4 ms·

> I'm not convinced you are. We totally agree here -- these are two different jobs. My point is that you can't tell that from the job title. > And can we re

> I'm not convinced you are. We totally agree here -- these are two different jobs. My point is that you can't tell that from the job title. > And can we really classify a job where you literally spend most of your time learning how to do a job a "degree required" job? Yes. There's a quantum leap between jobs where you're "learning on the job" trivial perturbations to things that hundreds of millions of other people already know and jobs where you're "learning on the job" stuff that no one / 10 other people know. ALL jobs require learning on the job. Some jobs require you to learn stuff that's not possible to learn without already having a degree or two's worth of background knowledge. Some jobs require you to learn stuff any high schooler can learn. That is the difference between mid-tech and high-tech! > There are always outliers. We have to be observing the typical or average case for this discussion to have any meaning. Yes. And I'm pointing out that understanding that typical/average case is difficult for "Software Development" because the job title encompasses so many different types of jobs.

randomdata 9y ago
> My point is that you can't tell that from the job title. Considering that you are not (primarily) a software developer, if you are being classified as a software developer in BLS data, the entire dataset is suspect. How many others are misclassified? What should we take from a dataset that is full of known inconsistencies? > Some jobs require you to learn stuff that's not possible to learn without already having a degree or two's worth of background knowledge. For instance? I'm honestly just curious about what you think is especially challenging. Everyone seems to have a different opinion about what this means. > Some jobs require you to learn stuff any high schooler can learn. That is the difference between mid-tech and high-tech! So do we agree that software development is, on average, mid-tech? And if so, why do we say it requires a degree? The software I was building in that after-school job in high school isn't a whole lot different to the dominant types of projects I see out there today. In my adult career, I've been involved with Silicon Valley startups to Fortune 500s and this seems to be the type of tech work that is most prevalent for software developers. I'll agree that a small segment of the industry is working on the fringes, doing things that only 10 other people will encounter in their lifetimes. But this is not the typical case that the BLS is drawing their conclusions from. They are looking at the entire industry as a whole.
- throwawayjava 9y ago
  > For instance? I'm honestly just curious about what you think is especially challenging Well, just look up active research labs at any top-tier university. Or peruse the work that's happening in top industry labs. And then realize that there are lots of groups within companies working on similar types of problems. And then a lot of stuff that's not "research" but still demands lots of background in some field of science/engineering in order to avoid Bad Things. Designing the control software for a new power plant, for example. > So do we agree that software development is, on average, mid-tech? I'm not sure the question is even well-formed. It's like asking if "health-care" is mid-tech. Well, certain nurses certainly are. Neurosurgeons certainly aren't. > The software I was building in that after-school job in high school isn't a whole lot different to the dominant types of projects I see out there today TBF maybe you shouldn't have been building that software without a lot more training? Maybe the lack of seriousness in the software industry is fundamentally unethical. See Equifax. As a high schooler I built software that interfaced with a database containing a shitload of social security numbers/dobs/names/addresses. I should not have been allowed to build public-facing interfaces to that database without oversight from a practiced software engineer. > But this is not the typical case that the BLS is drawing their conclusions from. They are looking at the entire industry as a whole. Well, yes, that's been exactly my point since the first post...

randomdata 9y ago

> My point is that you can't tell that from the job title. Considering that you are not (primarily) a software developer, if you are being classified as a software developer in BLS data, the entire dataset is suspect. How many others are misclassified? What should we take from a dataset that is full of known inconsistencies? > Some jobs require you to learn stuff that's not possible to learn without already having a degree or two's worth of background knowledge. For instance? I'm honestly just curious about what you think is especially challenging. Everyone seems to have a different opinion about what this means. > Some jobs require you to learn stuff any high schooler can learn. That is the difference between mid-tech and high-tech! So do we agree that software development is, on average, mid-tech? And if so, why do we say it requires a degree? The software I was building in that after-school job in high school isn't a whole lot different to the dominant types of projects I see out there today. In my adult career, I've been involved with Silicon Valley startups to Fortune 500s and this seems to be the type of tech work that is most prevalent for software developers. I'll agree that a small segment of the industry is working on the fringes, doing things that only 10 other people will encounter in their lifetimes. But this is not the typical case that the BLS is drawing their conclusions from. They are looking at the entire industry as a whole.

throwawayjava 9y ago
> For instance? I'm honestly just curious about what you think is especially challenging Well, just look up active research labs at any top-tier university. Or peruse the work that's happening in top industry labs. And then realize that there are lots of groups within companies working on similar types of problems. And then a lot of stuff that's not "research" but still demands lots of background in some field of science/engineering in order to avoid Bad Things. Designing the control software for a new power plant, for example. > So do we agree that software development is, on average, mid-tech? I'm not sure the question is even well-formed. It's like asking if "health-care" is mid-tech. Well, certain nurses certainly are. Neurosurgeons certainly aren't. > The software I was building in that after-school job in high school isn't a whole lot different to the dominant types of projects I see out there today TBF maybe you shouldn't have been building that software without a lot more training? Maybe the lack of seriousness in the software industry is fundamentally unethical. See Equifax. As a high schooler I built software that interfaced with a database containing a shitload of social security numbers/dobs/names/addresses. I should not have been allowed to build public-facing interfaces to that database without oversight from a practiced software engineer. > But this is not the typical case that the BLS is drawing their conclusions from. They are looking at the entire industry as a whole. Well, yes, that's been exactly my point since the first post...