A program that perceives its environment and takes actions to achieve goals

A human expert hired to label training data

Any physical robot equipped with a camera and microphone

A voice assistant that sets reminders and plays music

A machine that can reason competently about any subject like a human

A robot with fully general, human-level problem-solving ability

Expert systems rely on rules hand-coded by humans, while ML systems learn patterns from data

Expert systems require no human input of any kind

Expert systems can only be applied to processing images

Machine learning is a subset of deep learning, which is a subset of AI

Deep learning is a subset of machine learning, which is a subset of AI

AI, machine learning, and deep learning are unrelated fields of study

One of the earliest models of a single artificial neuron used for simple classification

A type of neural network used only for generating images

A dataset used to benchmark reinforcement learning agents

A single neuron's output value at one point in time

One complete pass of the entire training dataset through the network

The total number of layers contained in the network

The large number of layers stacked within the neural network

The size of the dataset measured in gigabytes

The complexity of the programming language used to build it

Reduce the number of layers a network needs for training

Pit two networks against each other so one learns to generate realistic data while the other judges it

Translate written text from one human language into another

Identify and classify names of people, places, and organizations within text

Convert spoken audio directly into written text

Compress large documents into much smaller files

Translating written text from one language into another

Detecting and labeling objects within a photograph

Converting spoken language into written text

Dividing an image into regions to identify the boundaries of distinct objects

Compressing an image file to reduce its storage size

Converting a color image into grayscale only

The ability to recognize a specific human face in a crowd

The ability to detect motion between frames of a video

The ability to read printed or handwritten text in images and convert it to digital text

The learner or decision-maker that interacts with the environment

The fixed dataset used to pretrain the model beforehand

The mathematical function that assigns numeric rewards

The programming framework used to build the learning algorithm

The set of labeled examples used for supervised training

The external system the agent interacts with and receives feedback from

Systematic and unfair discrimination produced by an AI system's outputs

A technique used to speed up how quickly a model trains

A method for encrypting sensitive training data

Reducing the storage size of a trained model

Making sure code runs efficiently across different hardware

Ensuring an AI system treats different groups of people equitably

Using data from equipment sensors to predict failures before they happen

Manually replacing machine parts on a fixed yearly schedule regardless of condition

Using AI purely to design the layout of a factory floor

By blocking every email that comes from an unknown sender

By learning patterns in emails to classify messages as spam or legitimate

By encrypting every incoming email message automatically

The moment computers first successfully ran a neural network

A hypothetical point where AI surpasses human intelligence and triggers rapid, unpredictable change

The point at which AI research funding is expected to peak

Genuine understanding of the meaning behind the symbols they manipulate

Access to a rulebook for generating responses

The physical ability to write characters on paper

Designing robots with human-like physical bodies

Programming an AI system using human language syntax

Attributing human traits, emotions, or intentions to AI systems that do not actually have them

That they can produce errors when given unfamiliar inputs

That they understand and consciously reason about the world the way humans do

That they are trained using large amounts of data

A system designed so a human reviews or approves key AI decisions before they take effect

A system that operates entirely without any human oversight

A system that only functions when physically wired to a human operator

Artificial Intelligence Quiz