Core Concepts
Technology that performs tasks requiring human-like intelligence.
AI
Learning from data with known answers.
Supervised learning
The decade when AI became a formal research field.
1950s
The researcher who coined the term artificial intelligence.
John McCarthy
When AI generates false or fabricated information.
hallucination
A subset of AI that learns patterns from data.
Machine Learning
Learning from data without predefined labels.
Unsupervised learning
The test proposed to evaluate machine-like human conversation.
Turing Test
The researcher known for the ImageNet project.
Fei-Fei Li
The risk of unfair results from skewed training data.
AI bias
AI that creates text, images, or other content.
Generative AI
Learning through rewards and penalties.
Reinforcement learning
The 1956 conference that helped establish AI.
Dartmouth workshop
The researcher who pioneered convolutional neural networks.
Yann LeCun
Data used to evaluate a model on unseen examples.
test data
The basic processing units of artificial neural networks.
Artificial neurons
Learning using both labeled and unlabeled data.
Semi-supervised learning
A period of declining AI funding and interest.
AI winter
The lead author of Attention Is All You Need.
Ashish Vaswani
The principle that correlation does not prove this.
causation
Machine learning that uses multiple neural network layers.
Deep learning
An unsupervised method that groups similar observations.
Clustering
The architecture introduced in 2017 that transformed language models.
the transformer
The two researchers awarded the 2024 Nobel Prize for AlphaFold.
Demis Hassabis and John Jumper
When a model performs well on training data but poorly on new data.
overfitting