Answers to some of the most commonly asked questions about AI
Artificial Intelligence (AI) is the world’s most exciting frontier for knowledge and technology. Everywhere you look, people are talking about intelligent machines improving our lives. For all of the excitement, many of the concepts and applications are still highly technical, and can be confusing if you’re not familiar with the basics of AI. If you have questions, you certainly aren’t alone!
Read on to learn answers to some of the most commonly asked questions about AI.
What is Artificial Intelligence?
This is an important first question; here are key definitions you should know:
Artificial Intelligence is a branch of computer science focused on building computers and machines that can simulate intelligent behavior. Artificial Intelligence systems are able to perform tasks traditionally associated with human intelligence, such as visual perception, speech recognition, decision-making, and translating languages.
An algorithm is a series of mathematical instructions created for a machine to follow. Think of it as simple step-by-step instructions: do A, then B, then C. In AI, programmers create algorithms that tell a computer to look at data, identify a problem, and learn from its attempts to solve the problem.
Machine learning is one of many algorithms used in AI. The machine learning field is concerned with designing programs that learn to make predictions from data, alone, without requiring assistance from a programmer. These algorithms are used in applications such as music recommendations, spam filtering, and fraud detection.
Deep learning is built on neural networks, a kind of machine learning model structured in a way that resembles neurons in a human brain. In a neural network, artificial neurons are arranged in interconnected layers. There is an input layer to receive data from the outside world, and there is an output layer which dictates how the system will respond to the information. Between these two layers, there are additional “hidden” layers of neurons, which process data by putting a numerical weight on the information they receive from the preceding layer, and passing this information to the next layer in the network. A neural network can solve very complex problems because of the huge quantity of neurons working together. Deep learning gets its name from “deep” neural networks, with dozens or even hundreds of hidden layers. These networks are powering the AI revolution with state-of-the-art object detection, machine translation, and audio synthesis.
Natural Language Processing
Natural language processing is how we get computers to understand, process, and manipulate human language. To achieve this, a computer needs to be able to “understand” a huge amount of information—from grammar rules and syntax, to different colloquialisms and accents. In a speech recognition system, for instance, human voice input becomes audio data, which then gets converted to text data, a difficult process in itself. This text data can then be used in an “intelligent” system for various applications such as translators, or controlling devices like TVs.
Computer vision is aimed at helping computers identify and process images in the same way humans do. Just as we learn to distinguish between the faces of different people, computer vision aims to teach machines to recognize different objects that it “sees” through a camera. It does this by looking at individual pixels, identifying different colors, and converting them to a numerical value, then looking for patterns so that it can identify groups of similarly colored pixels and textures. This helps it identify different objects.
Where is AI already being used?
AI is already present in many aspects of our lives. Examples include:
- Smart Assistants. Smart assistants, such as Siri, Alexa, and Cortana, use natural language processing to understand voice commands—setting reminders, finding music, answering questions, even adjusting your thermostat—all from a home speaker or your smartphone.
- Car “Autopilots.” Cars on the road today already use computer vision to operate a range of safety systems—such as tracking traffic around your car, and braking autonomously if the system perceives a danger ahead. To do this, the car needs to be able to rapidly identify different images, predict what could happen, and make a decision on what to do.
- Recommending Purchases. Popular shopping websites use AI to track what you browse, what you buy, and what you save to look at later. It then uses this information to better tailor the products and services it recommends to you. As the customer, this saves you time searching for what you want. For retailers, it means being able to predict demand for products so they have the right stock in the right places. This improves delivery times and maximizes their chances they are able to sell you something you actually need.
- Protecting your money. AI is used to constantly monitor bank accounts for potentially fraudulent activities. AI systems track all your purchases over time, and build a profile of your spending habits. The system can then rapidly flag any purchases that seem unusual. For example, if 99 percent of your purchases happen in your hometown, then suddenly a slew of purchases in another country show up, your bank can contact you to check if your card has been stolen.
- Ride-sharing. Ride-sharing apps like Uber use machine learning to accurately predict when the car you book will arrive. When your app tells you that your driver will be arriving in three minutes, machine learning has been used to analyze the data from millions of previous customer trips to hone that prediction. AI techniques are also used to determine how many cars Uber needs to have on the road at any given time, and in what areas; for example, helping ensure there are extra cars around major stations at peak commuting times.
What AI developments are set to change the world?
Here are some of the most exciting AI developments experts expect to see in the future:
- Fully automated transportation. AI will play a major role in the development of fully automated transportation systems—from self-driving cars to flying vehicles. Advanced AI systems will help vehicles react safely and intelligently to variable conditions such as other traffic, weather, and road conditions. This will result in transport that is much safer, quicker, and far less stressful than being in control ourselves. Autonomous transportation solutions will also reduce the amount of time people waste commuting through traffic, and free them up for more productive activities.
- Taking over dangerous jobs. Some jobs are inherently dangerous—such as working with hazardous chemicals. As AI develops, robots with the capacity to make intelligent, independent decisions can take over these roles and remove the need for people to risk their lives doing them.
- Faster and more accurate medical diagnosis. AI can help doctors increase the speed and accuracy by which they diagnose and treat medical conditions. Doctors will work with AI systems that can access a global database of medical conditions. The AI machine will compare patient symptoms with similar cases, and make recommendations almost instantly.
AI is poised to be the defining technology of the 21st Century, and this is an incredibly exciting time to enter this field. If you are ready to begin building your career in AI, Udacity’s School of Artificial Intelligence offers a full range of world-class programs for everyone from beginners to experienced AI professionals. You’ll learn about key AI tools and concepts, and gain critical hands-on experience building AI projects of your own. Enroll today!