Loved how you explained stuff and showed examples with Parks and Rec stuff. Looking forward more more contact like this! I love the style on how you teach/explain stuff!

I've been playing around with Github Copilot for some time now, I gotta say its a pretty impressive tool that acomplishes decent results. Not the best but not the worst either. During my tests, Copilot achieved some pretty nice results and some "meh" results, mainly due to how its constructed and how it sources its model from. Github has a lot of incredible and top-tier developers, but at the same time it has newbies and people who don't really know how to code. And this is kind of the bottleneck of Copilot, Copilot could practically revolutionice coding as we know it, but the way it sources its training models is the reason why its not the "next big thing" in my opinion. Lets put it like this, you buy the world's finest spices and then you cook a 50¢ ramen block and add the spices there, what do you think its gonna happen? Yeah, the ramen will taste better but is it really the best it can get? Of course not! You could either stick with your 50¢ ramen block or spend the time and money on buying and cooking a proper ramen block and then add the spices. Github Copilot's training models work in a similar way. In one side you could have the best pieces of code written by experts and in another side you could have some newbie's janky code, and as Copilot's AI Training (sometimes) can't really differenciate what is good code and what is bad code, it results on Copilot spitting out non-sense or code that is pretty broken most of the time. This could be solved by adding some filters to Copilot and maybe adding an option to repos in the lines of "Allow Github Copilot to get data from this repository". This would solve the "Github Copilot is breaking licenses" and the low-quality code that gets fed into the training model (unless a newbie or a troll enables that and just feeds it garbage code ofc). My conclusion is this: Github Copilot is a promising tool that is on somewhat of a downward spiral due to the low-quality code on Github and on how it sometimes can't differentiate between intent (for the people who don't have access to Copilot, this happens really often). Which at the end of the day, its something that will (probably) get fixed once Copilot fully releases (if it ever releases anyways), and as of now, the only thing we can do to help Copilot. Is using it and reporting bugs and mishaps, so that stuff like this doesn't repeat in the future.

I really liked your arguments about this, I started my "real" programming journey when I was about 10, I decided to learn Python because of all the buzzwords about how Python is the best language and how Python will get me a job at some major Silicon Valley company when I were to grow up. (funny that 4 years later I'm working as another profession) After learning Python and doing some projects, I decided to learn a "big boy" language and learned C++. Learning C++'s syntax was pretty difficult, not because of how C++ is structured but because of how out of place it felt. Switching from Python to C++ hit me like a freight train, and of course objects caught 12 year old me off guard. Anyways, nowadays when some friend ask me about them wanting to learn a programming language. I just direct them to either Nim or C. Both are amazing languages, specially Nim. As it offers a taste of what "real" programming languages are, while keeping a simple to understand and fast syntax. I'd also like to add some (constructive) criticism to your post and its that Hashnode allows the use of Markdown on posts. Reading though your post was kind of hard and distracting (specially when you have ADHD heh). So I'd suggest the use of paragraphs, line breaks and some Markdown stuff (like headers and dividers). Outside of that, your writing is amazing and I'd love to see more of it.