A traditional mindset applied to software results in the businesses we often see people complaining about. "We make the product but Sales gets all the money!" and that's very similar to your metallurgists. Code for hire as a product of labour. It takes an army of people to sell, to advertise, to make me want what it is you have (just like with music).
But a business who looks at software as complementary will see the value. A dollar spent brings many dollars back, just indirectly. It enhances the machinery that's already in place, and allows it to do entirely new things.
Where do you get that figure? The vast majority of people writing software are not paid enough salary + benefits to come anywhere close to 50% of the revenue their work is responsible for earning.
That’s the total expense, right? Not just employee remuneration.
It’s also not clear that typical grants are enough to close the gap to “more than half” as the op remarked.
Imagine there's no Version, it's easy if you try...
Really, no concept of version: you buy a piece of software as it is, take it or leave it! Either it works enough for you or it doesn't. No bugfix in the next release because there is NO next release. The musician, ehm... developer, has switched to the new shiny thing. You modified yoour environment slightly and you need a tiny change on the software to adapt? Too bad it does not exist!
Really, the point I'm trying to make is that comparing software developers to musicians doesn't hold a lot of water. Both professions include creativity, but a song shipped is as immutable as a piece of hardware. Adaptabillity is what makes software as worth as it is!
A song surely requires a skilled professional. But a game required dozens of those, with years of experience and lots fo time.
Creative industries get way lower pay because the skill level needed is low compared to many others, like law, medicine etc.
If you are talented you can paint great pictures or be a great guitar player in 1-2 years.
To be a good doctor you need what? like 10? 20 years?
First of all, even though "all in one" songwriter/producers exist today, you don't see them in the spotlight until they've already worked on their craft for many, many years. But in most cases, the commercial songs we listen to involve AT MINIMUM 1+ songwriters, 1+ producers, vocalist(s), professional instrumentalists (if it's not a band and/or a song where everything is programmed), an audio engineer, and probably a mastering engineer. And that's on top of the scaffolding of a bunch of other peoples' work (beatmakers, sample packs, recording studios/labels, etc etc). Each one of these people may not be formally educated, but you can bet they've spent years with trial and error honing their craft.
> If you are talented you can paint great pictures or be a great guitar player in 1-2 years.
What makes you think that? And if we're going to have such a low bar for "great", can't we say the same thing with programming? I mean, I know people who go to boot camp for 1 year, work for another, and then call themselves intermediate programmers and get paid $100k+. Same goes for law. I'm sure some real smart and performative people could become great lawyers within a year. The only reason that would never happen is because you literally have to go to law school for X years as mandated by law.
Don't mean to have such an aggressive reaction here, but I am really tired of the crafts being devalued so offhandedly.
If you want your [game, software, song, ...] to reach a certain level of polish and fame, you need a similar number of people working on it. Mixing, mastering, recording, composing. You need someone who's very good with sheet music and knowledgeable about all parts of an orchestra if you have any orchestral parts you want played on real instruments.
And so on.
Can I make a song in a week that'll impress a few tens of thousands people? Sure. People write in to me all the time saying "this is just what I needed!" They don't notice I banged it out in Ableton in an hour because it's what they needed. It's what I needed too, so it was easy to make it good enough. Just like you can make a game or a piece of software in a week that'll scratch a lot of people's itches even if it's not perfect.
Recently, I got back into photography, and it's the same. I can take a great picture on my own and impress a lot of people, but most famous photographers have a staff that rivals the biggest non-unicorn SV startups. Editors, lighters, second (or third) shooters, general A/V crew, marketing, office staff, etc.
Which explains why they are so in love with evergreen software that constantly changes things largely for the sake of changing them I suppose.
By contrast I'm a HW builder (mechanical engineer / industrial designer / design manager). I make good comp but maybe 70% of what my similarly leveled colleagues in SW do. The vast majority of the arguments in here still apply to my work, with a little less focus on maintenance. And yet my job role's comp is in a slow decline. Why? Because it's fully commoditized.
Anecdote: I used to work for a HW focused VC leading their internal incubator/consultancy's physical design team. One of our ML focused investments got acquired for a ton of money and every other startup in the building wanted to know how well compensated the team was going to be. It's odd to see some people at the water cooler suddenly be worth millions when your startup is still grinding, it makes for stark contrast and a very real "what if" to the onlookers. One of the Sr MEs at a struggling startup was clearly jealous and lamented to me, "man, screw machine learning, how come nobody is paying me a million bucks for my work?" To which I said "let me ask you something, when you were studying mech-e in school, did you ever use steam tables?" yeah. "Did you learn about the Carnot cycle?" yeah. "Have you considered the fact that our industry is 200 years old? It stands to reason that our jobs are not hot shit anymore. Be glad you like your work and are still paid well; ignore the new kids if you can."
This is because of leverage. Implement software that speeds up a process by 20% once, and every time that process is used from then on the speedup adds value.
We're lucky to be working in a field that enables us to add a lot of value, for relatively little capital costs, and pocket a good share of the difference.
I have this book from 1880 about steam locomotives, where the preface talks about how all the most ambitious people are going into steam engineering. Which they were, for a while.
The answer might be, until something better comes along.
I sometimes wonder how many decisions I make per day that an AI could replicate in some repeatable way. Who knows, but nothing lasts forever.
We seem to be subconsciously precocious about our fate too, as we deprecate things at a faster and faster rate.
I disagree, developer work will be automated as soon as it is technologically feasible, especially given the high cost of software development labor. I foresee AI software development platforms starting to take over new application development at some point, building applications that humans have minimal access to the source of. Existing apps by humans will likely need human maintenance until AI is able to reason about larger legacy applications.
We’re already starting to see no-code apps taking off, and AI will drive the proliferation of that technology as it becomes more sophisticated.
I don't think this is true. Take all the non-tech companies that rely on software and hire tons of engineers, such as insurance and banking.
There's SIC / NAICS / ISIC codes, or tier classifications, ranging from three (Qesnay) to five (Clark (1940), Hatt and Foote (1953), and Bell (1973)):
[A]n economy's major sectors, as delineated by Clark (1940), Hatt and Foote (1953), and Bell (1973), correspond to major stages in the essential life process. The primary sector -- agriculture, fishing, lumber, mining, oil and gas -- represents the extraction of matter from the environment to produce energy, including the calories to sustain individual organisms. The secondary sector -- processing primary goods, as in construction and manufacturing -- represents the synthesis of matter and energy into more organized forms (negentropy). The tertiary sector, including transportation and utilities, represents the infrastructure for distributing matter and energy about the system, while the quaternary sector -- trade, finance, insurance, and real estate -- constitutes a parallel infrastructure for the collection, processing, and distribution of information that is necessary in all living system for the control of material flows. Finally, the "highest" of all sectors in its remove from the physical environment is the quinary sector, including government, law, and education, representing the societal programming -- socialization, education, law making -- and collective or representative decision making to effect control.
-- James Beniger, *The Control Revolution,8 p. 179.
For participants, perhaps: unskilled labour, skilled labour, professionals, managers, finance, consumers?
Also unlike in some creative industries, software engineers tend to have higher education and are better able to understand their economic contributions.
Software solves hard problems it is constantly evolving and updating so you need lots of personnel to maintain it, with a lot of knowledge so it will never be cheap, and great engineers to create it in the first place.
Take away a bunch of artists nothing will change. Take away a bunch of critical software - the world will stop.
So no, it will last and it will last long, as the world becomes more and more technological.
It is unprecedented because software and computer world in general is a huge breakthrough and is only 30 years old. Music on the other hand is thousands of years.
Is that right?
Is there any evidence that a society can exist without art?
There’s an overwhelming body of evidence that suggests it can do so without software.
Art is a broad concept - it doesn't suggest any specific form or delivery. Software is fairly concrete - it has to be run on a computer.
A more appropriate comparison would be art versus scientific research, or art versus engineering. I think that's a more interesting question, to which I would say: perhaps we could survive without either, but I can't imagine we'd thrive without both.
Perhaps the difference is that musicians don't change a completed work. There is no maintenance cost. A song doesn't become useless and obsolete if you don't pay the musician to continually improve it.
For static software, I'd expect it to look more similar. How much of Tetris went to the people who wrote the software?
Startups pay well because speed is so important and faang pays well because their margins are enormous and the risk of those shrinking is worth paying top dollar for the absolute best talent in the world.
What's special about the software economy is that people who write the software take home a large fraction of it. Maybe more than half, depending how you count. Which is unprecedented.
It's not just because software can be zero marginal cost. Musicians don't take home half of music revenue. In most creative industries, the creatives take home less than 1/10th.
Enjoy it while it lasts.