The AI Cheat Sheet
Explaining Artificial Intelligence
As the ability and influence of Artificial Intelligence grows, so does its associated vocabulary. Due to AI’s sheer complexity, it’s becoming more and more important to understand the language used to describe it. Now, there are different types of AI itself. But what do developers and technologists really mean when they use these terms?
An algorithm is a formula for the completion of a task. It tells or programs computers to work in a certain way by writing logic into software. Algorithms can perform calculation, data processing and automated reasoning tasks by asking a series of logical questions. They provide the foundations for artificially intelligent technology.
2. Artificial Neural Network
In AI, an artificial neural network (ANN) refers to an artificial replica of the biological networks in our own brains. Artificial neural networks are a type of machine learning, and take inspiration from neuron activity to solve problems that are too complex for traditional programming. Instead of neurons, they use interconnected nodes to simulate the nervous system. At the moment, neural networks are far less powerful than the brains of living organisms, but can perform complicated tasks like playing chess.
3. Artificial Intelligence
AI research and development aims to equip computers with the ability to make decisions and solve problems. It’s generally used as an umbrella term for any part of AI technology, but it’s actually a field of computer science. Within AI, there are three main distinctions – transformative, DIY and faux. Faux, or fake, encompasses rules based applications without machine learning capabilities. DIY is the most common form, and processes information to make suggestions for human consideration. Transformative is the next level of AI which gathers data, formulates insights and gives instructions.
Autonomy is the ability to act independently. Autonomous software, therefore, is able to function without human intervention using machine learning techniques. Examples include autonomous robots and self driving vehicles. Autonomous systems are a key driver of automation, lessening and eventually removing the need for human agents.
Chatbots are conversational interfaces powered by AI. They live in apps and handle customer queries, providing a platform for communication between companies and consumers. They use advanced natural language processing and APIs to answer questions. Chatbots represent an application of AI in customer services.
Cognitive computing mimics the way the human brain thinks by making use of machine learning techniques. Cognitive AI is therefore AI that thinks, rather than AI that follows a set of protocols. As researchers move closer towards transformative artificial intelligence, cognitive will become increasingly relevant.
7. Data science
Data science is a field of study which explores the methods used and insights gained when analysing, or mining, data. Data science is offered as a degree at numerous universities, and spans the disciplines of mathematics, statistics and computer science. It has been criticised as a buzzword, but its acceptance as an academic discipline demonstrates its growing importance.
8. Deep learning
Also known as a deep neural network, deep learning uses algorithms to understand data and datasets. It’s a subfield of machine learning that has enabled practical applications in image recognition, speech recognition, natural language processing and the environmental awareness necessary for autonomous vehicles. Deep learning feeds data to a computer via artificial neural networks, aiming to solve any problem that requires thought.
9. Machine learning
Machine learning refers to the methods and algorithms used to improve the performance of data collecting software. Although the term is sometimes used interchangeably with Artificial Intelligence, machine learning is actually a statistical approach to creating AI. In short, it’s a process of learning from examples which allows machines to adapt to new data without reprogramming. Machine learning methods include pattern recognition, natural language processing and data mining.
10. Natural language processing
Through natural language processing (NLP), machines are able to understand human language. The way people communicate is typically full of nuances and colloquialisms that are particularly hard for software to comprehend, however tech giants are all heavily invested in improving voice search. Google, for example, aims to reach human level accuracy in the NLP systems used in Google Home devices.
This brief glossary covers the AI related language you’re most likely to come across, but it’s only a snapshot of what’s yet to come. At the moment, most of the terms we use are specific to DIY AI, which collects data and insights but leaves it up to us to do something useful with them. The development of transformative, cognitive AI will inspire a new vocabulary with whole new meanings. . . . either way, keeping track of the technology which will inevitably affect your life is definitely an intelligent move.