- The start of AI: Symbolic Learning Algorithims
- Applications of AI
- Limitations of AI
- Ethical implications of AI
- The potential for superintelligence
Return to Artificial Intelligence Course overview.
Previous lesson: Introduction to AI Part 1
Learn.
The first symbolic learning algorithm was used to solve first year integration math problems. By compiling a series of rules, the algorithm could perform the calculations using a systematic approach. James Slagle developed an one of the first symbolic integration program, known as SAINT in 1961. We are going to recreate his work during the practical’s this week. You can read the paper below for an extension.
Rule based algorithms eventually breakdown when the problem contains contradictory rules.
The following video entails Bill Gates explaining the current and potential applications of AI and some of it’s limitations.
The concept of superintelligence is touched on in the above video, superintelligence is general AI which is defined as being as smart as or smarter then human intelligence.
Currently, AI is no where close to superintelligence, however, we have to establish ways of controlling AI before it reaches that point.
Ok, we are going to conclude this weeks introduction to the theory of AI with a definition of AI.
Artificial Intelligence is the construction of models, or agents, which are able to perceive, think and take actions from a change in its environment, which is a stimuli.
Return to Artificial Intelligence Course overview.
Previous lesson: Introduction to AI Part 1
Master.
Question 1.
Explain three past limitations of AI which we associated with Symbolic learning algorithms.
Question 2.
Explain three current limitations of AI.
Question 3.
Contrast the origins of AI such as Moses symbolic program against the future for superintelligence
Return to Artificial Intelligence Course overview.
Previous lesson: Introduction to AI Part 1
