AI, Data Centers, and the Hardware that runs it





It goes without saying that Artificial Intelligence is the biggest and most significant advancement for mankind in our lifetime. With the potential to accelerate human development exponentially going forward, the technology is poised to find solutions to many questions that humans have asked and pondered since the emergence of human consciousness. However, with great power comes great responsibility. We are developing something that has the chance to rival our species and possibly surpassing, and destroying, everything we have become.

We are potentially playing with God, becoming it and creating it.

...But we are getting way ahead of ourselves here. This is all speculation and in actuality we are not even close to this supposed Artificial SuperIntelligence yet. So breath easy! With the introduction out of the way, this website will dive into Artificial Intelligence. We will learn about its basic processes and some of the basic models that make it run so effectively. We will discuss the Data Centers and the dangers that they pose to us which has made headlines as of recent. An argument can be made that the Data Centers are bigger threat to us than the AI it stores. Lastly, we will go into the hardware, the chips, that powers and makes AI possible and some of the economics of it. With all of that out of the way, let's dive into it!


The Basics of AI

AI is a type of technology that was created to replicate human behaviors, some examples would include decision making, speech recognition, and problem solving. It's able to accomplish these human-like abilities because of three things:
  • Algorithms
  • Data
  • Statistical Models
 The Markov Decision Process

Algorithms are the coded instructions that guide the computer or machine on what to do. These algorithms are fed data such as images, academic papers, and even models that they've carried out before with the recorded results. These three basic tools then drive the next three broader concepts:
  • Machine Learning
  • Neural Networks
  • Deep Learning
 A basic Neural Network comprised of 3 layers A neuron mirrors a neural network almost to a tee

Neural Networks are computational models that mimic neurons of a human brain. Neural Networks are comprised of three layers: an Input, a Hidden, and an Output. As programmers, we're familiar with how inputs and ouputs work with computers. The hidden layer though is a little more intricate. This is where all the big algortihms and mathematical models reside. Deep Learning models are a bunch of Neural Network models with hundreds of these hidden layers. The connections between neurons, called weights, process data through mathematical algorithms and deliver an eventual output based on the input and where over time and data become more accurate.

Machine Learning

Machine Learning is the one is the one I have yet to cover but is arguably the most important one which is why I will not be addressing it but including a video on the topic. Enjoy!


HTML Validation