Illustration by Steve Johnson, courtesy of Unsplash.

The last few years have seen an enormous explosion in the popularity of AI. The news is full of stories about the ways AI is going to change everyone’s lives, for better and for worse; the latest phones and computers boast exciting new AI features with names like Gemini and Copilot; even washing machines and refrigerators claim to come with AI-powered features that will make your clothes cleaner and your food fresher.

If you’re out of the loop, AI and all the new terms and vocabulary associated with it can seem unclear and confusing. That’s where the Jargon Buster comes in: below, you will find a list of all the important words related to AI with easy-to-understand definitions and relevant examples. 

Algorithm

Definition: The set of rules or instructions a computer program – not just AI – follows to solve problems and organise information. The expanded computing power provided by AI means that AI algorithms can be much more complex than those used for regular tasks.

Example: Social Media apps use algorithms to show you things you are more likely to interact with, based on what you have interacted with in the past.

Artificial Intelligence

Definition: Often shortened to ‘AI’, this is a broad term that covers all the ways computers can be trained to learn from new information and adjust how they work in response. Artificial Intelligence can be incorporated into a lot of different apps and programs to do different things.

Example: Hospitals have started using Artificial Intelligence to scan their patients’ x-rays for signs of cancer and other medical issues.

Automation

Definition: Taking something that was done manually and using technology to do it instead. AI automation allows complicated or repetitive tasks to be done by a computer on its own, often much faster than if it was done by a person using a computer.

Example: Using automation, lots of information in one format can be changed into a different format in seconds, which normally would have taken hours of manual work.

Chatbot

Definition: An AI program designed to be talked to as if it were a person. AI chatbots can be made for all-purpose interactions or for more specific uses, such as acting as a website’s customer service assistant. To avoid confusion, these chatbots are often used together with real customer service agents to handle problems that aren’t suitable for AI.

Example: Phone service providers, insurance companies and other businesses sometimes use Chatbots to answer frequently asked questions, freeing up live customer service agents for more complicated issues.

Deepfake

Definition: A video, audio clip or image of a person created with AI, designed to be very realistic. Deepfake technology is often used as a visual effect in films and television to make actors look younger, but it can also be used to depict people saying things they never said and situations that never happened.

Example: If you see something online that seems strange or outrageous, use sources you trust to find out if it’s real – it might be a Deepfake!

You can't trust everything you see online.
Image courtesy of Unsplash.

Generation

Definition: A term for an AI making something. Images, videos and text can all be generated from simple instructions depending on which AI you are using.

Example: Generating a menu for a family get-together may take some of the hard work out, but you should double-check it for accuracy and suitability.

Hallucination

Definition: Mistakes and errors that sometimes appear in generations. These happen when the AI tries to fill in missing information, or uses mislabelled or incorrect sources. You should always take hallucinations into account before doing anything suggested by an AI.

Example: A list of the top 10 restaurants in a specific town may include ones from other places, ones that have closed down, or even ones that don’t exist if the AI doesn’t have enough reliable, up-to-date sources and hallucinates.

Machine Learning

Definition: The way an AI learns for itself rather than just drawing on information provided to it. Machine Learning allows AI to infer details and draw conclusions from unclear instructions.

Example: AI programs that you can talk to use Machine Learning to match what you say to what you mean – also known as Natural Language Processing – so you never need to explain yourself!

Neural Network

Definition: A way for AI to solve problems that is modelled after the human brain. Neural Networks are better at making connections between unrelated information and identifying patterns, and can learn from things they have done before.

Example: If you ask an AI to rephrase some text you have given it, it will use its Neural Network to create a natural-sounding alternative.

Prompt

Definition: The question, request or other input you give an AI chatbot to make it do something. Prompts can be as simple or specific as you need, but more detailed prompts will often provide more useful results.

Example: The prompt “create a plan for a weekend trip to Amsterdam,” might be useful, but “create a plan for a weekend trip to Amsterdam for three people arriving by train, one of which can’t eat gluten,” will be better. 

Write a prompt into an AI chatbot and it will try to make it happen.
Image courtesy of Unsplash.

Virtual Assistant

Definition: A specific type of AI that features in phones, computers and other devices used by the general public. They can be interacted with in a variety of ways, including by writing and talking, and can be used to control a device’s functions and apps. Examples of Virtual Assistants include Siri, Alexa and Gemini.

Example: Some televisions feature Virtual Assistants, which let you change the channel, open streaming apps, select things to watch, adjust the volume and even start recordings using your voice.

Conclusion

With the help of our AI Jargon Buster, we hope that you will be able to follow and engage with conversations about AI more confidently. As the number of AI devices and features grows every year, keeping up with AI developments will only become more important. By breaking down the complex terms and concepts it uses, we hope to make the exciting world of artificial intelligence accessible to everyone.