACM, AI, Ethics and Society Summit, 2016~Present�Beijing Academy of Artificial Intelligence (BAAI), Annual Conference�European AI Alliance Workshops and Seminars, Futurism, 2019�FLI, Beneficial AI Conference, 2015, 2017, 2019�ITU, AI for Good, 2018, 2019�OECD, AI: Intelligent Machine and Smart Policy, 2017�Seoul National University, AI Today: Governance and Accountability, 2017~Present�Singularity University Global Summit�Tokyo University, AI and Society Symposium, 2017�UAE, World Government Summit, Global Governance of AI, Dubai, 2019�UNESCO, Principles of AI, Global Conference, 2019

20

21 of 23

Appendix C: Courses

AI for All

Coursera, AI for Everybody. 2019.

Diplo, AI: technology, governance, and policy frameworks, 2018.

MIT, Artificial General Intelligence, MIT 6.S099 (by Lex Fridman), 2018.

UC Berkeley, Introduction to AI, CS188.

U of Helsinki, Elements of AI, 2019.

21

22 of 23

Appendix D: List of AI Principles

Asimov’s Three Laws of Robotics, 1940.

Partnership on AI, Tenets, 2016.

Asilomar AI Principles, 2017.

Google, AI Principles, 2018.

IEEE, Ethics of Autonomous and Intelligent Systems, P7000, 2017.

FLI, Lethal Autonomous Weapon Pledge, 2018.

Japanese Government, AI R&D Principles, 2017.

Montreal Declaration on Responsible AI, 2018.

OECD, Going Digital – OECD AI Principles, 2019.

22

23 of 23