When you think of AI, a few things might come to mind, like the voice assistants that respond to your commands, the software that helps you write screenplays and the technologies that flag inappropriate content on social media. But these are only the tip of the iceberg. In fact, there are dozens of different types of artificial intelligence (AI), from the narrow AI that powers chatbots to the more broad and complicated AI used in productivity tools such as Microsoft 365 Copilot, facial recognition technology and more.

How does AI learn on its own?

Narrow AI, also known as neo-AI or weak AI, is the simplest form of the technology and often refers to systems that perform one task. AI Directory: Your Guide to AI Technology include the speech recognition features built into smartphones, image-recognition programs and tools that automate customer service responses. In many ways, these are the earliest examples of AI—the term ‘artificial intelligence’ wasn’t even coined until 1956.

Unlike non-AI computers, which are programmed to complete only a set of tasks, AI is designed to learn from its experience and adapt to new situations. The more advanced forms of the technology are referred to as strong AI or deep learning and can perform complex tasks such as image recognition, text translation and even writing songs and music.

The next step up the AI ladder are general purpose artificial intelligences, or AGIs. These have the potential to perform tasks that require human thinking and can be a key component of future robots. AGI isn’t quite ready to become a reality, however, and there are still some major obstacles that need to be overcome.