The words that decide whether you command the machines or are replaced by them — explained in plain language, free, for anyone navigating the age of artificial intelligence. Each entry tells you what the term means and why it matters to staying irreplaceable. This is the vocabulary of the human edge: the three literacies that keep a person valuable when the machines can do so much.
Artificial intelligence (AI)
Software that performs tasks we once thought required human thinking — recognizing speech, writing, deciding.
A Tool, Not a Wizard: AI is not magic and not a mind; it is powerful pattern-machinery. Seeing it clearly — as a tool with real strengths and real blind spots — is the first step to using it instead of fearing it.
Large language model (LLM)
An AI trained on vast text to predict and generate human-like writing — the engine behind chatbots.
A Prediction Engine, Not an Oracle: An LLM does not 'know' things; it predicts likely words. Grasping this explains both its brilliance and why it sometimes states falsehoods with total confidence.
Machine learning
A method where software improves at a task by finding patterns in data, rather than being explicitly programmed.
Teaching by Example, Not Instruction: Instead of writing every rule, you show the machine thousands of examples and it infers the pattern. This shift is why modern AI can do things no one could hand-code.
Algorithm
A precise set of steps a computer follows to solve a problem or make a decision.
The Recipe Running Your Feed: Algorithms decide what you see, what you're offered, and sometimes what you believe. Knowing they are human-made recipes — with goals and biases — is the start of not being steered blindly.
Prompt
The instruction or question you give an AI to get a useful result.
The New Literacy: In the AI age, the quality of your answer depends on the quality of your ask. Learning to prompt well — clear, specific, with context — is becoming as fundamental as learning to search.
Hallucination
When an AI confidently states something false as if it were fact.
Confidence Is Not Truth: An AI's smooth, certain tone is not evidence it is right. Knowing it can hallucinate is the reason a sovereign user verifies — never outsourcing judgment to a machine that cannot tell when it is wrong.
Training data
The information an AI learned from — which shapes everything it can do and every bias it carries.
You Are What You Eat: An AI is a mirror of its training data, flaws included. Understanding this explains both why these tools are powerful and why they inherit the blind spots of the data they were fed.
Automation
Using machines or software to do work that humans used to do by hand.
Do the Robot-Proof Work: Automation will take the repetitive and predictable. The response is not despair but direction: invest your effort in what cannot be automated — judgment, creativity, and care.
Artificial general intelligence (AGI)
A hypothetical AI that could match human ability across virtually any task — not yet built.
The Line Not Yet Crossed: Today's AI is narrow — dazzling at specific tasks, helpless outside them. AGI would be different in kind. Knowing the distinction cuts through both the hype and the fear in the headlines.
Bias (in AI)
Systematic unfairness in an AI's output, usually inherited from skewed training data.
The Machine Is Not Neutral: People assume a computer is objective. But AI learns from human-made data and can amplify its prejudices at scale. Healthy skepticism is the defense against being misled by a 'neutral' machine.
Data literacy
The ability to read, question, and reason about data and statistics.
The Second Survival Literacy: In an age drowning in numbers and charts, the person who can tell a real statistic from a misleading one holds power. Data literacy is armor against being fooled by graphs.
Technological literacy
Understanding how digital tools work well enough to use them deliberately, not just react to them.
Command the Tool, Don't Serve It: The first of the three robot-proof literacies. It is the difference between a person who directs technology toward their goals and one who is herded by its notifications.
Human skills
The capabilities machines cannot replicate — creativity, empathy, ethical judgment, genuine connection.
The Irreplaceable Core: As machines master the technical, the uniquely human becomes more valuable, not less. The third literacy is doubling down on what only a person can do.
Generative AI
AI that creates new content — text, images, audio, code — rather than just analyzing existing data.
The Tool That Makes: Generative AI is the leap from sorting information to producing it. Used well it is a tireless collaborator; used carelessly it floods the world with convincing nonsense. The user's judgment is everything.
Deepfake
A fake image, video, or audio clip created by AI to convincingly impersonate a real person.
Seeing Is No Longer Believing: Deepfakes mean a video of someone is no longer proof they did or said it. In this era, verifying the source matters more than trusting your own eyes.
Human-in-the-loop
Keeping a person in control of an AI system's important decisions, rather than fully automating them.
Never Hand Over the Wheel: The wise use of AI keeps a human judging the high-stakes calls. Human-in-the-loop is sovereignty as a design principle — the machine advises, the person decides.
Critical thinking
The disciplined evaluation of information for truth, logic, and reliability — the master skill of the AI age.
The Skill That Makes You Dangerous to Manipulate: When machines can generate endless persuasive content, the ability to question it becomes priceless. Critical thinking is the single most robot-proof capability a human owns.
Adaptability
The capacity to learn new skills and adjust as technology reshapes the world of work.
The Only Lasting Job Security: No specific skill is permanent anymore, but the ability to keep learning is. Adaptability is the trait that turns every wave of change from a threat into an opportunity.