Tensorflow is an AI deep learning open source software library from Google.

It is technically called anAI learning framework.

You would probably use it with the Python language on your laptop.

Now, you don’t really need to know the advanced math behind it to use it at a basic level. You could just look at some code or try to cut and paste from the library and maybe have a Raspberry Pi camera or some images and possibly get some results.

But if you think that you need to know whats under the hood then you will quickly get into some very complicated areas of AI involving discussions of matrices or arrays, vectors and inputs and outputs and 2d tensors or 3d tensors flowing god knows where!.

We will explain it in simple form soon but basically a tensor is a way to represent data in an array.

Picture a rubik cube being like a 3 by 3 by 3 array or matrix or what could be called a 3d tensor. It would either be an input 3d tensor or an output 3d tensor. And yes…it would flow…but we’ll explain that soon.

Or you could picture a game like 3d submarines or 3d chess or 3d suduko.

It all has to do with the x and y and z axis which is what they call cartesian coordinates and then there aer arrows or what they call vectors .

Now the programming will be using Python or some add on and you will be adding variables or placeholders and dealing with the arrays and setting up a loop to do the training step.

After each training loop your model gets better and eventually you test it and decide that its good enough.

So you use the graphics that come with Tensorflow and its called Tensorboard and you can see how the data looks visually. Not quite a Mona Lisa but this thing is for math nerds not flower children.

Basically all of these AI related sotware tools like Tensorflow, OpenCV GPT3, Gan etc etc have to do with the fact that we must work with HUGE amounts of data and represent it in some way usually in arrays or matrices.

Its then all about checking one thing against another like a face seen by a Raspberry Pi camera in a crowd and compared to 100,000 faces or parts of faces in a database and training your program to find the right face with a certain amount of probability that your math algorithms are using.

So just remember those few words….BIG data….organize it in arrays….throw math at it like calculus and probablility theory….and have it learn from its mistakes a million times and BOOM done deal.

You can throw a million dollars at it or hire some wiz kids in their mothers basement or a few super Wozniaks in 2 year Computer Science class at Waterloo or MIT.

Or just cry now and say IM NOT WORTHY!!!!

arrays.

(under construction Aug 19, 2021)