What is artificial intelligence?
It’s the capability of a computer system to mimic human-like cognitive functions such as learning and problem-solving.
How does artificial intelligence (AI) work?
Using maths and logic, a computer system simulates the reasoning that humans use to learn from new information and make decisions.
An artificially intelligent computer system makes predictions or takes actions based on patterns in existing data and can then learn from its errors to increase its accuracy. A mature AI processes new information extremely quickly and accurately, which makes it useful for complex scenarios such as self-driving cars, image recognition programmes, and virtual assistants.
How AI relates to machine learning
Machine learning is considered a subset of AI. Machine learning focuses on training machines to analyse and learn from data the way humans do. Therefore, machine learning is a technique that helps develop AI systems.What is machine learning
How AI relates to cognitive APIs
APIs—application programming interfaces—connect applications to other systems, services, or applications. When you use cognitive APIs, you’re requesting access to a library of domain-specific intelligent models.Learn more about Azure Cognitive Services
How AI relates to data science
AI and data science both involve gathering, analysing, and collecting large data sets—but they have different goals. AI focuses on how computers can make decisions based on data. Data science, on the other hand, focuses on the use of mathematics, statistics, and machine learning to extract insights from data.
How AI relates to robotics
A robot typically has both a physical form and the software that controls it. Robots that are controlled by AI software move autonomously—they don’t need direct instruction from a human. But not all robots are controlled by AI, and not all AI requires a physical form.
Types of artificial intelligence
Artificial narrow intelligence (Narrow AI)
Artificial narrow intelligence—sometimes called “weak AI”—refers to the ability of a computer system to perform a narrowly defined task better than a human can.
Narrow AI is the highest level of AI development that humanity has reached so far, and every example of AI that you see in the real world falls into this category—including autonomous vehicles and personal digital assistants. That’s because even when it seems like AI is thinking for itself in real time, it’s actually co-ordinating several narrow processes and making decisions within a pre-determined framework. The AI’s “thinking” doesn’t involve consciousness or emotion.
Artificial general intelligence (General AI)
Artificial general intelligence—sometimes called “strong AI” or “human-level AI”—refers to the ability of a computer system to outperform humans in any intellectual task. It’s the type of AI that you see in movies where robots have conscious thoughts and act on their own motives.
In theory, a computer system that has achieved general AI would be able to solve deeply complex problems, apply judgment in uncertain situations, and incorporate prior knowledge into its current reasoning. It would be capable of creativity and imagination on par with humans and could take on a far wider range of tasks than narrow AI.
Artificial super intelligence (ASI)
A computer system that has achieved artificial super intelligence would have the ability to outperform humans in almost every field, including scientific creativity, general wisdom, and social skills.
Machine learning is a process that computer systems follow to achieve artificial intelligence. It uses algorithms to identify patterns within data, and those patterns are then used to create a data model that can make predictions.
Machine-learning models are trained on subsets of data. When the data that’s used to train the model accurately represents the full dataset that will be analysed, the algorithm calculates more accurate results. When the machine learning model has been trained well enough to perform its task quickly and accurately enough to be useful and trustworthy, it’s achieved narrow AI.
Deep learning is an advanced type of machine learning that uses networks of algorithms that are inspired by the structure of the brain, known as neural networks. A deep neural network has nested neural nodes, and each question that it answers leads to a set of related questions.
Deep learning typically requires a large data set to train on; training sets for deep learning are sometimes made up of millions of data points. After a deep neural network has been trained on these large data sets, it can handle more ambiguity than a shallow network. That makes it useful for applications like image recognition, where the AI needs to find the edges of a shape before it can identify what’s in the image. Deep learning is also what trains AI that surpasses human skill in complex games like chess.
Examples of artificial intelligence
Businesses around the world already use AI in a wide variety of applications, and intelligent technology is a growing field. Here are some examples of AI in action today:
Some of the most complex examples of AI in the world are self-driving cars and other autonomous vehicles. These systems co-ordinate multiple processes to simulate the reasoning that human drivers use. They use image recognition to identify signs, signals, traffic flow, and obstructions. They optimise the routes they take to reach their destinations. And they send and receive data in real time to proactively diagnose issues and update their software.
Bots and digital assistants
Conversations are a natural way for people to communicate, and conversational interfaces have become more common as AI technology has advanced. Some interfaces serve a narrow purpose; people use them for one task, like booking film tickets or compiling Twitter threads into one story. Others behave more like personal assistants that can help with a wide range of tasks. But all conversational interfaces use natural language understanding (NLU) to interpret requests (also known as utterances) and reply with relevant information.
One of the most common uses for AI is to recommend items based on historical data. For example, when a media streaming service recommends what to watch or listen to next, it’s using AI to analyse what you’ve watched or listened to in the past, filter all of the available options based on their attributes, and surface the option that’s most likely to entertain you. When you’re shopping on a website and it recommends accessories or related items to add to your cart, it’s using AI in a similar way.
Many email platforms use AI to keep spam from cluttering your inbox. When a new email arrives in the system, the AI analyses it for signals that indicate spam. If the email meets enough criteria, it’s flagged as spam and quarantined. As you provide feedback—by fixing incorrect flags or flagging spam emails that weren’t caught by the filter—the system learns from that feedback and adjusts its parameters.
Smart home technology
Almost anything that automates your home uses AI. Examples include intelligent light bulbs that listen for commands, intelligent thermostats that learn your preferences and adjust themselves throughout the day, and intelligent vacuum cleaners that learn how to navigate the layout of your home without instruction.
Health data analysis
Health organisations around the world use AI to help with research, testing, diagnosis, treatment, and monitoring. Some use AI to analyse tissue samples and deliver more accurate diagnoses. Some companies use AI to analyse clinical data and discover gaps in patients’ treatment. And some companies use AI to analyse billions of compounds to help chemists reach discoveries faster and identify good candidates for clinical trials.
Benefits of artificial intelligence
AI offers real benefits that span almost every industry. Here are some of the top benefits that businesses have already seen:
Because computer systems don’t have the same biological needs as humans, an intelligent system can work throughout the day without breaks.
Communication at scale
Through bots and virtual agents, companies can provide guidance and support to more people in more places at once.
Automation of repetitive tasks
Using AI to do repetitive and time-consuming tasks helps people across your business focus on more strategic and impactful work.
Faster, more accurate decisions
AI reduces human error, which makes it helpful for decisions that are heavily informed by data and involve a lot of complex calculations.
More relevant recommendations
AI helps you make more relevant recommendations and suggestions to your customers, based on their interests and habits.
Start experimenting with AI
Build and train machine learning models, add intelligence to your apps, and turn unstructured information into searchable data through knowledge mining. Get started with 12 AI services free for 12 months.