The thing is the market for junior devs is very saturated right now. Things have changed a lot over the past two years.
1 Like
Yes, five years ago we could not find any local react developers. Now it is fairly easy to find someone with a few years experience (Nike has been laying lots of people off).
That is why I recommended Javascript. If you have that in your quiver; you are well positioned to adopt whatever the technical âflavor of the dayâ is at any one point in the market.
Javascript is also used on both the client and the server. This allows you to be very flexible.
People who are motivated to learn things on their own are diamonds in the rough, though. It is very obvious if someone is passionate about software development based on the projects they chose/choose to work on. I would prefer to work with someone who is self-taught but enjoys writing software to someone who has a B.S. in computer science but only did the work required for the degree.
1 Like
I should add that I primarily use C/C++ for work, but I do use python for prototyping and automating repetitive tasks. Python is also my preferred language for personal projects.
1 Like
Old C++ engineer here. Started coding C++ on the glockenspiel compiler back in the day.
My first programming experience was in the 90s on TI-82 and TI-83 calculators. I was trying to learn TI Basic on my own. I didnât understand the IF/THEN/ELSE syntax, so I used a lot of GOTO statements instead. I wrote some really ugly (but functional) code back then which I later rewrote after taking a C course in college.
I started with assembly code on a PDP-11.
2 Likes
As an allegedly senior software engineer this is one of the skills Iâd say is most important. Really self starting anything is key to success in software. Itâs usually cheap to prototype things, so have small ideas and explore them a little. Always be coding. I donât get as much coding time as I used to so I occasionally take on a 2-4 week side quest to touch code.
1 Like
I actually think this is a highly reasonable perspective. Generative AI is new enough that all the classes of security threats are not really well understood yet.
I agree that adoption is a matter of when and how, not if, and that weâre going to get there a lot faster if the large software companies keep pushing them front and center.
To bring the convo back to generative AI, Python is the primary language driving advances in that space, which is the new hotness in software, making it a super opportune time to learn it.
1 Like
Is python used for the actual AI network? Or are you referring to writing programs to use the AI network? I thought Cuda, C++ was the primary tool to actually implement the generative AI. Am I mistaken?
I thought most neural nets etc are implemented using python.
Itd be interesting to learn this is not the case! Haha.
The last time I developed neural nets (way back in 2005) I used C++. CUDA is a library that allows you to program the graphics card memory. We used it back then for neural-networks for trading. It was all C++. We also used it to speed complex financial calculations like black-scholes and synthetic options construction.
Before CUDA we had to use proprietary libraries for each graphic card to do this work.
1 Like
Python is only the wrapper. Would be way too slow
1 Like
Most cool python libs are c or c++ e.g. numpy, matplot, pytorch, tensorflow and so on
1 Like
Python is definitely the most popular tool for scientists and researchers, statisticians, engineers, etc.
My mechanical-engineer son works for Intel and writes software for his tools using python. There seems to be a library for just about everything.
2 Likes
Yeah is crazy how much python stuff is flying around. It is used really in all areas you can think of.
I think it mainly is so popular since you are super quick producing results through simply language and lib support. Is like JavaScript make things quick and can get out of hands if not planned properly for bigger applications.
I studied ML for a year now while working and implementing AI features for our company. Python is definitely the standard right now for ML but also other languages start to picking up quickly now which I would prefer like kotlin and swift. Simply because of stronger typing for IDE/compiler support and refactoring
1 Like
Also, JavaScript, WTF?
Haha yeah it is crazy 
My fav from back in the days is this one
Could not find the video anymore
1 Like