Hi everyone,
After decades in electronics engineering, I decided to switch to computer science, with a specific goal in mind: learning Artificial Intelligence.
My journey started in 2018, and these last 4 years have been awesome, from the technological side, but, to my regret, many things were left ‘unexplained’ as is usually the case with academic programs, especially practical things.
Also, during my previous career, I have been a researcher twice, so one could say I have a creative and curious mind, always trying to see things with a new perspective and come up with original and optimal solutions.
Finally, I plan on posting exhaustive summaries on specific topics. I like exhaustivity. Too many times do I see posts with a few techniques about, say, how to detect outliers. But I kept all those articles, and one day, I’ll post an exhaustive list of all the methods that I could find. One could say that it is what I wish I could find during my studies, or now still.
1. Why now
I started this blog after a conversation with Devansh - Machine Learning Made Simple about how I solved one of his problems using a dictionary to transform a probability distribution. This will be my first post actually. Stay tuned!
2. What kind of community am I targeting
I do not plan on posting about what everybody else has posted already, like the marvels of chatGPT for instance. But maybe I’ll post about how to use it in a very specific way, or the reason why it works so well. When I have figured that out myself first… I am a contrarian, so you won’t be spared the (constructive) criticisms on why things don’t work as well as they tell you (AI has “become sentient” every time a new model can do a new trick…).
I would like to meet people who love pure and original thinking, solving difficult problems and learning state-of-the-art stuff. I won’t be posting the x different ways to use a Python generator… Or the typical (known) ways to use vectorization with Numpy. But my second post will be about how to vectorize convolutions, like in Deep Learning, by broadcasting 2D meshgrids into 6D ones. THAT is my kind of thing.
My posts will be random, depending on what I find out or the problems I have solved. Don’t expect a post every week.