A.I.
Artificial Intelligence
What is A.I.?
Artificial Intelligence (AQA) - The research and development of computer systems that determine the relationships between inputs and output to make predictions, instead of following programmed instructions.
Coding a system that could act like a human has been sought for hundreds of years and often found in Science Fiction. Alan Turing created a test back in the 1950's in order to detect when a system becomes intelligent. It is called the Turing Test.
Expert Systems
Systems developed to be useful for a very specific purpose
Created by asking a human expert about their decision process
Cannot learn or improve by itself
Machine Learning
Systems that teach themselves using a fitness function
Machine learning using Genetic Algorithms - CLICK HERE
Machine learning using Neural Nets - for example MarIO - CLICK HERE
Google AlphaGo beats best Go player in the World - CLICK HERE
Deep Learning
Machine learning systems that use huge datasets (Big Data) in order to learn and improve
Craiyon.ai is a tool that uses vast datasets to draw any image you ask it to - CLICK HERE
ChatBots - 3rd Generation User Interface (UI)
ChatBots are here. We currently talk to them by typing, and they tend to be created for a specific purpose. However, when we can "talk" to them and they can access solutions to "any" problems, they will start to become a very powerful digital assistant.
BBC article on ChatBots - CLICK HERE
Google have also created an explanation video - CLICK HERE
Google Assistant can make real phone calls - CLICK HERE
Microsoft Twitter Bot becomes racist - and gets switched off - CLICK HERE
Facebook bots develops its language - and gets switched off - CLICK HERE
OpenAI ChatGPT is interviewed by Channel 4 - CLICK HERE
Creating an entire website using ChatGPT - CLICK HERE
Guardian newspaper ChatGPT article - CLICK HERE
In this topic we will study:
Human/Computer interaction: User Interface design
Artificial Intelligence
ChatBots
Program design using Flowcharts
Scratch coding practice
Computational Thinking and solving problems using technology
Computer Scientists used to have a nerdy image. This was because we understood how to talk to computers when others did not. We learned how to speak "computer" using commands (CLI) and onlookers thought we were magicians. We are not.
The Graphical User Interface (GUI) made computers easier to use and more people started using them. But there is still some level of learning to do in order to make a computer work for you.
What could be done so that computers are accessible to everyone? ChatBots.
Task 1 - Create a simple ChatBot.
Using the 'Solution Cycle', plan a simple ChatBot for a new student at your school.
What questions will it be able to answer?
What time is break?
How much lunch costs?
How many lessons per day?
Once planned, create it in Scratch. Marks will be awarded for using the 'x4 Principles of Coding' and the 'x4 Sexy Code Rules'.
20 million people in China use a ChatBot as their primary friend.
Could a person have a relationship with a ChatBot? Could it solve loneliness?
But what if the power was switched off?
What if it got "upgraded" and changed?
What if it just stopped...?
ps. Check out "Him" too - CLICK HERE
Task 2 - Create a MUD (Multi User Dungeon)
A MUD is a game that asks the user to type in commands. It is not graphical, it is all in the users imagination.
To write this MUD we will use Python 3.
Sign up to Trinket.io using your school Google account.
Task 3 - Create a more sophisticated ChatBot.
Sign up to: DialogFlow.ai (using your school Google account)
<Under construction>
There are lots of large tech companies that have developed an AI engine and offer it for anyone to use. Some charge, others (Facebook) make money out of people using their system and giving lots of user data.
Watson is becoming very powerful in answering natural language based questions - CLICK HERE
IBM also provide a chatbot interface for Watson called "Watson Conversation"