AI Domains

Language understanding involves analysing words, context, and intended meaning in written or spoken communication. Interpreting a written message is therefore a natural-language-processing example. Painting a wall, filling a container, and bending a key are physical activities; they may be automated mechanically but do not require an AI system to understand language.

AI can analyse medical images such as X-rays or scans and highlight patterns that may indicate a disease or abnormality. This can support trained healthcare professionals, but it does not replace clinical judgement or guarantee a diagnosis. Opening a box, cleaning a chair, and hammering a nail are physical tasks and do not represent this healthcare application.

A route-planning system can compare road distance, current or predicted traffic, road conditions, and sometimes travel time to suggest a suitable path. The system uses information relevant to movement through the road network. Wall thickness, chair colour, and paper weight do not describe traffic or route conditions, so they cannot normally guide this decision.

An AI recommendation system studies patterns in a user’s preferences, interests, and earlier choices or behaviour. It uses those patterns to suggest relevant films, songs, products, or courses. The thickness of paper, wall height, and machine weight are physical properties and do not explain what content a user may prefer.

Entertainment platforms can use AI recommendation systems to study patterns such as a user’s earlier choices, ratings, searches, or viewing history. The system then compares these patterns with available films and predicts which content may interest that user. Thus, recommending films according to personal preferences is an AI application in entertainment. The other options are simple manual activities and do not involve data-based learning or automated recommendations.

A conversational assistant interprets a user’s spoken or written request and responds through dialogue. It may answer questions, provide information, set reminders, or offer guidance, depending on its design. File organisation, machine repair, and internet-speed control are not the defining function of a conversational assistant.

Natural language understanding enables a system to interpret the meaning, intent, and sometimes context of human language. Understanding a spoken question and answering it therefore uses NLU. Identifying objects in an image is computer vision, lifting a box is robotic control, and measuring temperature is sensor-based measurement.

The AI analyses visual features and patterns in a face and compares them with authorised reference records to estimate identity. Merely storing photographs is data storage, not recognition. Battery life and manual name entry are separate hardware or human tasks. Such systems also require careful attention to consent, privacy, and accuracy.

AI systems examine data to find recurring patterns and relationships. Those patterns can support classification, recommendations, or predictions about new cases. Repeating fixed instructions does not necessarily involve learning, accurate results cannot be guaranteed without suitable data, and ethical rules should not be created without human oversight.

A spam filter learns or uses patterns in message wording, sender details, links, and other features to classify email as spam or non-spam. Turning on a computer, photocopying a document, and typing letters are fixed actions. They do not require analysing data patterns to make a classification.

An AI image-recognition system analyses visual features such as colour, shape, texture, and learned examples to identify an animal in a picture. Pushing a chair, wetting paper, and mounting a clock are physical activities. They do not, by themselves, require a system to recognise an object from visual data.

Images contain visual features such as colour, shape, texture, and arrangement. A machine can learn patterns in these features to distinguish flowers from leaves or identify different types. Drinking water, opening a door, and closing a book are physical actions and do not, by themselves, show learning from visual information.

Natural language understanding, or NLU, is a part of language-focused AI that helps a computer interpret human language. It considers words, intent, context, and relationships between parts of a sentence so that the system can respond or take an appropriate action. Therefore, understanding the meaning of a sentence is an NLU task. Painting, lifting, and heating are physical or control tasks; they do not require the machine to understand language.

A conversational assistant can interpret a user’s question or request and respond with information, guidance, or a useful action through dialogue. Its benefit is interactive assistance, although its answers may still need checking. Turning paper into stone, changing room size, or changing furniture colour is unrelated to conversational AI.

A recommendation system uses information such as a learner’s interests, earlier choices, or activity patterns to suggest content that may be relevant. Suggesting study material according to a student’s interests fits this purpose. Showing identical material, alphabetising titles, or posting a fixed list does not personalise suggestions using learned preferences.

Pattern recognition helps AI find repeated similarities, relationships, and trends in data. The system can use these learned patterns for classification, recommendations, or predictions about new inputs. It still needs suitable data and may make errors; pattern recognition does not give perfect answers without data or force one response for every input.

AI can assist trained healthcare professionals by finding patterns or possible signs of disease in reports and medical images. It supports, rather than replaces, professional judgement. Patient privacy must be protected, and every output should be checked before action. Therefore, making private data public, accepting every result blindly, or giving incorrect advice is inappropriate.

AI can combine weather forecasts, soil moisture, crop condition, and related data to recommend when irrigation may be needed and how much water could be suitable. This is decision support, so farmers should still consider local conditions and apply judgement. Hand painting, random distribution, and counting unrelated objects do not use AI for crop decisions.

A route-suggestion system can compare information such as road distance, current traffic, travel time, and sometimes road closures. AI uses these relevant inputs to estimate or recommend a suitable route. Wall colour, chair height, and paper weight do not normally describe road conditions, so they cannot guide a meaningful traffic route recommendation.

A music recommendation system can analyse listening history, likes, skips, and similarities among users to suggest songs that may interest a particular person. This data-based personalisation is an AI application in entertainment. Closing a door, folding paper, and lifting a table are physical actions and do not by themselves involve learning patterns or making recommendations.

An AI-based recognition system may compare a face or another credential with authorised records to assist access verification. Such systems must be tested carefully because recognition errors and privacy concerns can create risks. Nevertheless, verifying entry through face recognition is the only security-supporting option; ignoring errors, sharing private data, or unauthorised opening is unsafe.

A spam filter can examine words, sender details, links, and other features, then classify a message as spam or non-spam using learned patterns or rules. This is a data-based classification task associated with AI. Photocopying, pressing a switch, and printing page numbers follow direct physical or fixed procedures without such pattern-based classification.

Automation allows a system to perform suitable routine or repetitive steps with limited manual intervention. When designed and monitored properly, this can save time, reduce repetitive effort, and improve consistency. Automation does not guarantee that every result is wrong or that human effort always increases, and privacy violations are a risk to prevent, not a benefit.

Image recognition involves examining visual features such as shapes, colours, and textures and matching them with learned patterns to identify an object or animal. That is a common computer-vision application of AI. Pushing a chair, wetting paper, and placing a clock are physical actions; none requires interpreting image patterns to identify something.