AI

6.3 Artificial Intelligence

Understanding how machines simulate intelligent behaviour

Learning Objectives

By the end of this lesson, you will be able to:

  • Understand what is meant by artificial intelligence (AI)
  • Describe the main characteristics of AI as the collection of data and the rules for using that data, the ability to reason, and can include the ability to learn and adapt
  • Explain the basic operation and components of AI systems to simulate intelligent behaviour
  • Describe expert systems including the knowledge base, rule base, inference engine and interface
  • Explain what machine learning is and how it enables programs to automatically adapt their own processes and/or data
  • Identify the differences between AI and machine learning

Key Terms

  • Artificial Intelligence (AI)- a branch of computer science dealing with the simulation of intelligent behaviours by computers.
  • Weak AI / Narrow AI- AI trained and focused to perform a specific task or set of tasks (e.g. Apple's Siri, Amazon's Alexa, autonomous vehicles).
  • General AI / Strong AI / AGI- a theoretical form of AI where a machine would have intelligence equal to humans; it would have self-aware consciousness and the ability to solve problems, learn, and plan for the future.
  • Artificial Super Intelligence (ASI)- also known as superintelligence; would surpass the intelligence and ability of the human brain.
  • Expert system- a computer program that uses AI technologies to simulate the judgment and behaviour of a human or an organisation that has expertise and experience in a particular field.
  • Knowledge base- a repository of facts; stores all the knowledge about an area of expertise obtained from a number of expert resources. It is a collection of objects and their attributes.
  • Rule base- a set of inference rules used by the inference engine to draw conclusions; usually involves a series of 'IF' statements.
  • Inference engine- the main processing component of an expert system; acts like a search engine examining the knowledge base for data that matches queries and uses inference rules from the rules base to make decisions.
  • User interface- the method by which the user interacts with the expert system; typically includes dialog boxes, prompts and other input methods that present a series of yes/no questions.
  • Machine learning- a subset of AI; the science of training a device or application to perform a task and improve its capabilities by feeding it data, information and scenarios so it can learn over time from past experiences.
  • Deductive reasoning- where a number of correct facts are built up to form a set of rules which can then be applied to other problems.
  • Cognitive functions- mental processes such as reasoning, speech and sight that AI attempts to replicate.

What is Artificial Intelligence?

Artificial Intelligence (AI)is a branch of computer science dealing with thesimulation of intelligent behaviours by computers. AI involves replicating the cognitive functions of the human brain. These cognitive functions include reasoning, speech, and sight. Machines can replicate cognitive functions and can be measured against human benchmarks.

AI is a system that can:

Learn

Acquire new information

Decide

Analyse and make choices

Act Autonomously

Take actions without human input

Why AI Matters

Most AI can be broken down into two main categories:

Applied / Weak / Narrow AI

Most common form of AI. Designed to manage a specific task — e.g. image recognition. It is anything but weak; it enables some very robust applications, such as Apple's Siri, Amazon's Alexa, IBM Watson, and autonomous vehicles.

Generalised AI

Can evolve and improve to handle other tasks. Emerging and developing area of AI closely linked to machine learning. A theoretical form of AI where a machine would have intelligence equal to humans.

Activity 6.3A: AI or Not AI?

Difficulty: Easy • Estimated time: 5 minutes

Explain whether each of the following is an example of AI and why:

  1. A calculator performing 2 + 2 = 4.
  2. A self-driving car navigating through traffic.
  3. A chatbot answering customer questions on a website.
  4. A washing machine following a fixed wash cycle.
  1. Not AI— it simply follows a fixed mathematical rule with no learning, reasoning or adaptation.
  2. AI— it can learn, decide and act autonomously using sensors, cameras and machine learning to navigate safely.
  3. AI— it uses AI to understand and respond to typed or voice messages, learning from interactions over time.
  4. Not AI— it simply repeats the same programmed instructions over and over again with no AI involved.

Check Your Understanding: What is AI?

Artificial intelligence (AI) is a branch of computer science dealing with the simulation of intelligent behaviours by computers. It involves replicating the cognitive functions of the human brain, including reasoning, speech, and sight.

  • Learn — acquire new information
  • Decide — analyse and make choices
  • Act autonomously — take actions without human input
  • Weak AI (Narrow AI):AI trained and focused to perform specific tasks. It drives most of the AI that surrounds us today, such as Siri, Alexa, and autonomous vehicles.
  • Generalised AI:Can evolve and improve to handle other tasks. It is a theoretical form of AI where a machine would have intelligence equal to humans, with self-aware consciousness and the ability to solve problems, learn, and plan for the future.

Artificial Super Intelligence (ASI), also known as superintelligence, would surpass the intelligence and ability of the human brain.

Main Characteristics of AI

In very simple terms, the main characteristics of AI can be divided into the following categories:

1

Collection of Data

AI systems require large amounts of data to perform tasks. The data is used to train and improve the system.

2

Rules for Using Data

The data is processed using rules or algorithms that enable the system to make decisions and predictions.

3

Ability to Reason

AI systems can use logical reasoning to evaluate information and make decisions based on that information. It can change its own rules and data.

4

Ability to Learn and Adapt

AI can be described as machine learning if it possesses the ability to learn and adapt. The system improves over time from past experiences.

Summary Table

Collection of DataRules for Using DataAbility to Reason
AI systems require large amounts of data to perform tasks.The data is processed using rules or algorithms that enable the system to make decisions and predictions.AI systems can use logical reasoning to evaluate information and make decisions based on that information. It can change its own rules and data.

Deductive Reasoning in AI

Reasoning is the ability to draw reasoned conclusions based on given data.Deductive reasoningis where a number of correct facts are built up to form a set of rules which can then be applied to other problems. By carrying out a sequence of steps, the AI machine can learn, and next time it will know how to do the task more effectively and even apply it to a new situation.

Activity 6.3B: Characteristics of AI

Difficulty: Easy • Estimated time: 5 minutes

Describe the main characteristics of AI. [3]

  • Collection of data:AI systems require large amounts of data to perform tasks.
  • Rules for using that data:The data is processed using rules or algorithms that enable the system to make decisions and predictions.
  • Ability to reason:AI systems can use logical reasoning to evaluate information and make decisions based on that information. It can change its own rules and data.
  • Ability to learn and adapt (machine learning):The ability to automatically adapt its own processes and/or data.

Check Your Understanding: Characteristics of AI

  • Collection of data
  • Rules for using that data
  • Ability to reason
  • Ability to learn and adapt (machine learning)

Deductive reasoning is where a number of correct facts are built up to form a set of rules which can then be applied to other problems. By carrying out a sequence of steps, the AI machine can learn, and next time it will know how to do the task more effectively and even apply it to a new situation.

An artificial intelligence can be described as machine learning if it possesses the ability to learn and adapt — that is, the ability to automatically adapt its own processes and/or data.

Expert Systems

Anexpert systemis a computer program that uses artificial intelligence (AI) technologies to simulate the judgment and behaviour of a human or an organisation that has expertise and experience in a particular field. Expert systems attempt to mimic human knowledge and experience. In essence, they use a combination of knowledge and inference to solve problems or answer questions in a similar way to a human expert.

Expert systems are usually intended tocomplement, not replace, human experts. It is important to remember that an expert system is not used to replace human experts; instead, it is used to assist humans in making complex decisions.

Examples of Expert Systems

Equipment troubleshooting and repair

Diagnosing faults in mechanical and electronic equipment.

Technical support

Providing expert advice to users.

Medical diagnosis system

Diagnosing a patient's illness based on symptoms.

Oil and mineral prospecting

Suggesting potential new sites for resource extraction.

Tax and financial calculations

Performing complex financial and tax calculations.

Strategy games (e.g. chess)

Making strategic decisions in games.

Logistics

Efficient routing of parcel deliveries.

Identification of plants, animals & compounds

Identifying chemical and biological compounds.

Components of an Expert System

Expert systems share four key characteristics / components:

1. User Interface

This is the method by which the user interacts with the expert system. It will typically include dialog boxes, prompts and other input methods that present a series of yes/no questions based on responses to previous questions. The questions being asked usually only have Yes/No answers and are based on the responses to previous questions.

2. Knowledge Base

The system is initially set up with a repository of facts. It attempts to store all the knowledge about a specific area of expertise, which can be obtained from a variety of resources. The knowledge base is essentially a collection of objects and their attributes.

3. Rules Base

The rules base is essentially a set of logical statements, much like IF statements. These inference rules are used by the inference engine to draw conclusions. Inference rules follow logical thinking usually involving a series of 'IF' statements.

4. Inference Engine

The inference engine is the main processing component of an expert system. In essence, it acts like a search engine. It searches the knowledge base, using the inference rules from the rules base to make decisions. It is responsible for gathering information from the user by asking a series of questions and applying responses where necessary.

Structure of an Expert System


User

User Interface

Inference Engine

Rules Base

Knowledge Base

Simple Example of an Expert System

ObjectTravels land/air/waterMethod of powerNumber of wheels
TrainLandEngineMany
CarLandEngineFour
HelicopterAirEngineNone
Cruise shipWaterEngineNone
BikeLandManualTwo
PlaneAirEngineMany
Rowing boatWaterManualNone
SkateboardLandManualFour

Does it travel on water?NO→ Does it travel in the air?NO→ Is it powered manually by a human?YES→ Does it have two wheels?YES→Conclusion: It is a bike.

Setting Up an Expert System

  • Information needs to be gathered from human experts or from written sources such as textbooks, research papers or the internet.
  • Information gathered is used to populate the knowledge base that needs to be first created.
  • A rules base needs to be created; this is made up of a series of inference rules so that the inference engine can draw conclusions.
  • The inference engine itself needs to be set up; it is a complex system since it is the main processing element making reasoned conclusions from data in the knowledge base.
  • The user interface needs to be developed to allow the user and the expert system to communicate.
  • Once the system is set up, it needs to be fully tested; this is done by running the system with known outcomes so that results can be compared and any changes to the expert system made.

Medical Diagnosis Example

Input Screen
  • An interactive screen is presented to the user.
  • The system asks a series of questions about the patient's illness.
  • The user answers the questions (multiple choice or YES/NO).
  • A series of questions are asked based on the user's responses to previous questions.
Expert System
  • The inference engine compares the symptoms entered with those in the knowledge base looking for matches.
  • The rules base (inference rules) is used in the matching process.
  • Once a match is found, the system suggests the probability of the patient's illness being identified accurately.
  • The expert system also suggests possible solutions and remedies to cure the patient or recommendations on what to do next.
  • The explanation system gives reasons for its diagnosis.
Output Screen
  • The diagnosis can be in the form of text or images of the human anatomy to indicate where the problem may be.
  • The user can request further information from the expert system to narrow down the possible illness and its treatment.

Advantages and Disadvantages of Expert Systems

AdvantagesDisadvantages
High level of accuracy and expertise.Only as good as the data entered into the system.
Traceable solutions and diagnostics.Correct, optimal use requires training.
Consistent results.Assuming the system is infallible can lead to poor decisions.
Faster response times than human experts.Responses can be cold and lack the humanity of a real expert.
Can store large amounts of facts and data.Set up and maintenance costs are very high.
Unbiased reporting and analysis based on facts. 
Possibility of multiple expertise. 
Indicate the probability of any suggested solution being correct. 

Activity 6.3C: Expert System Components

Difficulty: Medium • Estimated time: 8 minutes

Describe the role of each of the four main components of an expert system.

  • User Interface:The method by which the user interacts with the expert system. It presents a series of yes/no questions based on responses to previous questions.
  • Knowledge Base:A repository of facts; stores all the knowledge about an area of expertise obtained from a number of expert resources. It is a collection of objects and their attributes.
  • Rules Base:A set of inference rules used by the inference engine to draw conclusions. They follow logical thinking usually involving a series of 'IF' statements.
  • Inference Engine:The main processing component. It acts like a search engine, examining the knowledge base for data that matches queries and using the inference rules from the rules base to make decisions.

Check Your Understanding: Expert Systems

An expert system is a computer program that uses artificial intelligence (AI) technologies to simulate the judgment and behaviour of a human or an organisation that has expertise and experience in a particular field. It uses a combination of knowledge and inference to solve problems or answer questions in a similar way to a human expert.

  • User Interface
  • Knowledge Base
  • Rules Base
  • Inference Engine

The inference engine is the main processing component of an expert system. It acts like a search engine, examining the knowledge base for data that matches queries. It is responsible for gathering information from the user by asking a series of questions and applying responses where necessary. It uses the inference rules in the rules base to draw conclusions.

Advantages:

  • High level of accuracy and expertise.
  • Consistent results.
  • Faster response times than human experts.
  • Can store large amounts of facts and data.
  • Unbiased reporting and analysis based on facts.

Disadvantages:

  • Only as good as the data entered into the system.
  • Correct, optimal use requires training.
  • Assuming the system is infallible can lead to poor decisions.
  • Responses can be cold and lack humanity.
  • Set up and maintenance costs are very high.

The user interacts with the expert system through a user interface. This typically includes dialogue boxes, command prompts or other input methods. The questions being asked usually only have Yes/No answers and are based on the responses to previous questions.

Machine Learning

Machine learningis a subset of AI — in other words, it is one way to achieve artificial intelligence. Machine learning is the science of training a device or application to perform a task and improve its capabilities — this is achieved by feeding it data, information and scenarios so it can learn over time from its past experiences.

Machine learning uses algorithms to analyse data and identify patterns or relationships. It is possible for the system to make predictions or even take decisions based on previous scenarios. They can offer fast and accurate outcomes due to very powerful processing capability. One of the key factors is the ability to manage and analyse considerable volumes of complex data.

Predictions and Decisions

Possible to make predictions and decisions based on previous scenarios.

Manage Vast Data

Ability to manage and analyse vast volumes of complex data.

Fast, Accurate Outcomes

Can provide fast, accurate outcomes due to powerful processing.

Difference Between AI and Machine Learning

Artificial IntelligenceMachine Learning
Represents simulated intelligence in machines.This is the practice of getting machines to make decisions without being programmed to do so.
The aim is to build machines that are capable of thinking like humans.The aim is to make machines that learn through data acquisition, so that they can solve new problems.

Exam Tip

Machine learning is part of AI, not separate from it. In the exam, avoid saying they're "different things". Say ML is a method used to train AI systems using data.

Examples of Machine Learning

Email filtering (spam)

Self-drive cars

Speech and image recognition

Gaming and education

Automatic tagging suggestions

Video surveillance

Google translation

Virtual assistants

Example 1: Categorising Email as Spam

Consider email messages such as 'You have won $2 million in the National Lottery'; how can machine learning determine that this email should be put into your spam folder?

  1. A machine learning algorithm collects data about emails, such as email content, headers, sender's name/email address and so on.
  2. It carries out a 'cleaning' process by removing stop words (for example, the, and, a) and punctuation, leaving only the relevant data.
  3. Certain words/phrases are frequently used in spam (for example, lottery, earn, full-refund) and indicate that the incoming email is very likely to be spam.
  4. The machine learning model is built and a 'training data set' is used to train the model and make it learn using past email known to be spam.
  5. Once it is evaluated, the model is fine-tuned and tested live.

Example 2: Detection of Fraudulent Activity

Data Collection
Data Cleaning
Exploration & Analysis
Building a Model
Model Evaluation
  • Data is gathered by a survey or web scraping; for example, to detect credit card fraud, information about customers is gathered (types of transactions, shopping habits and certain personal data).
  • Redundant data is then removed; this needs to be carefully done to remove the possibility of wrong predictions.
  • The most important machine learning step: the algorithm is trained through real examples of customer purchasing behaviour.
  • A model is built based on learning from the training data, and the machine learning algorithm can now be used to detect fraud (for example, if a customer spends an unusual amount on a piece of jewellery, there is a high chance fraudulent activity has taken place).
  • The machine learning model is then fully tested with known data and known outcomes; the system is modified if it hasn't met its criteria to detect fraudulent activity.

Example 3: Search Engines

The search engine will learn from its past performance, meaning its ability to carry out searches becomes more and more sophisticated and accurate. If the user selects one of the websites found on page 1 of the search engine's hits, the search engine classes this as a success since the relevant pages were found on page 1. If the user has to go to pages 2, 3 or 4 to find the information, the search engine classes this as a failure since the relevant pages were not found on page 1.

Advantages and Disadvantages of Machine Learning

AdvantagesDisadvantages
Reduces the need for manual work, saving time and effort.Needs vast amounts of quality data to perform well.
Can detect patterns and make predictions more accurately than humans in many cases.Requires high processing power and resources.
Continuously improves performance as it processes more data. 

Activity 6.3D: Machine Learning in Action

Difficulty: Medium • Estimated time: 8 minutes

Explain how a search engine uses machine learning to improve its results over time.

  • The search engine uses search bots to locate websites matching the user's search criteria.
  • If the user selects one of the websites found on page 1 of the search engine's hits, the search engine classes this as a success since the relevant pages were found on page 1.
  • If the user has to go to pages 2, 3 or 4 to find the information they are looking for, the search engine classes this as a failure.
  • The search engine will learn from its past performance, meaning its ability to carry out searches becomes more and more sophisticated and accurate.

Check Your Understanding: Machine Learning

Machine learning is a subset of AI. It is the science of training a device or application to perform a task and improve its capabilities — this is achieved by feeding it data, information and scenarios so it can learn over time from its past experiences.

  • AI:Represents simulated intelligence in machines. The aim is to build machines capable of thinking like humans.
  • Machine Learning:The practice of getting machines to make decisions without being programmed to do so. The aim is to make machines that learn through data acquisition so they can solve new problems.
  • Machine learning is a subset of AI — one way to achieve AI.
  1. A machine learning algorithm collects data about emails (content, headers, sender's name/email address).
  2. It removes stop words (the, and, a) and punctuation, leaving only relevant data.
  3. Certain words/phrases frequently used in spam (lottery, earn, full-refund) indicate the email is likely spam.
  4. A model is built and a 'training data set' is used to train the model using past emails known to be spam.
  5. Once evaluated, the model is fine-tuned and tested live.

Advantages:

  • Reduces the need for manual work, saving time and effort.
  • Can detect patterns and make predictions more accurately than humans in many cases.
  • Continuously improves performance as it processes more data.

Disadvantages:

  • Needs vast amounts of quality data to perform well.
  • Requires high processing power and resources.
  1. Data collection:Information about customers is gathered (types of transactions, shopping habits, personal data).
  2. Data cleaning:Redundant data is removed carefully to avoid wrong predictions.
  3. Exploration and analysis:The algorithm is trained through real examples of customer purchasing behaviour.
  4. Building a model:A model is built based on learning from the training data; the algorithm can now detect fraud.
  5. Model evaluation:The model is fully tested with known data and outcomes; modified if it hasn't met its criteria.

Key Takeaways

  • Artificial Intelligence (AI) is a branch of computer science dealing with the simulation of intelligent behaviours by computers.
  • AI can learn, decide and act autonomously — it is not just a robot.
  • Weak/Narrow AI performs specific tasks (Siri, Alexa); Generalised AI can evolve and handle other tasks; Strong AI/ASI would surpass human intelligence.
  • The main characteristics of AI are: collection of data, rules for using that data, ability to reason, and ability to learn and adapt.
  • Expert systems simulate human experts using a knowledge base, rule base, inference engine and user interface.
  • The knowledge base is a repository of facts; the rule base contains inference rules (IF statements); the inference engine applies rules to facts; the interface allows user interaction.
  • Expert systems are used in medical diagnosis, equipment troubleshooting, financial calculations, logistics and more.
  • Machine learning is a subset of AI that uses data to enable machines to be trained and learn from past experiences.
  • Machine learning is used in spam filtering, fraud detection, search engines, self-driving cars and facial recognition.
  • The difference between AI and ML: AI aims to build machines that think like humans; ML aims to make machines that learn through data acquisition.
  • Advantages of AI include increased efficiency and accuracy; disadvantages include job losses and potential for biased decision making.

Question Bank

Marking Scheme
  • Definition:A branch of computer science dealing with the simulation of intelligent behaviours by computers. [2]
  • Collection of data:AI systems require large amounts of data to perform tasks. [1]
  • Rules for using data:The data is processed using rules or algorithms that enable the system to make decisions and predictions. [1]
  • Ability to reason:AI systems can use logical reasoning to evaluate information and make decisions based on that information. [1]
  • Ability to learn and adapt:The ability to automatically adapt its own processes and/or data. [1]
Additional Notes for Slow Learners
  • Think of AI as a system that can learn from data, make decisions, and act on its own — just like a human would.
  • But AI is not a robot — it's the intelligence behind the machine.
Marking Scheme
  • User Interface:The method by which the user interacts with the expert system; presents yes/no questions based on previous responses. [2]
  • Knowledge Base:A repository of facts; stores all knowledge about an area of expertise; a collection of objects and their attributes. [2]
  • Rules Base:A set of inference rules used by the inference engine to draw conclusions; follows logical thinking involving 'IF' statements. [2]
  • Inference Engine:The main processing component; acts like a search engine examining the knowledge base and using inference rules to make decisions. [2]
Additional Notes for Slow Learners
  • Remember it like this: Theinterfaceis the door, theknowledge baseis the library, therules baseis the rulebook, and theinference engineis the detective who solves the case.
Marking Scheme
  • Definition:A subset of AI; the science of training a device or application to perform a task and improve its capabilities by feeding it data so it can learn from past experiences. [2]
  • Example 1 — Spam filtering:Collects data about emails, removes stop words, identifies spam words, builds a model, trains with past spam, tests and fine-tunes. [2]
  • Example 2 — Fraud detection:Gathers customer data, cleans redundant data, trains algorithm with real examples, builds a model, evaluates and tests. [2]
Additional Notes for Slow Learners
  • Machine learning is like teaching a child — you show them lots of examples, and eventually they learn to recognise patterns on their own.
Marking Scheme
  • AI:Represents simulated intelligence in machines. The aim is to build machines that are capable of thinking like humans. [2]
  • Machine Learning:The practice of getting machines to make decisions without being programmed to do so. The aim is to make machines that learn through data acquisition, so they can solve new problems. [2]
Additional Notes for Slow Learners
  • AI is the big idea — making machines smart. Machine learning is one method used to achieve that goal.
  • Machine learning is asubsetof AI, not a separate thing.
Marking Scheme
  • Advantage:High level of accuracy and expertise. [1]
  • Advantage:Consistent results. [1]
  • Advantage:Faster response times than human experts. [1]
  • Disadvantage:Only as good as the data entered into the system. [1]
  • Disadvantage:Correct, optimal use requires training. [1]
  • Disadvantage:Responses can be cold and lack the humanity of a real expert. [1]
Additional Notes for Slow Learners
  • Expert systems are like having a super-smart assistant who knows everything about one subject — but they can't replace human empathy and intuition.
Marking Scheme
  • Collection of data:AI systems require large amounts of data to perform tasks. [1]
  • Rules for using data:The data is processed using rules or algorithms that enable the system to make decisions and predictions. [1]
  • Ability to reason:AI systems can use logical reasoning to evaluate information and make decisions based on that information. [1]
  • Ability to learn and adapt:AI can learn from data, experiences, and interactions to improve performance over time. [1]
  • Can change its own rules and data:AI systems can adapt to new situations without human intervention. [1]
Additional Notes for Slow Learners
  • Think of AI as a system that: collects data (like gathering ingredients), has rules for using data (like a recipe), can reason (like tasting and adjusting), and can learn and adapt (like improving the recipe over time).
Marking Scheme
  • Data is gathered about customers (types of transactions, shopping habits, personal data). [1]
  • Redundant data is removed carefully to avoid wrong predictions. [1]
  • The algorithm is trained through real examples of customer purchasing behaviour. [1]
  • A model is built based on learning from the training data. [1]
  • The algorithm can now detect fraud (e.g. unusual spending on jewellery). [1]
  • The model is fully tested with known data and outcomes; modified if it hasn't met its criteria. [1]
Additional Notes for Slow Learners
  • It's like teaching a security guard what normal shopping looks like, so they can spot when something unusual happens.
Marking Scheme
  • An interactive screen is presented to the user (doctor). [1]
  • The system asks a series of questions about the patient's illness. [1]
  • The inference engine compares symptoms entered with those in the knowledge base. [1]
  • The rules base (inference rules) is used in the matching process. [1]
  • Once a match is found, the system suggests the probability of the illness being identified accurately. [1]
  • The system suggests possible solutions and remedies, and the explanation system gives reasons for its diagnosis. [1]
Additional Notes for Slow Learners
  • Think of it like a super-smart doctor's assistant that asks the right questions and helps narrow down what's wrong.
Marking Scheme
  • Advantage:Increased efficiency. [1]
  • Advantage:Increased accuracy. [1]
  • Disadvantage:Job losses. [1]
  • Disadvantage:Potential for biased decision making. [1]
Additional Notes for Slow Learners
  • AI can do things faster and more accurately than humans, but it can also take jobs and make unfair decisions if the data it learns from is biased.
Marking Scheme
  • Collects data. [1]
  • Stores rules for using the data. [1]
  • The ability to reason. [1]
  • The ability to learn / uses machine learning / adapting what it does from mistakes to not make them again. [1]
  • Result from previous decisions impacts future / by changing its own rules / by changing its own data / by being trained. [1]
Additional Notes for Slow Learners
  • AI needs data to learn, rules to follow, the ability to think logically, and the ability to improve over time.