How does artificial intelligence learn? - Briana Brownell

AI works with a simple set of instructions that creates many rules and strategies.

AI is used, for example, to help doctors diagnose patients, pilots fly commercial aircraft, and help predict traffic.

• Unsupervised learning. This approach is ideal for analysing untagged data to find general similarities and patterns. 
• Supervised learning. A computer algorithm is trained on input data to create the desired output.
• Reinforcement learning. The program uses an Iterative approach to gather feedback, then compares the data and suggests an optimal outcome. 

Each technique has its strengths and weaknesses, but when combined, researchers can build complex AI systems.

However, these models become more self-directed, making it hard for computer scientists to determine how these algorithms arrive at their solutions.

As machines continue to develop, we must consider how to teach them to teach each other to operate ethically.