Artificial intelligence is the branch of computing concerned with building systems that can perform tasks normally requiring human intelligence — such as recognising speech, classifying images, or making recommendations. AI is no longer science fiction; it shapes your daily life through search engines and voice assistants. At KS3 it is now an explicit part of the computing curriculum.
What does "artificial intelligence" actually mean?
The term artificial intelligence (AI) was coined in 1956 by the American computer scientist John McCarthy, who defined it as "the science and engineering of making intelligent machines." The word "artificial" simply means made by humans rather than occurring naturally; "intelligence" refers to the ability to learn, reason, and solve problems.
An AI system is not intelligent in the way humans are. It does not have feelings, consciousness, or genuine understanding. Instead, it is very good at identifying patterns in large amounts of data and using those patterns to make predictions or decisions. Ask it to do something outside its training and it fails — sometimes spectacularly.
What is the difference between narrow AI and general AI?
| Type | Also called | What it can do | Examples |
|---|---|---|---|
| Narrow AI | Weak AI | Performs one specific task very well | Spam filter, chess engine, voice recognition, recommendation system |
| General AI | Strong AI / AGI | Performs any intellectual task a human can | Does not yet exist |
| Superintelligence | — | Surpasses human intelligence across all domains | Theoretical only |
All AI in existence today is narrow AI. A chess-playing AI like Stockfish is unbeatable at chess but cannot recognise a cat in a photograph, hold a conversation, or drive a car. A spam filter that achieves 99.9% accuracy cannot sort your laundry.
General AI (sometimes called Artificial General Intelligence, AGI) — a system that can perform any intellectual task as well as or better than a human — does not yet exist. Despite headlines about AI, no system today comes close to genuine general intelligence.
How do AI systems make decisions?
Most modern AI systems are built using machine learning, a technique in which the system is not programmed with explicit rules but instead learns patterns from data. However, some simpler AI systems use rule-based approaches — pre-programmed if/else logic.
Rule-based AI example (simple chatbot):
IF user says "hello" THEN reply "Hello! How can I help?"
IF user says "what is your name" THEN reply "I am an AI assistant."
ELSE reply "I'm not sure I understand."
This is easy to understand and predictable but brittle — any question not covered by a rule returns the fallback.
Machine learning AI learns from examples (training data) and discovers its own rules. Given millions of emails labelled "spam" or "not spam", a machine learning spam filter learns which patterns predict spam without any human writing the rules explicitly.
What are real examples of AI in everyday life?
You encounter AI many times each day, usually without thinking about it:
| AI system | Where you encounter it | What it does |
|---|---|---|
| Spam filter | Email inbox | Classifies incoming messages as spam or legitimate |
| Recommendation engine | Netflix, YouTube, Spotify | Predicts which content you are likely to enjoy |
| Voice assistant | Siri, Alexa, Google Assistant | Converts speech to text, interprets intent, responds |
| Search engine ranking | Google, Bing | Ranks billions of pages by likely relevance to your query |
| Face unlock | Smartphone | Recognises your face in varied lighting and angles |
| Autocorrect | Keyboard | Predicts the most likely intended word |
| Fraud detection | Banks | Flags transactions that deviate from your normal pattern |
| Image recognition | Medical imaging, self-driving cars | Identifies objects, features, or anomalies in images |
What are the limitations of AI?
AI systems have significant limitations that are important to understand before concluding that AI is "intelligent":
They need vast amounts of data. A human child learns to recognise a dog from a handful of examples. An image-recognition AI may need millions of labelled photographs.
They can be fooled. Adding carefully designed noise to an image — invisible to the human eye — can cause an AI image classifier to misidentify a panda as a gibbon with high confidence. These are called adversarial attacks.
They reflect their training data. An AI trained on biased data will produce biased outputs. This is not a bug in the algorithm; it is an inherent consequence of learning from imperfect human-generated data.
They have no understanding. A large language model can write a grammatically perfect, plausible-sounding paragraph about a topic it knows nothing about. This is called "hallucination" and is a significant limitation for high-stakes applications.
What does the curriculum say about AI?
The DfE computing programme of study asks students to understand how digital technologies affect "how we communicate, collaborate, and solve problems" and to appreciate the wider implications of technology (gov.uk/government/publications/national-curriculum-in-england-computing-programmes-of-study). AI is now central to both the ethical and technical strands. GCSE specifications from AQA and OCR include AI and machine learning as topics, and Teach Computing's KS3 curriculum includes units on AI and data.
Frequently asked questions
What is artificial intelligence in simple terms for KS3?
Artificial intelligence is the ability of a computer system to perform tasks that would normally require human intelligence — such as recognising speech, identifying images, making recommendations, or playing games. Modern AI systems achieve this not through genuine understanding, but by finding patterns in large amounts of training data. All AI today is "narrow" — excellent at one specific task, unable to generalise.
Is AI the same as machine learning?
No. AI is the broader field concerned with making machines behave intelligently. Machine learning is one approach within AI, in which systems learn from data rather than being programmed with explicit rules. Other AI approaches include rule-based systems and search algorithms. Machine learning is currently the dominant technique in AI research, which is why the terms are sometimes used interchangeably, but they are not the same thing.
Can AI think for itself?
Not in the way humans do. Current AI systems process inputs and produce outputs based on patterns learned from training data. They do not have goals, desires, or consciousness. A language model that generates a creative essay is not being creative in a human sense — it is predicting the statistically most likely sequence of words given a prompt. The question of whether any machine could ever genuinely "think" is a philosophical debate that remains unresolved.
What are some dangers of AI that KS3 students should know about?
Key dangers include: algorithmic bias (AI systems that perpetuate or amplify unfairness in their training data); deepfakes (AI-generated fake images or videos that can spread misinformation); job displacement (AI automating tasks previously done by humans); privacy erosion (AI-powered surveillance and data analysis); and security risks (AI tools used by criminals to automate attacks). At KS3, the most important skill is being able to identify an AI ethical issue in a scenario and explain multiple perspectives on it.
For Socratic computing tutoring at KS3 and GCSE — see aitutors.me.