6.3 Artificial Intelligence
AI
A branch of CS that involves creating systems that can perform tasks alone that would normally require human intelligence to complete. The goal is to stimulate intelligent behaviour in machines.
Machine Learning (ML) is a subset of AI that focuses on giving the machine the ability to learn and improve from data without being programmed to do so.
There are 3 different types of AI:
Weak AI - designed to perform a specific task
Strong AI - designed to perform any intellectual task that a human can do
Superintelligence - hypothetical AI that would surpass human intelligence in all areas
Characteristics
AI systems must be able to collect data and rules. The data that they collect is processed using algorithms that enable the system to make decisions and predictions.
They must have the ability to reason so that logic can be used to evaluate information and make decisions based on it.
Lastly, they must be able to adapt and learn and can therefore change its rules and acquire more data over time.
With those in place, the AI system can adapt over time, learn from its past mistakes and improve over time adjusting its behaviour and logic.
Components
There are two main types of AI systems:
Expert system
Machine learning (ML)
Expert systems have a knowledge base to generate rules to solve problems. They have a rule base which is a set of rules that is used to apply the knowledge in the knowledge base to specific problems. They have a knowledge base and an interface for users to interact with the system.
Sample answer (Describe an expert system):
They have a knowledge base They have a rule base They have an interface for the user to interact with the system They apply the rule base to the knowledge base to solve problems (Some may have ML - they can learn and adapt themselves)
ML systems have the ability to automatically its own rules and data. It uses algorithms to analyse data and identify patterns. They system can adapt and learn over time improving itself.
Sample answer (Describe ML capabilities):
It's a form of AI, Artificial Intelligence This means it can adapt or change its own processes and algorithms It can edit its own data It can be trained, meaning it can learn without human interaction It analyses patterns and stores successful and unsuccessful results to influence future decisions to avoid errors
Subroutines
A sub-program or sub-system is used to perform frequently used operations within a program. These can be called when they are needed and can be used by other programs.
This is a question or topic you're more likely to be asked in Paper 2 (only the subroutines part is included in this note)
Last updated