thesmallobject.com » The Basics of Machine Learning

With machine learning, AI software learns from experience, the same way a human learns a new skill by practicing it, failing at it several times, and finally mastering it.

Programmers set up their machine learning software by giving it lots of examples, the same way ChatGPT has learned to create human-like content by browsing the web and learning from billions of different documents and web pages.

As an example, if you want to create an application that is able to detect human faces, you must give it (let’s say) 20,000 photos, with 10,000 of them being labeled ‘face’, and the other 10,000 being labeled 'not face’.

Then, when you input an unknown image into your program, it will be able to determine if it represents a human face or not based on its previous training.

The more data you feed into the program, the smarter it will be, of course. So, if your application is trained using 100,000 images instead of 20,000, it will make fewer errors.

Machine learning is very powerful because it allows computers to derive new information based on already existing knowledge. The resulting programs have a wide array of applications, ranging from weather predictions to stock market forecasts.