Читать книгу Outsmarting AI - Brennan Pursell - Страница 19

AI is just math! Math and its companion, statistics.

Coded AI systems are expressions of mathematics and logic. Statistics rely on the same. AI algorithms use calculus and linear algebra to work over data in numeric form to get results. The math can get very complicated and sophisticated, but for all that, it’s still math.

An honest AI pro tweeted: “It’s AI when you’re trying to raise money, machine learning when you’re trying to hire developers, and statistics when you’re actually doing it.” This says it all. Statistics is just applied mathematics, in AI, for data analysis.

You may well be wondering, “So if AI is just math, mathematical procedures done on numeric data, then how is AI different from data analytics, predictive analytics, data science, and big data?” Well, they are all part of the same family of algorithms performed by software. A key difference among AI algorithms is their ability to self-optimize—some would say, “to learn.” (We will revisit this important matter in chapter 2.)

The beauty of recent AI software advances is that you do not need to learn, memorize, and key in the algorithms in order to get the outputs you need from your data. Nor do you need specialized hardware. You can even build your own AI data analytics system online in the cloud by dragging and dropping elements into place.