G-Ramps

About

In detail (currently draft version): G-Ramps - Curiosity Driven Natural Mastery of AI

In brief:
The goal is to create a “help”, a platform which we are calling G-Ramps, such that by taking help from it, any learner can master AI naturally, where the only prerequisite is curiosity.

“Master AI” means having a skill which lets one master (or deeply understand) any AI algorithm naturally and a lot more, which I talked about in detail in the above linked blog.
The process of mastering AI involves learning this skill. And as explained in the blog, learning this skill is nothing but mastering the required maths and how to apply it.
One can learn this skill by first mastering some number (let's say N number) of AI algorithms. And this is where G-Ramps comes in. The blog explains how to take help from G-Ramps and master AI naturally. In the blog, I have also mentioned what I mean by “AI” and “AI algorithm”.

An AI algorithm solves some problem, which I called “main problem” in the blog. Mastering an AI algorithm means learning from first principles which is nothing but acting as a creator of the algorithm and trying to resolve the “main problem” where one only uses the knowledge which they have already mastered. And come up with the algorithm on their own. And only taking help from G-Ramps whenever needed. I have explained in the blog that to come up with the algorithm, the process of resolving, let them think of all the knowledge required to fill the knowledge gap between their currently mastered knowledge and mastery of the algorithm. Resolving is a step-by-step process and can be looked as a series of steps where the initial steps involve understanding the problem well enough and the rest of the steps involve solving it iteratively. We can also say that this series of steps is nothing but the logical derivation of an AI algorithm (explained in the blog what I mean by “logical”).

“Curiosity” here means curiosity to master all AI algorithms. And it is the only prerequisite to master AI via G-Ramps, which means one is not required to be good at the famously required maths like linear algebra, probability, statistics, multivariate calculus, optimisation etc. But to build G-Ramps, we have to assume some knowledge which almost all learners who are curious will satisfy. We assumed that a learner who is curious to master AI must already be good at school-level maths. It is a very weak assumption. Very few learners who are not good at school-level maths will be able to learn it easily from plenty of available resources. If one cannot even learn school-level maths, then there are high chances that a learner is not curious at all because school-level maths is very easy to understand and anyone with even little curiosity can understand it.

“Naturally” here means putting in required efforts for mastering an AI algorithm effortlessly. And while mastering, this only happens when there is no doubt factor about whether your curiosity is going to be satisfied or not. With G-Ramps, the process of mastering AI is natural. And after becoming a master of AI, the process of mastering any AI algorithm is also natural.
Here, let's talk about how the process of mastering AI via G-Ramps is natural. There is no doubt factor because right from the beginning, G-Ramps asks you to master N numbers of AI algorithms. It means right from the beginning you are doing what you are curious about. And it also tells you how to master an AI algorithm, i.e., learning from first principles as mentioned above. And it also contains all the help which any learner may require while learning from first principles, i.e., it contains all the steps for each AI algorithm in the series of N AI algorithms. Each learner knows that every step they take while solving moves them closer to the solution, i.e., with every step they are satisfying their curiosity of mastering an AI algorithm. And they know they can take these steps, because G-Ramps provide them the most amount of help one can get. G-Ramps realises a learner that they are naturally good problem solver and every step for any learner is easy to figure out. It really makes them feel like a creator. So, there is no doubt factor. As they take every step, their curiosity keeps getting satisfied, which makes them keep putting in the efforts effortlessly. After mastering N numbers of AI algorithms, one learns the skill mentioned above and masters AI.

This is in contrast to learning from generally available learning resources on the internet, which does not contain all the steps. And also some of them do not even mention how to master, i.e., they do not mention that one has to learn from first principles so that at least they can try to figure out steps by themselves. Learning from such resources feels like they introduce the algorithm out of the blue without explaining the problem and creating the necessity or motivation to learn the algorithm. Because of missing steps, there is a doubt factor and hence learning is not natural. Learning feels effortful and hence many learners end up doing shallow learning.

Structure

Here I am using the other word "ramp" for "learning resource" because any learning resource takes you from one level of knowledge to a higher level. Using the word “ramp” lets you easily classify different learning resources according to how steep or gentle they are. “G-Ramp” means gentle ramp, which is in contrast to steep ramp (or S-Ramp). A G-Ramp for an algorithm contains all the steps to master it, while an S-Ramp does not contain all the steps. The word "ramp” will be used to refer to either G or S-Ramp, and it will be clear which type of ramp is referred to depending on the context.

G-Ramps is the series of N "gentle" ramps where each gentle ramp (or G-Ramp) lets you master an AI algorithm naturally with curiosity as the only prerequisite (detailed explanation in the above linked blog). Each ramp is made up of a sequence of posts (or maybe I should call them "chapters") where each post either corresponds to a single step or a subsequence of steps of that ramp. Except for the first and last post. The first post in each ramp gives a brief overview of that ramp. It will talk about in brief all the steps of that ramp. Last posts called "Take Away" posts will summarise what you have mastered in a natural manner. Each ramp assumes that you have already gone over previous ramps.