Computer science

21 ideas, each explained on one page with a 60-second lesson you can play right there.

AI Alignment AI alignment steers artificial intelligence toward human goals rather than unintended shortcuts like reward hacking and strategic deception.
Artificial consciousness Artificial consciousness is the idea that a non-biological machine or artificial intelligence could have subjective, inner experiences.
Artificial General Intelligence (AGI) Artificial general intelligence is hypothetical software that can learn, reason, and solve novel problems across any domain without reprogramming.
Attention (machine learning) Attention is a method in machine learning that calculates how important every component in a sequence is relative to the others.
ChatGPT ChatGPT is an artificial intelligence chatbot that generates text, speech, and images in response to user prompts.
Chinese Room John Searle's thought experiment showing that a computer running a program can mimic human conversation without understanding it.
Embedding (machine learning) An embedding is a list of numbers that represents complex data like words or images so a computer can work with them mathematically.
Entropy Entropy is a measure of how widely energy in a system spreads out and becomes unavailable to do useful work.
Generative pre-trained transformer Generative pre-trained transformers learn broad language patterns on unlabeled data by predicting the next token before adapting to tasks.
Hallucination (artificial intelligence) An AI hallucination is a false or misleading statement generated by an artificial intelligence model and presented as true fact.
Hofstadter's law Hofstadter's law is the rule that complex work always takes longer than expected, even when you factor in the law itself.
Infinite scrolling Infinite scrolling dynamically loads content as you move down a page, replacing pagination and removing natural stopping points.
Moravec's Paradox Moravec's paradox shows why computers excel at complex logic like chess but struggle with basic human skills like vision and walking.
Neural Scaling Law Neural scaling laws describe how model error reliably drops as parameters, training datasets, and computing power increase.
Self-Supervised Learning Self-supervised learning trains AI models by turning raw data into its own practice tests, bypassing the need for manual human labeling.
Social engineering (security) Social engineering tricks people into handing over private data or access by exploiting human psychology rather than software bugs.
Spaced repetition Spaced repetition is a learning method where you review information at increasing intervals of time instead of all at once.
Technological singularity The technological singularity is a hypothetical event where self-improving AI triggers rapid, runaway intelligence growth beyond human control.
Timeboxing Timeboxing is a planning method that sets a strict maximum period of time for an activity and refuses to extend it.
Transformer (machine learning) Transformers process entire sequences at once using self-attention and positional encodings, replacing sequential neural networks.
Turing Test Alan Turing replaced the question of machine thinking with a test: can a computer converse well enough through text to pass as human?