Welcome to the @Ai Journal,
I’m Lem, founder of esy.com and a professional software developer for 7 years now.
I’ve been itching to jump into the Ai space for years but never seemed to have the time, now I’m going full blast, that is, every single day I will be studying some form of Ai/ML and journaling about it. I’m letting you know now, that I honestly don’t know shit outside of 3-4 videos that I watched from Andrew Ng a few years ago, which is pretty much (Supervised Learning vs Unsupervised Learning). I don’t have much of a curriculum either, so this will be an honest ‘learn as we go’ styled approach. I don’t want to get too caught up with the planning as I’m sure it’ll stress me out, so if you’re looking for order, go elsewhere, this is legit just one random guy exploring this interesting field and sharing his journey.
So what can you expect?
I’ll probably set up an outline or something later, but generally speaking I’m just going to go through a few books on the subject, try out some cool prompts in ChatGPT, play around with tools like MidJourney, and record anything I learn or do related to the subject here. Most of it will be a brain dump of takeaways and experiences relevant to Ai.
What is my goal?
As an engineer I’m deeply fascinated and curious to understand the science behind how these models loosely replicate activity in the brain, or so I’m told. I read a ‘Thousand Brains’ by Jeff Hawkins and have been following his work with Numenta so that got me super excited as well. Obviously, the Ai hype is pulling me in more than anything right now, so like everybody else I’m going to ride the bandwagon on this one with full intent to take it to the end.
Secondly, it’s a challenge to deeply understand some of these concepts so I’m going to take a stab at breaking some of them down, primarily for my own benefit - “learn by teaching”. I want to look back a year from now and see how far I’ve come on this but I hope others can get some value here as well.
Regarding content, I’ll maintain a weekly newsletter on some interesting events but I don’t want to turn into another Twitter threadboi regurgitating Ai-news bullet points. I’m more interested in studying the science and engineering details of these new technologies. I suppose there aren’t many people doing that, because it’s just too hard, can’t copy/paste, or release content as quickly for the ad algorithm, hence the personal challenge and obvious casual, journalistic style. I get a brain freeze trying to ‘stiff write’ jargon on tech topics I don’t even understand yet.
What is the curriculum?
None, really. I’m starting Hand-On Machine Learning with Scikit-Learn, Kara, and TensorFlow and just started Andrew Ngs Deep Learning Specialization. I’ll likely work on those and jump into the Fast Ai tutorial by Jeremy Howard. We’ll see how it goes? I’m also thinking maybe doing a Kaggle competition and recording my efforts there as well. Who knows, I may even discover after a few books and competitions that I don’t even like this shit, better I find out fast!
That’s all I got for this entry, in the next one I’ll cover some learnings from Hands-on ML and do a brief into LLMs, then rewind back to WTF is Machine Learning, remember what I said earlier, no-order.
Later