AI basics: what it is and how it got here
Start here. What artificial intelligence means, how machine learning and neural networks work, where the training data comes from and what AI costs to run.
Artificial intelligence is not one technology, and the word gets stretched over a spam filter, a chatbot and a self driving car with equal confidence. What connects them is a change in how software gets built: instead of a person writing a rule for every case, a system is shown a large number of examples and works out the patterns itself. Once that idea is clear, the arguments you keep hearing about AI get easier to follow and the marketing gets easier to discount.
Read the plain definition of the term first, because it draws the line between the science fiction version and the software actually shipping. Then how a program learns from examples gives you the mechanism, and what is really inside a trained model shows you the weights and layers doing the work. Finish with where the training material comes from, which explains more about a model's habits and blind spots than any benchmark will.
People often skip ahead to whether machines will one day think like us, then argue past each other because nobody agreed what the word means. The debate about general intelligence makes far more sense after the basics rather than before them.
Articles in this hub
- What is artificial intelligence? A plain English definitionA clear definition of artificial intelligence, how it differs from ordinary software and automation, and what today's systems can and cannot actually do.
- Machine learning explained: how software learns from examplesHow machine learning works in practice: training data, patterns, predictions, and why a model that scores well in testing can still fail badly in the real world.
- Neural networks explained: what is actually inside a modelWhat a neural network really is: layers, weights, training by error correction, and why nobody can point to the part of the model that produced a given answer.
- Where AI training data comes from and why it mattersHow AI models are fed: web scraping, licensed archives, user chats and synthetic data, plus why the source shapes what a model knows, repeats and gets wrong.
- What AGI means, and why smart people disagree about itArtificial general intelligence explained without hype: the competing definitions, what today's systems still cannot do, and how to read a confident prediction about it.
- AI jargon decoded: 30 words you keep seeing explainedThe AI words that turn up in headlines and settings screens, each defined in one or two sentences: tokens, context, weights, RAG, agents, inference, distillation and more.