What is AI? AI is short for artificial intelligence. It means machines and computer programs that can imitate smart human behavior. They can make guesses called predictions, such as whether an X-ray shows signs of a tumor. AI helps phones understand the words that people say out loud. It is also used in translation apps. Those apps change words from one language into another. You may have met it already.
In 1950 the mathematician Alan Turing published a paper. It asked, "Can machines think?" In the game he described, a person asks questions. The person tries to tell which of two hidden players is the machine and which is the person. The answers are typed. The name "artificial intelligence" comes from a 1955 proposal for a summer project at Dartmouth College. The field of AI is said to have started officially in 1956, at a summer conference there.
Computers can be good at some tasks that are hard for people, such as looking through huge amounts of data very fast. In May 1997 a chess computer called Deep Blue played the world champion in a six-game match. It won, 3½ to 2½. Some game-playing programs work by brute force. They try a huge number of possible moves and replies. Deep Blue could examine about 200 million chess moves each second.
Many kinds of AI learn by looking at lots of examples. One kind of example is a picture that people have labeled as a dog. People usually give an AI the examples that it learns from. What an AI learns from is called training data. It usually needs a very large amount of training data to work well. Machine learning is one way to make AI. The computer finds patterns in data.
A large language model is the kind of AI behind many chatbots. It works by predicting which word is likely to come next. A language AI works a bit like autocomplete, but on a much larger scale. It breaks what you type into small pieces called tokens. AI can make mistakes and can say things that are not true. AI chatbots can make up facts. They can even make up the sources they name.
An AI does not think the way a person does. AI can be wrong even when it sounds sure, so a confident answer is not proof. A chatbot may not know about very recent events. It can only use what it learned from. A spam filter is a program that sorts junk email from email you want. It can wrongly flag a good email as spam. This is called a false positive.
People should check what an AI tells them before they trust it. A good way to check an AI answer is to open a new tab and look the facts up on other reliable pages. This is called lateral reading. If a chatbot names a book or article, a person should search to confirm that it really exists. An AI should not be asked to check its own work. A person should look for outside sources.
Words to know
- brute: try everything
- tokens: small word pieces
- autocomplete: word guesser
- lateral: sideways checking
- predictions: guesses
