3 ms·
>On the other hand, if you're a highly paid software engineer, it's likely you're already working with highly automated and abstracted systems, so there's much
by wakawaka28 2y ago
>On the other hand, if you're a highly paid software engineer, it's likely you're already working with highly automated and abstracted systems, so there's much less opportunity for LLMs.
I think you're wrong about this. There are people today getting paid lots of money for basic jobs, and people with real difficult jobs that don't get paid much. Pay rate is as much a function of location and your particular company or industry as it is the actual nature of what you are doing.
>Yes there's Copilot/etc which will reduce boilerplate, but the only way a company achieves "70% less coders" is by wholesale automation of large parts of their process that must be horribly basic already.
The basics often only seem basic because we've done them a great deal. When the machines are charged with doing a lot of stuff, will the basics still appear basic? Will you be able to spot the critical yet tricky errors buried inside reams of code for a process that you never have to develop yourself? Furthermore, will you be able to skip all the basics in your career development and work on advanced things? I have my doubts. A lot of "horribly basic" "drudgery" will have to be done manually, because it is too important to leave to the whims of a LLM and people have no other way to develop the necessary skills to sustain the industry.
- danpalmer 2y agoFair points, although I think you may be reading my points as more concrete than intended. When I'm talking about pay I'm thinking location-agnostic, i.e. low to high pay within a comparable region. For every engineer in London earning £100k there's many earning £26k doing more basic work. And when I say "basic" I don't just mean working with low level systems, I mean working with systems that are un-automated. Working with processes where someone has made the judgement call that it's cheaper to throw large numbers of cheap engineers at the problem rather than fewer highly paid engineers to automate it all away. That's not to say that highly paid engineers don't need to get into the basics, they definitely do, but that digging into the depths of an automated deployment system is not the same as running repetitive manual tasks.
- soco 2y agoHere is a question worth answering yet often overlooked: if AI helps the newbs generate useful stuff but which they don't properly understand, and also helps experienced find the issues on AI generated stuff based on their experience, WHERE is the middle ground where newbs can gain the experience to bring them to that senior level? I see no path going from there to there, unless the juniors start doing things by hand again, for which nobody will want to pay them (hey why don't you use AI instead). And seniors don't live forever...