(The below text version of the notes is for search purposes and convenience. See the PDF version for proper formatting such as bold, italics, etc., and graphics where applicable. Copyright: 2023 Retraice, Inc.)

Re107: Three Kinds of AI (Day 5, AIMA4e Chpt. 5)

retraice.com

War, peace and commerce AI in our multi-agent world. Considering multi-agent environments as economies, adversarial games, or merely nondeterministic; commerce, peace, war; Bostrom's `capacity building' and `strategic analysis'; adversary arguments; the computer control game; solving for mobilization and war.

Air date: Thursday, 5th Jan. 2023, 10:00 PM Eastern/US.

At least three stances toward other agents

Russell & Norvig (2020) p. 148:

There are at least three stances we can take towards multi-agent environments. The first stance, appropriate when there are a very large number of agents, is to consider them in the aggregate as an economy, allowing us to do things like predict that increasing demand will cause prices to rise, without having to predict the action of any individual agent.

Second, we could consider adversarial agents as just a part of the environment--a part that makes the environment nondeterministic. But if we model the adversaries in the same way that, say, rain sometimes falls and sometimes doesn't, we miss the idea that our adversaries are actively trying to defeat us, whereas the rain supposedly has no such intention.

The third stance is to explicitly model the adversarial agents with the techniques of adversarial game-tree search. That is what this chapter covers.^1

War, peace and commerce AI

Russell & Norvig (2020) p. 168:

Bobby Fischer declared that `chess is war,' but chess lacks at least one major characteristic of real wars, namely, partial observability. In the `fog of war,' the whereabouts of enemy units is often unknown until revealed by direct contact. As a result, warfare includes the use of scouts and spies to gather information and the use of concealment and bluff to confuse the enemy.

* Commerce AI: markets and prices, supply and demand; * Peace AI: Bostrom's^2 capacity building (as long as we model the relevant environment as having one agent, and we aren't doing AOSE^3); * War AI: search changes at the point of an `adversary argument'^4 ; now we're into (the results of) Bostrom's `strategic analysis'.^5

Questions and answers

Laymen's questions that arose (mostly) before Retraice's `technical turn' at Retraice (2022/11/20) are starting to get technical answers: * Why does the CC (computer control) game feel like a (game-theoretic) game and war? (Retraice (2022/11/19)) See Russell & Norvig (2020) chpt. 4, p. 136 on `adversary arguments'; chpt. 5, p. 168 on `chess is war'. * What do players do? (Retraice (2022/11/16); Retraice (2022/11/18)) Solve. (Retraice (2023/01/03)) * What will war-AI and general mobilization look like? (Retraice (2022/12/03)) Solving, and deploying solutions to, hard game-problem-environments^6 (partially observable, multi-agent, nondeterministic, sequential, dynamic, continuous, unknown). (Russell & Norvig (2020) p. 47; cf. Retraice (2023/01/02).

References

Bostrom, N. (2014). Superintelligence: Paths, Dangers, Strategies. Oxford. First published in 2014. Citations are from the pbk. edition, 2016. ISBN: 978-0198739838. Searches: https://www.amazon.com/s?k=978-0198739838 https://www.google.com/search?q=isbn+978-0198739838 https://lccn.loc.gov/2015956648

Retraice (2022/11/16). Re52: Big Questions About AI. retraice.com. https://www.retraice.com/segments/re52 Retrieved 17th Nov. 2022.

Retraice (2022/11/18). Re54: Implications and Endgames. retraice.com. https://www.retraice.com/segments/re54 Retrieved 19th Nov. 2022.

Retraice (2022/11/19). Re55: The Computer Control Game. retraice.com. https://www.retraice.com/segments/re55 Retrieved 20th Nov. 2022.

Retraice (2022/11/20). Re56: A Valuable Brick: `Artificial Intelligence: A Modern Approach' 4th ed. retraice.com. https://www.retraice.com/segments/re56 Retrieved 21st Nov. 2022.

Retraice (2022/12/03). Re69: TABLE-DRIVEN-AGENT Part 5 (ECMP and AIMA4e p. 48). retraice.com. https://www.retraice.com/segments/re69 Retrieved 4th Dec. 2022.

Retraice (2023/01/02). Re104: Agent Functions, Agent Programs, Task Environments (Day 2, AIMA4e Chpt. 2). retraice.com. https://www.retraice.com/segments/re104 Retrieved 3rd Jan. 2023.

Retraice (2023/01/03). Re105: Solve or Be Solved (Day 3, AIMA4e Chpt. 3). retraice.com. https://www.retraice.com/segments/re105 Retrieved 4th Jan. 2023.

Russell, S., & Norvig, P. (2020). Artificial Intelligence: A Modern Approach. Pearson, 4th ed. ISBN: 978-0134610993. Searches: https://www.amazon.com/s?k=978-0134610993 https://www.google.com/search?q=isbn+978-0134610993 https://lccn.loc.gov/2019047498

Footnotes

^1 Cf. p. 168 on `war', p. 136 on `adversary arguments', p. 169. on `evasive moves'.

^2 Bostrom (2014) pp. 317 ff.

^3 Retraice (2022/12/03).

^4 Russell & Norvig (2020) p. 136.

^5 Bostrom (2014) pp. 317-317.

^6 Russell & Norvig (2020) p. 42.