A technology company develops software that can identify cats, dogs, and birds in photographs after being trained on thousands of examples. Although it performs this task accurately, it cannot solve mathematical problems, drive a vehicle, or perform unrelated activities.

A research team claims to have developed a machine that can learn new languages, solve scientific problems, compose music, and understand human emotions without separate programming for each task. The machine can transfer knowledge between different fields.

A bank uses software that studies previous loan applications and their outcomes. After training, the system predicts whether a new applicant is likely to repay a loan without being given specific rules for every situation.

An online store trains a model using customer information along with labels showing whether each customer purchased a product. The model then predicts future purchasing behavior based on similar data.

A supermarket analyzes customer purchase records without assigning any categories beforehand. The system automatically discovers groups of customers with similar shopping habits and buying patterns.

A robot learns to move through a maze. It receives rewards when it approaches the exit and penalties when it hits obstacles. After many attempts, it finds the shortest path to the goal.

A student creates a neural model where several input values are connected directly to a single output node. The model contains no hidden layers and performs simple classification tasks.

A medical imaging system uses a neural network with many layers to identify tumors in X-ray images. Each layer extracts increasingly complex features before making a final prediction.

A self-driving car uses cameras to recognize traffic signs, pedestrians, and other vehicles. The AI system specializes in processing visual information obtained from real-world images.

A translation application processes sentences word by word while considering the meaning of previous words. The order of information is important for generating accurate translations.

A company reuses an image-recognition model already trained on millions of pictures and adapts it to identify different species of plants. This approach reduces training time and data requirements.

A machine learning model achieves almost perfect accuracy on training data but performs poorly when tested on new data. Investigation reveals that it memorized the training examples rather than learning patterns.

A data scientist divides a dataset into several parts and repeatedly trains and tests a model on different combinations of these parts. The goal is to estimate performance on unseen data.

A hospital uses an AI system that analyzes symptoms and medical records to assist doctors in diagnosing diseases. The system imitates the reasoning process of experienced medical specialists.

An organization uses an AI hiring system trained on biased historical data. As a result, qualified candidates from certain groups are unfairly rejected despite having suitable skills and experience.

A busy office worker uses a virtual assistant on her smartphone to set reminders, schedule meetings, search for information, and send messages through voice commands. As a result, she spends less time on routine tasks and can focus on more important work.

A streaming platform analyzes a user's viewing history, favorite genres, and watch time patterns. It then suggests movies and television shows that closely match the user's interests, increasing satisfaction and engagement.

A hospital installs an AI-powered imaging system that examines X-ray and MRI scans. The system highlights suspicious areas for doctors, helping them identify diseases such as cancer at an earlier stage.

A credit card company continuously monitors customer transactions. When an unusual purchase pattern appears, the system immediately flags the transaction and alerts security staff before significant financial loss occurs.

A school introduces intelligent tutoring software that adjusts lesson difficulty according to each student's performance. Faster learners receive advanced exercises while struggling students receive additional support and explanations.