Lesson 1 of 28 · AI foundations

What is artificial intelligence (AI)?

Artificial intelligence (AI) is the field of building computer systems that do tasks which normally need human intelligence, such as understanding language, recognising images, making predictions and solving problems. Most AI today learns these skills from large amounts of data rather than following hand-written rules. Chat assistants, spam filters, maps and recommendations all use AI.

What AI means, in plain words

Artificial intelligence is an umbrella term, not a single machine or program. It covers any computer system that does something we would normally call intelligent: understanding a sentence, spotting a face in a photo, or predicting next month's sales. The term was introduced by John McCarthy and colleagues in a 1955 proposal for a summer research workshop at Dartmouth College, held in 1956.

A helpful way to picture the difference: ordinary software follows step-by-step instructions that a programmer wrote in advance. Most AI software is instead shown many examples and works out the pattern for itself, much as a child learns what a dog is by seeing lots of dogs rather than by reading a definition.

How AI learns: rules versus examples

Early AI systems were mostly built from rules written by experts, such as "if the temperature is above 38°C and the patient has a cough, suggest a flu test". Rules are easy to read, but they break when the real world does not match them, and nobody can write a rule for everything.

Most AI today uses machine learning instead. The system is trained: it looks at thousands or millions of examples and gradually adjusts itself until its outputs match them well. The result of training is called a model, essentially a file of numbers that turns an input (an email, a photo, a question) into an output (spam or not, "cat", an answer). Using a trained model on new input is called inference.

The main kinds of AI you will hear about

AI terms are nested inside each other, like boxes within boxes:

Machine learning: AI that learns patterns from data instead of following hand-written rules.
Deep learning: machine learning with many-layered neural networks, behind most speech, image and language AI.
Generative AI: AI that creates new text, images, audio, video or code, such as chat assistants and image generators.
Narrow AI versus AGI: every AI system in use today is narrow, meaning it is good at specific tasks. Artificial general intelligence (AGI), a system as capable as a person across almost any task, remains a goal and a subject of debate.

What AI is not

AI is not a thinking mind. Today's systems find statistical patterns and use them to produce outputs; they have no feelings, intentions or understanding of the world in the way people do. A chat assistant can sound confident and still be wrong, because sounding right and being right are different things.

AI is also not the same as a robot, and not every piece of automation is AI. A robot is a physical machine, while AI is software that may or may not run inside one. A timer that switches your heating on at 7 a.m. is automation; a thermostat that learns your routine and adjusts itself is using AI.

Where you already meet AI

You probably use AI many times a day without noticing: the spam filter in your inbox, face unlock on your phone, photo search that finds every picture of your dog, arrival times in map apps, product and video recommendations, voice typing, translation and, more visibly, chat assistants such as ChatGPT.

At work, AI sits inside writing tools, meeting-note apps, customer-support chatbots, coding assistants and analytics software. Mindber's directory groups these AI products by category and task, so you can see how this course's ideas become real tools you can compare.

Example

Rules versus learning: two spam filters

Worked example

Rule-based filter (written by a programmer): IF the subject contains "free money" THEN mark as spam → misses "fr3e m0ney" and every wording nobody thought of Machine-learning filter (trained on examples): Shown 100,000 emails, each labelled "spam" or "not spam" → learns which words, senders and links tend to signal spam → flags new variations it has never seen before

Real spam filters combine both approaches. The point is the shift: from telling the computer exactly what to do, to showing it examples and letting it find the pattern.

Key points

AI is software that performs tasks that normally need human intelligence.
Most modern AI learns patterns from data instead of following hand-written rules.
All AI in use today is narrow: strong at specific tasks, not generally intelligent.
AI can be wrong, so important output still needs a human check.
You already meet AI in search, maps, spam filters, recommendations and chat assistants.

Try it on Mindber

Next lesson: What is machine learning?

Browse AI assistants in the directory

See how AI tools rank on Mindber

AI fundamentals guide

Frequently asked questions

What does AI stand for?

AI stands for artificial intelligence. The term comes from a 1955 proposal by John McCarthy and colleagues for a 1956 research workshop at Dartmouth College. Today it describes computer systems that perform tasks such as understanding language, recognising images or making predictions, usually by learning from data.

Is AI the same as a robot?

No. A robot is a physical machine; AI is software. Some robots use AI to see their surroundings or plan movements, but most AI has no body at all and runs inside apps, websites and servers. A chat assistant or a spam filter is AI without being a robot.

Can AI think like a human?

Not in the way people do. Current AI systems find statistical patterns in data and use them to produce outputs. They can write fluently or recognise images well, but they have no feelings, intentions or lived experience, and they can make mistakes that a person would easily avoid.