Instance PeerTube généraliste, une bonne alternative à YouTube et autres plateformes de streaming contrôlées par des géants du WEB.
General PeerTube instance, a good alternative to YouTube and other streaming platforms controlled by WEB giants.
Because in my experience I was constantly reading their unreadable code, then telling them why it’s wrong or bad or not fitting in a digestible manner and then waiting weeks for a refactor.
Iterate that for a month. Mentoring them took way longer than it would have taken me to write it from scratch. Not that dissimilar to trying to using AI for where it sucks (larger, a tad more complex problems).
It only makes sense if you look at it as an investment, because they will eventually improve.
When I was a junior, I was given an entire front-end app to develop entirely on my own with very little guidance from the team-lead. It was some ridiculously bad code, especially since it was my first time working with React with basically zero preparation.
Few months later, project is delivered, I get some time to read docs and guides before starting the next one. Since I was learning theory on what I would practise earlier, I was digesting it extremely fast and it helped me patch up all the holes in my thinking and learn how things should actually be done.
Soon after the next project came and it was definitely much more of a smooth ride. The code was alright and even the early decisions I made were pretty sustainable much later. It was another project I was working all alone, then some people joined in and I was teaching them, but I would always guide them too much and they weren’t growing very fast.
Even after a few months, these people were not ready or willing to work independently, which was my personal failure as a mentor. That’s what really assured me that people should be given a lot of space to properly grow.
My whole career is me working on increasingly larger projects with decreasing assistance. And it’s extremely effective. 4 years in the field and I just became a software architect.
I’d love to not handhold, but where I worked that was not possible, because my task came from upper management and was to handhold juniors while guarding our production code from anything they did (which often didn’t meet company standards). The tasks the juniors were getting came from somebody else.
But also, this is not really about the style (throwing into deep waters on one extreme, babysitting on the other, and everything in between) but rather that I experienced way better results from chatGPT for 50$ / month than from a junior who had under a year of experience and costs 3000-4000$ a month. Both require some degree of attention and time investment from seniors if you want to use the resulting code in prod.
So the real goal is to pay them to get better than chatGPT and hope they stay at your company long enough to get a return out of that investment.
If you just want junior-grade code, let your seniors deploy chatGPT for a much cheaper price.
You’ve worked with juniors before?
Because in my experience I was constantly reading their unreadable code, then telling them why it’s wrong or bad or not fitting in a digestible manner and then waiting weeks for a refactor.
Iterate that for a month. Mentoring them took way longer than it would have taken me to write it from scratch. Not that dissimilar to trying to using AI for where it sucks (larger, a tad more complex problems).
It only makes sense if you look at it as an investment, because they will eventually improve.
When I was a junior, I was given an entire front-end app to develop entirely on my own with very little guidance from the team-lead. It was some ridiculously bad code, especially since it was my first time working with React with basically zero preparation.
Few months later, project is delivered, I get some time to read docs and guides before starting the next one. Since I was learning theory on what I would practise earlier, I was digesting it extremely fast and it helped me patch up all the holes in my thinking and learn how things should actually be done.
Soon after the next project came and it was definitely much more of a smooth ride. The code was alright and even the early decisions I made were pretty sustainable much later. It was another project I was working all alone, then some people joined in and I was teaching them, but I would always guide them too much and they weren’t growing very fast.
Even after a few months, these people were not ready or willing to work independently, which was my personal failure as a mentor. That’s what really assured me that people should be given a lot of space to properly grow.
My whole career is me working on increasingly larger projects with decreasing assistance. And it’s extremely effective. 4 years in the field and I just became a software architect.
I’d love to not handhold, but where I worked that was not possible, because my task came from upper management and was to handhold juniors while guarding our production code from anything they did (which often didn’t meet company standards). The tasks the juniors were getting came from somebody else.
But also, this is not really about the style (throwing into deep waters on one extreme, babysitting on the other, and everything in between) but rather that I experienced way better results from chatGPT for 50$ / month than from a junior who had under a year of experience and costs 3000-4000$ a month. Both require some degree of attention and time investment from seniors if you want to use the resulting code in prod.
So the real goal is to pay them to get better than chatGPT and hope they stay at your company long enough to get a return out of that investment.
If you just want junior-grade code, let your seniors deploy chatGPT for a much cheaper price.