Pete 2_Big_Data.docx

**MEGA-TREND 2: BIG DATA** We've just heard about the "Internet of Things" and how it's going to grow. That goes hand-in-hand with our second mega-trend: Big Data. Unless you live under a very, very analogue rock, you're already familiar with the term "Big Data". But I want you to think about this...

**MEGA-TREND 2: BIG DATA** We've just heard about the "Internet of Things" and how it's going to grow. That goes hand-in-hand with our second mega-trend: Big Data. Unless you live under a very, very analogue rock, you're already familiar with the term "Big Data". But I want you to think about this for a second. The International Telecommunication Union -- so, the branch of the United Nations that is specialized in information and communication technologies -- has forecasted that by 2030, global monthly data traffic will be 84 times higher than it was in 2020. So here's where we were in 2020, and here's where the United Nations expects us to be in 2030. Let's make that a little more tangible. Here is the lovely European country of Austria. And here is the lovely Latin American country of Brazil. That's the relative size growth the International Telecommuniation Union is talking about. If 2020 global monthly data traffic is the size of Austria, 2030 global monthly data traffic will be the size of Brazil. Mainly -- but not only -- because of the growth in the Internet of Things, Big Data is about to get a lot bigger. "So what?" you may be asking yourself. "My job has nothing to do with building data networks. Why should I care?" You should care because this data explosion is going to catapult the quality of Artificial Intelligence and Machine Learning lightyears beyond what it is today. For any type of Artificial Intelligence to be useful, it needs to draw upon accurate and relevant data. The more such data it has, the better it will be able to identify patterns, make predictions, and provide recommendations. In the past, the amount of data that Artificial Intelligence worked with was limited by two factors: the speed at which computers were able to PROCESS data, and the availability of the data itself. Modern day processors have freed AI of the first of these limitations: computers can now process billions of units of data per second. But until recently, the SECOND limitation -- access to enough accurate and relevant data -- has required Artificial Intelligence to limit its computations to statistically significant SAMPLES of data. So let's take an example where we want to use AI to make a prediction about how EU households will respond to a new environmental law. In the past, instead of looking at the data of ALL the households in the EU, Artificial Intelligence would consider the data of a SUBSET of these households that has been selected in a way that makes this subset statistically repreSENTATIVE of the whole. This data would then be analyzed by the AI and -- using statistical methods - it would make predictions about how EU households would react to the new law. That's already pretty astounding. But statistical methods are ALWAYS subject to sampling errors and "confidence limits". What this means -- for the purpose of our discussion -- is that there's ALWAYS a certain probability that conclusions drawn on the basis of statical methods will be... wrong. And here's where Big Data comes in. As more and more data becomes available -- so, Brazil instead of Austria -- Artificial Intelligence won't NEED to draw upon statistic samples -- so SUBSETS -- of a dataset anymore. It will have access to the ENTIRE dataset. And computer processors will be fast enough to sift through that entire dataset in the blink of an eye. We'll talk more about what that means for the future of AI in a later video. HERE, we can preemptively say that AI will be able to recognize patterns, spot outliers, and identify trends with an accuracy and at a speed that human beings simply can't match. Obviously, the forthcoming data explosion will bring its own set of challenges, such as making sure that the data is accurate and trustworthy, and that protective data privacy is protected. One CENTRAL challenge related to Big Data is the question of where organizations will STORE or ACCESS all this data. And that's where our THIRD mega-trend, Cloud Computing, comes in.