Phase 1:

In your career, the very first stage is about gathering experience and skills. At this point, it rarely matters (to anyone, really) if you get peanuts for your work. The sad reality is that most of the 'peanut-eaners' are the ones doing all the ‘dirty’ work and that's that. All for the phrase, “I have 187 years of experience in, blah-blah-blah. ” Have you noticed how the interns are the ones who run around for everything (coffee included) and the ‘veterans’ become the ones to take all the credit? While we are all advocating for that to change (as soon as yesterday), let’s use it to learn a concept, or two, in ML.

Grab your coffee and take a ride with me!😁

Source: RODNAE Productions

Machine learning has a similar concept. The first stage has nothing to do with solving real-world problems (at least, not directly). In the first stage, we are only feeding in the predictors and the predicted. The questions and answers. The hope is to see our model ‘figure it out' and ‘gain experience.’ If anything goes wrong, “the model is not ready” and if it somehow develops a way to discover new elements, then “Brian built an awesome model.” None of the good credit is awarded to the model under normal circumstances. Much like how the entry-level intern has limited respect attached to their name.

Phase 2:

You have dropped the title, ‘intern.’ At this point, you might feel like you have made it in life because your paycheck is starting to move closer to your worth. The Senior Techies still see you in diapers but, at least, you’re not necessarily the ‘coffee guy (or girl)’ anymore. Also; you probably know better than to mess with the legacy code now! However, you are not there yet. You are not even close. You still have a lot to learn. The paycheck looks ‘better,’ but that’s still not the most important part. Sharpen those skills. If you can’t do it in your sleep, then there are a few things you need to learn. Learn them!

Source: Christina Morillo

This paragraph is mostly for those who keep finding flaws in #ChatGPT. ChatGPT is probably on phase 2 as well. Data has been fed and the training has been done over a long period. It is getting all the hype at this moment and that’s a good thing. BUT!!!: It is still learning. Of course, it won’t solve ALL your riddles, but it is learning how to. It is sharpening skills based on what is happening. Much like someone in stage 2 of their career, functionality is now there, and real-world problems are being solved in an increasingly direct manner. But there is still a lot to learn.

Phase 3:

Source: Anna Shvets

Here, you can afford to walk out of the interviewing room. Your experience is no longer a ‘conversation piece,’ everyone in the room now knows what you are capable of. All you have to do is name your price because you are now a brand. They are buying your reputation and your skills. Everything you have gathered in the past stages is now working for you, you are no longer working for it. This is where you want to be.

This is also where your machine-learning model wants to be. Where you provide the current and past stock prices as input values and get told whether to go short or long. Where you ask for a painting of a girl and get exactly that, not a boy with long hair. This is the point at which AI starts to solve problems we didn’t know we even had: The most ideal phase.

What phase are you in? How are you working on getting to the most ideal phase? Let’s talk about it down below.