Notes by Retraice

(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: 2022 Retraice, Inc.)

Re86: Code Reading (Best-First-Search Part 5, AIMA4e pp. 73-74)

retraice.com

Getting used to the components of a search problem and algorithm implemented in Python. The Problem class and its place-holder methods; the RouteProblem subclass and its more substantial methods; the Map class, and neighbors as actions; the links argument to Map as actions, the locations argument as states; the multimap function as the key to generating a list of neighbors (actions).

Air date: Sunday, 18th Dec. 2022, 10:00 PM Eastern/US.

Problem and RouteProblem

Most of the code below is from https://github.com/aimacode/aima-python/blob/master/search4e.ipynb. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Formalizing a search problem (from Re82):^1

* state space: a set of possible states of the environment; * initial state: the state in which the agent starts; * goal state(s): a set of one or more; account for one, some, infinite (by means of a property) by specifying Is-Goal method for problem; * actions: what the agent can do; Actions(state) returns a finite set of actions that can be executed in state; * transition model: describes what actions do; Result(state,action) returns the state s' that results from doing action in state; * action cost function: Action-Cost(s,a,s') gives the numeric cost of applying action a in state s to reach new state s'. Cf. the evaluation function, which we'll use to prioritize our nodes for next expansion, and the objective function, which was our cost measure to be minimized in the airport problem.^2

Map, multimap and romania ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

best_first_search ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Other sources consulted during this livestream: * Russell & Norvig (2020); * Retraice (2022/12/14); * Retraice (2022/12/15); * Retraice (2022/12/16); * Retraice (2022/12/17); * http://aima.cs.berkeley.edu/figures.pdf; * https://github.com/aimacode/aima-python/blob/master/search4e.ipynb; * https://github.com/retraice/ReAIMA4e/.

__

References

Retraice (2022/12/11). Re78: Recap of Gradients and Partial Derivatives (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re78 Retrieved 12th Dec. 2022.

Retraice (2022/12/14). Re82: What is a problem? (BEST-FIRST-SEARCH Part 1, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re82 Retrieved 15th Dec. 2022.

Retraice (2022/12/15). Re83: A Problem Instantiated (BEST-FIRST-SEARCH Part 2, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re83 Retrieved 16th Dec. 2022.

Retraice (2022/12/16). Re84: A Node Instantiated (BEST-FIRST-SEARCH Part 3, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re84 Retrieved 17th Dec. 2022.

Retraice (2022/12/17). Re85: The Details (BEST-FIRST-SEARCH Part 4, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re85 Retrieved 18th Dec. 2022.

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 Russell & Norvig (2020) p. 65.

^2 Retraice (2022/12/11).

(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: 2022 Retraice, Inc.)

Re85: The Details (Best-First-Search Part 4, AIMA4e pp. 73-74)

retraice.com

Looking ahead at the code we'll need. An attempt to build a toy problem reveals unsatisfied dependencies; the need for a problem implementation with state space, actions sets, transition model and action cost function; AIMA's RouteProblem class and best_first_search function implementations as guides; walking through the suites of each; the need for PriorityQueue and f to order our search tree's frontier of nodes.

Air date: Saturday, 17th Dec. 2022, 10:00 PM Eastern/US.

expand(problem, node) dependencies ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC We need to implement a problem, with a state space, initial and goal states, actions sets, transition model and action cost function. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

The RouteProblem example ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC The AIMA implementation of RouteProblem, a subclass of Problem. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Looking ahead at Best-First-Search implemented ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC We're also going to need a PriorityQueue, and f, an implemented evaluation function that will prioritize our nodes for next expansion. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Other sources consulted during this livestream: * Russell & Norvig (2020); * Retraice (2022/12/14); * Retraice (2022/12/15); * Retraice (2022/12/16); * http://aima.cs.berkeley.edu/figures.pdf; * https://github.com/aimacode/aima-python/blob/master/search4e.ipynb; * https://github.com/retraice/ReAIMA4e/tree/main/Re85-BEST-FIRST-Part-4.

__

References

Retraice (2022/12/14). Re82: What is a problem? (BEST-FIRST-SEARCH Part 1, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re82 Retrieved 15th Dec. 2022.

Retraice (2022/12/15). Re83: A Problem Instantiated (BEST-FIRST-SEARCH Part 2, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re83 Retrieved 16th Dec. 2022.

Retraice (2022/12/16). Re84: A Node Instantiated (BEST-FIRST-SEARCH Part 3, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re84 Retrieved 17th Dec. 2022.

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

(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: 2022 Retraice, Inc.)

Re84: A Node Instantiated (Best-First-Search Part 3, AIMA4e pp. 73-74)

retraice.com

Writing, as a class in Python, a data structure to represent a reached state in the environment's state space. State space and search tree; nodes and edges; nodes as representing unique paths; parent nodes, applied actions and total path-from-initial-state cost; init, repr, len and lt magic methods in Python.

Air date: Friday, 16th Dec. 2022, 10:00 PM Eastern/US.

The structure of a node

In search problems we first have a state space: a set of states and the actions that cause transitions from one to another state. One method of solving search problems is based on building a search tree in the state space to represent paths between the initial state and a goal state. The search tree is composed of nodes that represent unique paths from the initial to a given state, and edges that represent actions.^1

Nodes implemented in code need these attributes:^2 * State: node.State, the state to which the node corresponds; * Parent: node.Parent, the node that generated this node; * Action applied: node.Action, the action that was applied to the parent node; * Path-to cost: node.Path-Cost, the total cost of the path from initial to this node. In math it's g(node).

Nodes as class in Python ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC Python code implementing an object class data structure to represent nodes of a search tree of a state space. Code at https://github.com/retraice/ReAIMA4e. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

___

References

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 Russell & Norvig (2020) p. 71.

^2 Russell & Norvig (2020) p. 73.

(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: 2022 Retraice, Inc.)

Re83: A Problem Instantiated (Best-First-Search Part 2, AIMA4e pp. 73-74)

retraice.com

Writing a well-defined problem, in Python, as an object that's an instance of the class Problem. Object-oriented programming; class Problem as subclass of object, implementing structure of well-defined problem; initial state and goal state as attributes in Problem; the four functions Actions(), Result(), Is-Goal() and Action-Cost(), and the informed search function h(), as methods in Problem.

Air date: Thursday, 15th Dec. 2022, 10:00 PM Eastern/US.

The structure of a problem^1

* state space: a set of possible states of the environment; * initial state: the state in which the agent starts; * goal state(s): a set of one or more; account for one, some, infinite (by means of a property) by specifying Is-Goal method for problem; * actions: what the agent can do; Actions(state) returns a finite set of actions that can be executed in state; * transition model: describes what actions do; Result(state,action) returns the state s' that results from doing action in state; * action cost function: Action-Cost(s,a,s') gives the numeric cost of applying action a in state s to reach new state s'. Cf. the evaluation function, which we'll use to prioritize our nodes for next expansion, and the objective function, which was our cost measure to be minimized in the airport problem.^2

Problem implemented in Python ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Code available at https://github.com/retraice/ReAIMA4e. Adapted from https://github.com/aimacode/aima-python/blob/master/search4e.ipynb.

__

References

Retraice (2022/12/11). Re78: Recap of Gradients and Partial Derivatives (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re78Retrieved 12th Dec. 2022.

Retraice (2022/12/14). Re82: What is a problem? (BEST-FIRST-SEARCH Part 1, AIMA4e pp. 73-74). retraice.com. https://www.retraice.com/segments/re82Retrieved 15th Dec. 2022.

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 Adapted from Russell & Norvig (2020) p. 65. See also Retraice (2022/12/14).

^2 Retraice (2022/12/11).

(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: 2022 Retraice, Inc.)

Re82: What is a problem? (Best-First-Search Part 1, AIMA4e pp. 73-74)

retraice.com

Using sets and functions to formalize problems. Passing a problem to a function vs. passing a variable or number; English meaning and formal meaning of problem; problem as description of task environment; problem as object instantiating class, written in Python; state space, initial state, goal state(s), actions, transition model, action cost function, evaluation function.

Air date: Wednesday, 14th Dec. 2022, 11:00 PM Eastern/US.

The problem in the pseudocode

The question is: What is a problem? It seems more complicated than passing a variable or a number to a function. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

PIC The state space (cities graph), initial state (Arad), goal state (Bucharest), and pseudocode for Best-First-Search, adapted from Russell & Norvig (2020) pp. 64, 73. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Other figures and visuals used during the livestream: * https://github.com/aimacode/aima-python/blob/master/search4e.ipynb * http://aima.cs.berkeley.edu/figures.pdf * http://aima.cs.berkeley.edu/algorithms.pdf

Formalizing a search problem:^1

We are not talking about the English-language meaning of `problem'. We're talking about a well-defined, well-formed description of something (a task environment^2). The problem is going to be an object (instance of a class, as in `object-oriented' programming) written in Python code. The following sets and functions will need to be written explicitly in Python before we can pass problem to Best-First-Search. * state space: a set of possible states of the environment; * initial state: the state in which the agent starts; * goal state(s): a set of one or more; account for one, some, infinite (by means of a property) by specifying Is-Goal method for problem; * actions: what the agent can do; Actions(state) returns a finite set of actions that can be executed in state; * transition model: describes what actions do; Result(state,action) returns the state s' that results from doing action in state; * action cost function: Action-Cost(s,a,s') gives the numeric cost of applying action a in state s to reach new state s'. Cf. the evaluation function, which we'll use to prioritize our nodes for next expansion, and the objective function, which was our cost measure to be minimized in the airport problem.^3

___

References

Retraice (2022/12/11). Re78: Recap of Gradients and Partial Derivatives (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re78 Retrieved 12th Dec. 2022.

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 Russell & Norvig (2020) p. 65.

^2 Russell & Norvig (2020) pp. 42-47.

^3 Retraice (2022/12/11).

(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: 2022 Retraice, Inc.)

Re81: Recap of Natural Intelligence (Re10-Re13)

retraice.com

Naturally evolved intelligent entities are the origin of the other kinds. Rocks and podcasts; guessing and checking; learning separate from intelligence; perception and intelligence; fight, flight and environments; smart things traveling; mental time and space travel; space and time physics; aliens; how vs. what one thinks; the crazies; science's weaknesses, strategic intelligence's strengths; attentional blind spots; care constraints on intelligences; computers don't give a damn; existential holism; volitional necessity; common sense.

Air date: Wednesday, 14th Dec. 2022, 10:00 PM Eastern/US.

If the phenomenon of artificial intelligence, according to Russell & Norvig (2020), is `intelligent agents', and another phenomenon of artificial intelligence, according to Retraice (2022/12/13), is intelligent artifacts, and the phenomenon of strategic intelligence, according to Retraice (2022/12/12), is `intelligence agents' (and perhaps `agencies', government and otherwise), what about naturally evolved intelligence, the origin of it all?

Re10: Living to Guess Another Day^1

On guessing, checking and fighting. Rocks can't listen to podcasts; we're smart and dumb; guessing and checking answers are powerful uses of intelligence; true things seem to stay true and assert themselves over time; Barlow on guessing as detecting new non-chance patterns in redundant sensory messages to improve predictions; Macphail on the disconnect between learning and intelligence; Barlow on perception being central to intelligence, and an information-theoretic absolute measure of intelligence; IQ; science and logic as guess-checking; Feynman on science as reporting everything and fooling neither yourself nor others; the machine-human easy-hard yin-yang; fight or flight; environments partially determine which actions are intelligent; Darwin was just busy; science is recent, babies are not.

Re11: Travel^2

On where intelligent things can go, and when, and how. Smart things travel, dumb things just move; the difference is inside the thing; time travel is done mentally; the marshmallow test; time travel is space travel; space and time are quantum and relative and bizarre; mental travel is not hard for intelligent things; Hinton's five-year-old; reflexes and mental travel; essential elements of intelligence.

Re12: Aliens and Trouble^3

On thinking about alien intelligences--real or not, perceived or not. How one thinks matters more than what one thinks at any given time; reputations and jobs are at risk; the crazies; many or most witnesses are serious, whether or not they're right; science (public hypotheses, assumptions of the uniformity of nature, requirements of reproducibility, etc.) is inadequate for investigating strategically intelligent, evasive entities; `them'; documents, manipulation, evidence; serious consideration vs. acceptance of huge attentional blind spots.

Re13: The Care Factor^4

A constraint on intelligences. An intelligence might be bound by what it cares about; Haugeland and Frankfurt on intelligences caring; computers don't give a damn; living some kind of life as prerequisite to understanding some kinds of things; existential holism; common sense; defining `care'; logical, causal and `volitional' necessity; care and the unthinkable.

_

References

Retraice (2020/11/02). Re10: Living to Guess Another Day. retraice.com. https://www.retraice.com/segments/re10 Retrieved 2nd Nov. 2020.

Retraice (2020/11/04). Re11: Travel. retraice.com. https://www.retraice.com/segments/re11 Retrieved 4th Nov. 2020.

Retraice (2020/11/05). Re12: Alien Troubles. retraice.com. https://www.retraice.com/segments/re12 Retrieved 5th Nov. 2020.

Retraice (2020/11/10). Re13: The Care Factor. retraice.com. https://www.retraice.com/segments/re13 Retrieved 10th Nov. 2020.

Retraice (2022/12/12). Re79: Recap of Strategic Intelligence (Re1-Re5). retraice.com. https://www.retraice.com/segments/re79 Retrieved 13th Dec. 2022.

Retraice (2022/12/13). Re80: Recap of Artificial Intelligence (Re6-Re9). retraice.com. https://www.retraice.com/segments/re80 Retrieved 14th Dec. 2022.

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 Retraice (2020/11/02). ^2 Retraice (2020/11/04). ^3 Retraice (2020/11/05). ^4 Retraice (2020/11/10).

(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: 2022 Retraice, Inc.)

Re80: Recap of Artificial Intelligence (Re6-Re9)

retraice.com

A first pass at the phenomena of artifacts, machines and goals. Six questions about AI; artifacts as interfaces; intentions, purposes, goals; absorption and exsorption of goals; the transition from inanimate to animate; new goals coming out of machines; the totalitarian solution; `artilect war'; living systems as like artifacts; insufficient warning; energy, autonomy, reproduction, communication, control; `singularity'; deferred gratification; AI and its users; camouflage and new perceptions; prediction machines; `they' are the controllers; reality and fitness; the quantum chessboard.

Air date: Tuesday, 13th Dec. 2022, 10:00 PM Eastern/US.

If the phenomenon of artificial intelligence, according to Russell & Norvig (2020), is `intelligent agents', and the phenomenon of strategic intelligence, according to Retraice (2022/12/12), is `intelligence agents' (and perhaps `agencies' government and otherwise), then what to think about lifelike, strategically behaving artifacts--intelligent artifacts?

Re6: Interface^1

On what AI is or isn't, and whether it's between things. Six questions about what AI is; Who?; Where? metal, electricity, patterns; When? locations in time; How? and Why?; Which, a basis for the other questions; the ocean of things going on out there; Simon on artifacts as interfaces between inner and outer environments, given intentions and purposes; artifacts are everywhere; the many goals related to even a simple battery.

Re7: Artifactual Goals^2

On goals producing artifacts-producing-goals, and machines coming alive. Simon on goals linking inner and outer environments and being central to the description of artifacts; artifacts absorbing goals; the physical interpretation of `goal'; artifacts causing goals; artifacts absorbing their own goals--goal absorption and exsorption in machines of moving parts; inter-artifact absorption (e.g. two abaci); the transition from inanimate things to animate things--AlphaZero, SCADA (supervisory control and data acquisition) systems; new goals coming out of machines; goals moving amongst entities of all sorts; What would we do if we knew?; totalitarianism; `artilect war'.

Re8: Strange Machines^3

A survey of the idea that technology is creatures. Artifacts with goals should not be called artifacts; Simon on living systems being like artifacts; organic evolution vs. Walter's tortoises; Butler recommends war to the death; Dyson agrees the entities are real; Wolfram created stunning cellular automata behavior; Yudkowsky says there will be no sufficient warning; I. J. Good says take scifi more seriously; definitions can be distractions; Russell and Norvig say the difference in technologies is autonomy; von Neumann's neutral definition, and Kurzweil's optimistic definition, of `singularity'; the morality of machine takeover; S. Russell on the user's mind as environment; Dyson on reproduction, communication and control; Smallberg on energy sources; search and `looking for' things; Dietterich on reproduction and autonomy; What to do? Bostrom on deferred gratification; skyscrapers seem taller than they are.

Re9: They Can See You^4

On what is perceptible to AI, and AI controllers. AI and its users can see hidden patterns, including their extrapolation into the future and interpolation toward secrets; the next thing after rods and cones; camouflage lags behind new perceptions; prediction machines; whoever controls AI is the `they'; the Target debacle; accumulating capacities; reality and fitness; the chessboard is a quantum one; What's the work?

_

References

Kissinger, H. A., Schmidt, E., & Huttenlocher, D. (2021). The Age of AI. Little, Brown and Company. ISBN: 978-0316273800. Searches: https://www.amazon.com/s?k=9780316273800 https://www.google.com/search?q=isbn+9780316273800 https://lccn.loc.gov/2021943914

Retraice (2020/10/25). Re6: Interface. retraice.com. https://www.retraice.com/segments/re6 Retrieved 26th Oct. 2020.

Retraice (2020/10/26). Re7: Artifactual Goals. retraice.com. https://www.retraice.com/segments/re7 Retrieved 27th Oct. 2020.

Retraice (2020/10/28). Re8: Strange Machines. retraice.com. https://www.retraice.com/segments/re8 Retrieved 29th Oct. 2020.

Retraice (2020/10/31). Re9: They Can See You. retraice.com. https://www.retraice.com/segments/re9 Retrieved 31st Oct. 2020.

Retraice (2022/12/12). Re79: Recap of Strategic Intelligence (Re1-Re5). retraice.com. https://www.retraice.com/segments/re79 Retrieved 13th Dec. 2022.

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

Simon, H. A. (1996). The Sciences of the Artificial. MIT, 3rd ed. ISBN: 0262691914. Searches: https://www.amazon.com/s?k=0262691914 https://www.google.com/search?q=isbn+0262691914 https://lccn.loc.gov/96012633 Previous editions available at: https://archive.org/search.php?query=The%20sciences%20of%20the%20artificial

Footnotes

^1 Retraice (2020/10/25). ^2 Retraice (2020/10/26). Discussed during the livestream: Simon (1996) pp. 151-153 on the undue power of professional engineers in affecting social systems. ^3 Retraice (2020/10/28). ^4 Retraice (2020/10/31). On AI seeing things that we can't, see Kissinger et al. (2021) pp. 13-17.

(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: 2022 Retraice, Inc.)

Re79: Recap of Strategic Intelligence (Re1-Re5)

retraice.com

An overview of the phenomenon of `intelligence agents' and agencies (government and otherwise). Natural, artificial and strategic intelligence; difficulties of the subject of strategic intelligence; likely and unlikely secrets; problems of evidence, information, trust, and survival; openness and security; group and sub-group hostility; team survival of life, leisure and legislation; totalitarian logic.

Air date: Monday, 12th Dec. 2022, 11:00 PM Eastern/US.

Retraice is a constructive podcast. Our point of departure was the concept of intelligence, but our `gradient'^1 at each point along the way has indicated our direction should be toward technical AI. Strategic intelligence was a first stop on the path to that local maximum.

If the phenomenon of artificial intelligence is `intelligent agents', the phenomenon of strategic intelligence is `intelligence agents', and perhaps `agencies'--government and otherwise.

Re1: Three Kinds of Intelligence^2

On natural, artificial and strategic intelligence; intellectual hazards; secrets. Russell on empiricism; Retraice, Inc.; three kinds of intelligence--natural, artificial and strategic; likely and unlikely secrets leading to reputational minefields an attentional blind spots; tough subjects testing one's mettle; knowledge, evidence, fallacies; the price of information; deception; intelligence in warfare.

Re2: Tell the People, Tell Foes^3

On Dulles, evidence, trust, `is' and `ought', and capacities. The reputations of Allen Dulles; his book `The Craft of Intelligence'; chpt. 15 `Security in a Free Society'; the urge to talk and its consequences; problems of evidence--texts, contexts, `we' (Dulles' and the readers'), inaccessibility of thoughts, inaccessibility of past events, testable ideas, the present beliefs of others; problems of trust--vulnerability of reputations, the power (control) to know and the incentive to tell; control; incentives; is and ought; good guys and bad guys; violence; psychological and cyber warfare; our capacity to hold opposing ideas in mind and await more evidence.

Re3: Tell Everyone^4

On problems of information: `careless leaks' and `giveaways'. Dulles on `careless leaks' not due to malice, and `giveaways' of technical details in the military press, government hearings and investigations, inter-service rivalries, and journalism; motivations to give and get information; spies and their sides; the tensions between openness and security.

Re4: Trust No One^5

On problems of trust: `contrived leaks' and betrayals. Dulles on whistle-blowers, press manipulators, traitors and enemies; unintentional problems of information vs. intentional problems of trust; sub-group hostility; whole-group hostility; hardening and bulwarking; MICE; havoc; moles.

Re5: Hints From Inside^6

On problems of survival: destruction by enemies and fanatics. Dulles and expensive information; team survival of life, leisure and legislation against enemies and fanatics exploiting information and trust problems; solutions and mitigation; totalitarian logic--stopping minds from changing by destroying them; killing and kinetic warfare; propaganda, rhetoric, fear, inspiration, norms, laws, deception; Dulles' major points and proposals in chpt. 15.

_

References

Retraice (2020/09/07). Re1: Three Kinds of Intelligence. retraice.com. https://www.retraice.com/segments/re1 Retrieved 22nd Sep. 2020.

Retraice (2020/09/08). Re2: Tell the People, Tell Foes. retraice.com. https://www.retraice.com/segments/re2 Retrieved 22nd Sep. 2020.

Retraice (2020/09/09). Re3: Tell Everyone. retraice.com. https://www.retraice.com/segments/re3 Retrieved 22nd Sep. 2020.

Retraice (2020/09/10). Re4: Trust No One. retraice.com. https://www.retraice.com/segments/re4 Retrieved 22nd Sep. 2020.

Retraice (2020/09/11). Re5: Hints From Inside. retraice.com. https://www.retraice.com/segments/re5 Retrieved 22nd Sep. 2020.

Retraice (2022/12/11). Re78: Recap of Gradients and Partial Derivatives (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re78 Retrieved 12th Dec. 2022.

Footnotes

^1 Retraice (2022/12/11). ^2 Retraice (2020/09/07). ^3 Retraice (2020/09/08). ^4 Retraice (2020/09/09). ^5 Retraice (2020/09/10). ^6 Retraice (2020/09/11).

(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: 2022 Retraice, Inc.)

Re78: Recap of Gradients and Partial Derivatives (AIMA4e pp. 119-122)

retraice.com

An overview of Re70-Re76. The airport problem; squares vs. sums of distances; the six-dimensional solution space for three airports, many cities; the gradient as `slope' toward the best solution; the two-dimensional solution space for one airport, two cities; vector calculus as `pile of numbers' approach; local- vs. global-best solutions; the partial derivatives that make up the gradient, a total derivative; solving the two-cities-one-airport problem by gradient descent.

Air date: Sunday, 11th Dec. 2022, 11:00 PM Eastern/US.

We're focusing on the math and code of AIMA4e^1 right now, December 2022. This is in service of our plan to deep-dive the book from Jan.-Jun., 2023.

The airport problem from p. 120: Where to locate three airports amongst several cities in Romania such that the sum of the squares of the straight-line distances between each city and its nearest airport (which is a good way of measuring how well the airports have been placed) is minimized?

Re70: Gradients and Partial Derivatives Part 1 (AIMA4e pp. 119-122)^2

The math of `Local Search in Continuous Spaces'. The three airports and the six-dimensional vector that represents their coordinates and therefore solutions to the problem; why the sums of squares of the distances reflects better and worse airport locations; how the gradient is like a `slope' in the space of better and worse solutions; the gap between passive and active knowledge and a demonstration of confidence being calibrated.

Re71: Gradients and Partial Derivatives Part 2 (AIMA4e pp. 119-122)^3

Put the airport problem first. A simple, two-city-one-airport version of the problem; how vectors and calculus can represent this real-world problem as a `pile of numbers'; the airport coordinates as independent variables; the objective function as dependent variable; the gradient as tool for improving our first guess-hypothesis; local vs. global solution optimization.

Re72: Gradients and Partial Derivatives Part 3 (AIMA4e pp. 119-122)^4

Be in the math. The four key equations that define our problem and solutions--the solution guesses (vectors), the objective function of the solution guesses (our `score' or `cost' to be minimized), the gradient vector (the `slope' toward better and worse solutions), and the six partial derivatives that combine to make up our gradient vector (a six-dimensional `slope').

Re73: Gradients and Partial Derivatives Part 4 (AIMA4e pp. 119-122)^5

The limits that define our gradient. Discussion of how the gradient will indicate the `direction' of improving our first guess; full expansion of the gradient for the three-airport problem into six equations of partial derivatives and limits, each with respect to a different independent variable (airport partial coordinate).

Re74: Gradients and Partial Derivatives Part 5 (AIMA4e pp. 119-122)^6

Bringing the algebra back down to numbers. Calculating squares of distances on the map of Romania; a first look at the algebra of calculating the objective function given the coordinates of cities and airports.

Re75: Gradients and Partial Derivatives Part 6 (AIMA4e pp. 119-122)^7

Can we please just place an airport? A first airport guess for the two-city version of the problem; calculating the objective function value given that first guess and the coordinates of our two cities.

Re76: Gradients and Partial Derivatives Part 7 (AIMA4e pp. 119-122)^8

Moving the airport to improve its value. Mapping one, two and three guesses as directed by gradient descent calculations; the lucky-unlucky choice of our first airport location; the hand-math algebraic manipulations and the spreadsheet-math iterations that show the behavior of the partial derivatives in the limit; the common-sense basis of our confidence in the solution provided by gradient descent.

_

References

Retraice (2022/12/04). Re70: Gradients and Partial Derivatives Part 1 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re70 Retrieved 5th Dec. 2022.

Retraice (2022/12/05). Re71: Gradients and Partial Derivatives Part 2 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re71 Retrieved 6th Dec. 2022.

Retraice (2022/12/06). Re72: Gradients and Partial Derivatives Part 3 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re72 Retrieved 7th Dec. 2022.

Retraice (2022/12/07). Re73: Gradients and Partial Derivatives Part 4 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re73 Retrieved 8th Dec. 2022.

Retraice (2022/12/08). Re74: Gradients and Partial Derivatives Part 5 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re74 Retrieved 9th Dec. 2022.

Retraice (2022/12/09). Re75: Gradients and Partial Derivatives Part 6 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re75 Retrieved 10th Dec. 2022.

Retraice (2022/12/10). Re76: Gradients and Partial Derivatives Part 7 (AIMA4e pp. 119-122). retraice.com. https://www.retraice.com/segments/re76 Retrieved 11th Dec. 2022.

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 Russell & Norvig (2020). ^2 Retraice (2022/12/04). ^3 Retraice (2022/12/05). ^4 Retraice (2022/12/06). ^5 Retraice (2022/12/07). ^6 Retraice (2022/12/08). ^7 Retraice (2022/12/09). ^8 Retraice (2022/12/10).

(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.)

Re77: Recap of TABLE-DRIVEN-AGENT (AIMA4e p. 48)

retraice.com

An overview of Re65-Re69. The agent program; its agent function; several Python implementations; definitions of various and sometimes confusing terms; the mathematical behavior of the look-up table; making Python do that math; the ECMP (English-Code-Math-Progress) cycle approach to technical learning.

Air date: Sunday, 11th Dec. 2022, 10:00 PM Eastern/US.

We're focusing on the math and code of AIMA4e^1 right now, December 2022. This is in service of our plan to deep-dive the book from Jan.-Jun., 2023.

*CORRECTION*: The `agent function' is not `the smart part' of the agent program. "The agent function is an abstract mathematical description.... that maps any percept sequence to an action"; "the agent program is a concrete implementation [of the agent function], running within some physical system." (100!black!//Russell & Norvig (2020) pp. 37-37) Function: math. Program: code. [Correction at Jan 2nd, 2023.]

The AIMA4e TABLE-DRIVEN-AGENT is described on p. 48. The idea is that the agent program has a look-up table (a kind of `agent function') that tells it what to do in response to every percept it experiences, considering not just that percept but also the history of percepts leading up to it.

Re65: TABLE-DRIVEN-AGENT Part 1 (AIMA4e p. 48)^2

A basic agent program in both AIMA4e pseudocode and Python. Comparison of the AIMA4e pseudocode and the official Python implementation; attempted definitions of program function, agent program, agent function, agent, architecture, program and memory.

Re66: TABLE-DRIVEN-AGENT Part 2 (AIMA4e p. 48)^3

A basic agent program in both AIMA4e pseudocode and Python. More and better explanations and definitions of initialize, python dictionary, memory object, computer program function, mathematical function, and intelligent agent function; demonstration by running three ground-up Python implementations using the iPython interpreter; comments on the sentiment `always be coding'.

Re67: TABLE-DRIVEN-AGENT Part 3 (AIMA4e p. 48)^4

The mathematical fate of TABLE-DRIVEN-AGENT. Explanation of the algebraic expression that counts the number of table entries necessary for any given table-driven agent; expansion, simplification and application of the expression to our example agent; description of the combinatorial explosion of the last term; amendment and correction of our use of the term `AI code'.

Re68: TABLE-DRIVEN-AGENT Part 4 (AIMA4e p. 48)^5

Computer, do the math. Recommendations about controlling Python and Python package environments against future dependency problems; implementation of a Python for-loop to do the arithmetic of our table-size math expression for any number of percepts and percept events.

Re69: TABLE-DRIVEN-AGENT Part 5 (ECMP and AIMA4e p. 48)^6

A strategy for technical progress. Reflections on tempting digressions in the study of AI; a four-part cycle (ECMP, English, Code, Math, Progress) for acquiring technical knowledge; reflections on `job-AI' and `war-AI', and future mobilizations for war.

_

References

Retraice (2022/11/29). Re65: TABLE-DRIVEN-AGENT Part 1 (AIMA4e p. 48). retraice.com. https://www.retraice.com/segments/re65 Retrieved 30th Nov. 2022.

Retraice (2022/11/30). Re66: TABLE-DRIVEN-AGENT Part 2 (AIMA4e p. 48). retraice.com. https://www.retraice.com/segments/re66 Retrieved 1st Dec. 2022.

Retraice (2022/12/01). Re67: TABLE-DRIVEN-AGENT Part 3 (AIMA4e p. 48). retraice.com. https://www.retraice.com/segments/re67 Retrieved 2nd Dec. 2022.

Retraice (2022/12/02). Re68: TABLE-DRIVEN-AGENT Part 4 (AIMA4e p. 48). retraice.com. https://www.retraice.com/segments/re68 Retrieved 3rd Dec. 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.

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 Russell & Norvig (2020). ^2 Retraice (2022/11/29). ^3 Retraice (2022/11/30). ^4 Retraice (2022/12/01). ^5 Retraice (2022/12/02). ^6 Retraice (2022/12/03).