@esrtweet I think part of it is also the quality of the open source code written in a language that can be used to train the robot friends. Golang code for a variety of reasons tends to be pretty high quality. So the generated code tends to inherit that quality.
@BarbellFi Because inflation can eat your lunch that way. If you bought $1m in 30 year bonds in 1965 by 1995 your returned principle was worth ~$212k in 1965 dollars. You lost almost 80% of your principle for a 4-5% yield.
@asymmetricinfo I actually tested this in grad school. What we found was that dominant cursive writers wrote faster in cursive, dominant printers wrote faster printing, and there was no statistically significant speed difference between dominant cursive or print writers.
@Chaos2Cured@yacineMTB Mostly I have just made a *huge* number of errors at high rates... and learned from at least some of them, sometimes after repeating the mistake for the second or third time π
@Chaos2Cured@yacineMTB People who are incredibly competent at only one thing can easily miss the work involved on digging in on other things. Once you hit at least two, they recognize the work needed, and dig in properly.
@Chaos2Cured@yacineMTB My experience is likely a slight nuance on your observation: people who are incredibly competent are at least two things are incredibly competent at anything they turn their mind to. The 'at least two' is crucial.
@ibuildthecloud Sort of like mutexes in Golang. Are they sometimes the right tool. Sure. But infrequently enough that you should be suspicious whenever you see them in a PR.
@mccricardo More critically, as engineers, itβs our duty to ask ourselves whether a pattern applies to a problem at all. Much mischief has been wrought by engineers with a limited tool box of patterns beating on a problem with one of them instead of seeking out a more appropriate tool.
@CTOAdvisor I do am concerned about how we train seniors without the experience of being junior. There's a lot of learning that goes into being able to get good results out of GenAI (and correct its mistakes). If we just replace the juniors with AI, where do the next seniors come from?
@CTOAdvisor If someone is really *up* to performing at a more senior engineer level but is in a junior role, GenAI code gen can look like its making them more senior... but its just enabling them to show they are more senior. 2/
@CTOAdvisor GenAI for code tends to be useful in the hands of developers senior enough on a topic to 'supervise' it properly. Seniors will often spend a lot of their time 'supervising' more junior engineers who write a lot of the code. GenAI can replace those juniors. 1/