Foundations of AI
Foundations of AI
• How does the mind arise from a physical brain?
• Where does knowledge come from?
• How does knowledge lead to action?
Aristotle(384-322 B.C): Proposed syllogism to represent the rational part of mind.
All dogs are animals and all animals have 4 legs => All dogs have 4 legs.
Ramon Lull(1315): proposed reasoning system.
Leonard da vinci(1452-1519): designed 1st mechanical calculator.
Blaise pascal(1623-1662): built Pascaline.
Leibniz(1646-1716): built mechanical device that operates on concepts rather than numbers but its scope was limited.
Foundations of AI
Rationalism & Dualism: Human brain’s physical and logical part(animals don’t).
Materialism: which holds that the brain’s operation according to the laws of physics constitutes the mind.
Empiricism: Brain understand the senses.
Induction: rules are acquired by exposure to repeated associations between their elements.
Logical Positivism: All the knowledge can be characterized by the observation of sentences that corresponds to sensory inputs.(rationalism + materialism).
Confirmation theory: acquisition of knowledge through the experience by quantifying the degree of beliefs that should be assigned to the logical sentences based on their connection to observation that confirm and disconfirm them.
Newell and Simon designed general problem solution(GPS) which is greedy regression planning system. Method based on logical planning to reach definite goal.
Utilitarianism decision makers should consider the interests of the many individuals, rational decision making based on maximizing utility should apply to all spheres of human activity, including public policy decisions made on behalf of many individuals.
Consequentialism whether the policy is right or wrong is determined by the expected outcome of the policy.
Deontological ethics doing right think is determined by universal laws like don’t lie, don’t kill etc.
Foundations of AI
2. Mathematics
Deals with what are the formal rules to draw the valid conclusions?
What can be computed?
How do we reason with uncertain information?
Formal logic: George boole worked on propositional or Boolean logic then Gottlob frege extends bools logic to get first order logic which include objects and relations.
Probability: generalization to the logic with uncertain information that can be applied to find the possible outcomes of any events.
Statistics: statistics is the discipline that concerns the colletion, organization, analysis, interpretation and presentation of data. In applying statistics to a scientific, industrial or social problem.
Algorithm: finite sequence of un ambiguous statements to solve problem GCD(m,n) was the first trivial algorithm.
Foundations of AI
Incompleteness: there are true statements that do not have proof within the theorem.
Computability: whether the problem can be solved in an finite amount of time.
Tractability: whether the problem can be solved in an polynomial amount of time or the problem keeps on growing exponentially.
NP-completeness: whether the NP hard can be converted into P class.
Foundations of AI
3. Economics:
• How should we do this when others may not go along?
• How should we do this when the payoff may be far in the future?
Decision theory: combines probability theory with utility theory provides formal and complete framework for individual decision making.
Game theory: is the study of mathematical models of strategic interactions among rational agents.
If they both confess to being bank robbers, then each will serve 3-year imprisonment for both can theft and robbery.
If only one of the confesses to being a bank robber and the other does not, then the person who confesses will serve 1-year and others will serve 10-year imprisonment.
Foundations of AI
Operation research: how to make rational decisions when the payoffs are not immediate but in future and depends on the several actions taken in the sequence (Markov decision and reinforcement learning)
Satisficing: making decisions that are enough rather than laborious calculation of optimal solutions.
Foundations of AI
4. Neuroscience:
Foundations of AI
4. Neuroscience:
Foundations of AI
4. Neuroscience: study of nervous system especially brain. How brain enables thought is the mystery.
Paul Broca(1824-1880) investigation of aphasia in 1861 in the left hemisphere of brain which is responsible for speech production, it was noticed that brain is the collection of large cells called neurons
1873: Camillo Golgi developed a staining technique allowing observation of brain cells which was used by Ramon Cajal to point out that collection of simple cells an lead to thoughts, actions, and consciousness, Brain caused mind.
1929 – Hansberg(EEG) fMRI(ogean) provide the details of ongoing cognitive actions in brain.
Optogenteics (crick,1999) allows both measurement and control of individual neurons on light sensitivity operate.
Foundations of AI
Brain Machine interface was developed (Lebedev and Nicolelis, 2006) for both sensing and motor control not only promises to restore the function to disabled individuals, but also sheds light on neural systems, remarkable finding of this study is brain is able to adjust itself to interface successfully with an external device treating the devices as an sensing organ or limb.
Foundations of AI
5. Psychology (How do humans and animals think and act?)
German physicist Herman von Helmholtz(1821-1894) and his student Wilhelm wundit (1832-1920). Wundt opened the first laboratory of experimental psychology, at the university of Leipzig.
Wundit insisted on carefully controlled experiments in which his workers would perform a perceptual or associative task while introspecting on their thought processes.
Behaviourism H.S.Jennings (1906) Behaviourists insisted on studying only objective measures of the percept's given to an animal(stimulus) and its resulting actions. Behaviourism discovered a lot about rats and pigeons but had less success at understanding humans.
Cognitive psychology views the brain as an information-processing device. Cralk(1943) specified the three key steps of a knowledge-based agent:
1) the stimulus must be translated into an internal representation.
2) the representation is manipulated by cognitive processes to derive new internal representation
3) these are in turn retranslated back into action.
Foundations of AI
5. Psychology (How do humans and animals think and act?)
Geroge Miller presented the Magic Number seven(it is often interpreted to argue that the number of objects an average human can hold in short-term memory is 7+-2. this has occasionally been referred to as Miller’s law.
The second cognitive limitation Miller discusses is memory span. Memory span refers to the longest list of items (digits,letters,words) that a person can repeat back in the correct order is only 50% of trials immediately after the presentation.
Noam Chomsky presented Three Models of language.
Allen newel and Herbert simon presented the logic theory machine.
These three influential papers showed how computer models could be used to address the psychology of memory, language and logical thinking respectively.
“a cognitive theory should be like a computer program” (Anderson, 1980); that is, it should describe a detailed information-processing mechanism whereby some cognitive function might be implemented.
Foundations of AI
6. Computer Engineering (how can we build efficient computer)
1940: the first operational computer was the electromechanical Heath Robinson, built by Alan Turing’s team for a single purpose: deciphering German messages
1943 the same group developed the colossus, a powerful general-purpose machine based on vacuum tubes.
1941 the first operational programmable computer was the Z-3, the invention of Konrad Zuse in Germany.
1940 and 1942 the first electronic computer, the ABC was assembled by john Atanasoff and his student Clifford berry between at lowa state university.
it was the ENIAC, developed as part of a secret military project at the University of Pennsylvania
by a team including John Mauchly and John Eckert, that proved to be the most influential forerunner of modern computers.
Performance doubled every 18 months or so until around 2005.
Foundations of AI
6. Computer Engineering (how can we build efficient computer)
Charles Babbage (1792–1871) designed two machines, neither of which he completed. The Difference Engine was intended to compute mathematical tables for engineering and scientific projects. It was finally built and shown to work in 1991 at the Science Museum in London (Swade, 2000).
Babbage’s Analytical Engine was far more ambitious: it included addressable memory, stored programs, and conditional jumps and was the first artifact capable of universal computation.
Ada Lovelace wrote programs for the unfinished Analytical Engine and even speculated that the machine could play chess or compose music.
Software side of computer science has supplied the operating systems, programming languages, and tools needed to write modern programs.
Foundations of AI
7. Control Theory and Cybernetics(How can artifacts operate under-their own control?)
Control theory was created by Norbert Wiener (1894-1964) who studied biological and mechanical control systems and their connection to cognitive systems.
Ktesibios of Alexandria (c. 250 B.C) built the self-controlling machine: a water clock with a regulator that maintained a constant flow rate.
Steam engine governor, created by james watt (1736 – 1819)
Thermostat, invented by cornelis Drebbel (1572-1633)
Cybernetics: feedback mechanism.
Homeostatic: intelligence could be created by the use of homeostatic devices containing appropriate feedback loops to achieve stable adaptive behaviour.
Objective and cost function: Modern control theory, especially the branch known as stochastic optimal control, has its goal the design of systems that maximize an objective function and reduces cost function over time.
Foundations of AI
7. Linguistic : (How does language relate to thought)
In 1957, B.F Skinner published verbal behaviour. This was a comprehensive, detailed account of the behaviourist approach to language learning.
Noam Chomsky, who had just published a book on his own theory, Syntactic Structures. Chomsky pointed out that the behaviorist theory did not address the notion of creativity in language—it did not explain how a child could understand and make up sentences that he or she had never heard before.
Computational linguistics or natural language COMPUTATIONAL LINGUISTICS processing: understanding language requires an understanding of the subject matter and context, not just an understanding of the structure of sentences.