3 ms·

First jobs were replaced by simple tools, then by livestock, then by machines. What makes software any different from any other advancements made throughout hi

by joobus 9y ago

First jobs were replaced by simple tools, then by livestock, then by machines. What makes software any different from any other advancements made throughout history?

dragonwriter 9y ago
All of the earlier advances require human maintenance that scale with the scope of deployment. The whole reason software is a hot area of investment is that it's support requirements don't scale that way, which means is good for the capitalist from a cost perspective, but bad for the wage-labor dependent class.
defined 9y ago
It's the only advancement where making a million copies of an item takes a few minutes and no raw material. Try that with tools, livestock and machines. Or anything physical.
- convolvatron 9y ago
  but before that you have to go through several iterations of failure with a bunch of costly entitled developers before you get something which sort of approximates what you wanted and is so internally broken its almost impossible to extend not really ragging on the developers, but its not a cheap or risk free process. and if you aren't actively investing in maintenance it will die. the costs are all just per-type not so much per-instance
  - defined 9y ago
    > but before that you have to go through several iterations of failure with a bunch of costly entitled developers before you get something which sort of approximates what you wanted and is so internally broken its almost impossible to extend This is too often true, subject to the following observations: - Project failure is almost always a management failure (no citations needed, really), because: - The thing to be built is poorly conceived and explained by the leadership/stakeholders, who are seldom the devs - The delivery schedule is wildly optimistic or infeasible, because unrealistic promises were made to customers or investors - The operational parameters of the system (peak users, load, transactions, data size, availability, etc) are overestimated by the leadership by a couple of orders of magnitude and pushback tends to be career-limiting - When the leadership eventually realizes the failure of its vision, it "pivots", and keeps pivoting at the speed of a turbine, "reframing" the so-called vision. - The "costly, entitled" developers often come from some body shop which pays them $25/hour and charges $75 (illustrative values only) - Costly, unentitled, competent developers probably ran for the hills or were rejected because they weren't "with the program" - they said something realistic in the interview. - The process is not cheap or easy or risk free, yet many companies hire and pay as if it is. - The mere act of getting cohesive requirements out of people is one of the hardest parts of software, and one of the largest reasons for failure. - The rest isn't easy, either, unless it's a conceptual copy of some vanilla CRUD system that does nothing novel or "web scale" (whatever that is). OTOH, I read some devs stating that software development is just bolting legos together, what's the problem? If you are referring to these devs, I see your point, but they and their ilk are the creations of corporates and startups who pushed the view that devs are commodity items - resources - who aren't really skilled and can be churned out of bootcamps and will be good enough to meet the deadlines. Now I am not criticizing bootcamps or the people who make it through them. There are many smart people who have no formal CS background who could make good devs, given time and experience. There are even more people who should not be allowed near a computer, sadly, who IMO are more commonly encountered. I am criticizing the social forces that put money above useful education, profit above knowledge, ideology above science and engineering, the short term against the entire future.
- dragonwriter 9y ago
  > It's the only advancement where making a million copies of an item takes a few minutes and no raw material. And, mire directly to the point, doesn't require additional labor inputs. (If extraction is automated by software tools driving hardware, even things that demand raw materials can have this feature.)
chrisdbaldwin 9y ago
The scale of software tool utility is pretty much unparalleled by traditional tools. Traditional tools such as livestock and machines require breeding, mining... some sort of "manufacturing" process. Okay, software has manufacturing cost too, but typically only once to create, and once to maintain. Imagine if you could dynamically spawn livestock without the overhead of mining, breeding, and trading. The jobs that software has/will eat are, imo, beyond current human comprehension.
- evgen 9y ago
  Electrification was far more significant and far-reaching, occurred on a similarly quick timescale, and led to a massive boom in productivity across the board. If you want to predict how software and the computerization of everything will pan out then electrification is a better example to consider.
  - nugget 9y ago
    The revolutions that occurred in electrification (up to and including sophisticated home appliances), mass production (clothing, food, shelter), and transportation (from horses to cars) brought a lot more people into the workforce and into the same connected market economy. Prior to these inventions, people were spread out and almost entirely disconnected from a shared economy: in the 1870s about 75% of American families were rural and very time-intensively grew/raised most of what they needed to survive themselves. Over the next hundred or so years a bunch of wonderful inventions moved everyone together into the same very tightly-connected market economy, and reduced non-leisure time commitments at home by 80%+ (notably allowing women to join and remain in the workforce in greater numbers). There is a strong argument that this connective effort is what led to massive GDP growth - that these were a series of "one trick" ponies, by definition - and that while there are amazing inventions on the horizon, no doubt, they will have much more impact on quality of life than on traditional GDP and productivity, at least in terms of Wall Street's expectations. Even if autonomous robots come and reduce non-leisure domestic commitments to absolute zero, it wouldn't be nearly as impactful as the advancements that we've already experienced, from a productivity standpoint. I tend to agree that "it's different this time" is one of the most dangerous phrases in the English language but at some point, maybe it's true.
knodi123 9y ago
It's not different. But all of those things involved massive crashes in certain sectors of the economy.
tbirrell 9y ago
Ease of scale basically.

by joobus 9y ago

First jobs were replaced by simple tools, then by livestock, then by machines. What makes software any different from any other advancements made throughout history?

dragonwriter 9y ago

All of the earlier advances require human maintenance that scale with the scope of deployment. The whole reason software is a hot area of investment is that it's support requirements don't scale that way, which means is good for the capitalist from a cost perspective, but bad for the wage-labor dependent class.

defined 9y ago

It's the only advancement where making a million copies of an item takes a few minutes and no raw material. Try that with tools, livestock and machines. Or anything physical.

convolvatron 9y ago
but before that you have to go through several iterations of failure with a bunch of costly entitled developers before you get something which sort of approximates what you wanted and is so internally broken its almost impossible to extend not really ragging on the developers, but its not a cheap or risk free process. and if you aren't actively investing in maintenance it will die. the costs are all just per-type not so much per-instance
- defined 9y ago
  > but before that you have to go through several iterations of failure with a bunch of costly entitled developers before you get something which sort of approximates what you wanted and is so internally broken its almost impossible to extend This is too often true, subject to the following observations: - Project failure is almost always a management failure (no citations needed, really), because: - The thing to be built is poorly conceived and explained by the leadership/stakeholders, who are seldom the devs - The delivery schedule is wildly optimistic or infeasible, because unrealistic promises were made to customers or investors - The operational parameters of the system (peak users, load, transactions, data size, availability, etc) are overestimated by the leadership by a couple of orders of magnitude and pushback tends to be career-limiting - When the leadership eventually realizes the failure of its vision, it "pivots", and keeps pivoting at the speed of a turbine, "reframing" the so-called vision. - The "costly, entitled" developers often come from some body shop which pays them $25/hour and charges $75 (illustrative values only) - Costly, unentitled, competent developers probably ran for the hills or were rejected because they weren't "with the program" - they said something realistic in the interview. - The process is not cheap or easy or risk free, yet many companies hire and pay as if it is. - The mere act of getting cohesive requirements out of people is one of the hardest parts of software, and one of the largest reasons for failure. - The rest isn't easy, either, unless it's a conceptual copy of some vanilla CRUD system that does nothing novel or "web scale" (whatever that is). OTOH, I read some devs stating that software development is just bolting legos together, what's the problem? If you are referring to these devs, I see your point, but they and their ilk are the creations of corporates and startups who pushed the view that devs are commodity items - resources - who aren't really skilled and can be churned out of bootcamps and will be good enough to meet the deadlines. Now I am not criticizing bootcamps or the people who make it through them. There are many smart people who have no formal CS background who could make good devs, given time and experience. There are even more people who should not be allowed near a computer, sadly, who IMO are more commonly encountered. I am criticizing the social forces that put money above useful education, profit above knowledge, ideology above science and engineering, the short term against the entire future.
dragonwriter 9y ago
> It's the only advancement where making a million copies of an item takes a few minutes and no raw material. And, mire directly to the point, doesn't require additional labor inputs. (If extraction is automated by software tools driving hardware, even things that demand raw materials can have this feature.)

chrisdbaldwin 9y ago

The scale of software tool utility is pretty much unparalleled by traditional tools. Traditional tools such as livestock and machines require breeding, mining... some sort of "manufacturing" process. Okay, software has manufacturing cost too, but typically only once to create, and once to maintain. Imagine if you could dynamically spawn livestock without the overhead of mining, breeding, and trading. The jobs that software has/will eat are, imo, beyond current human comprehension.

evgen 9y ago
Electrification was far more significant and far-reaching, occurred on a similarly quick timescale, and led to a massive boom in productivity across the board. If you want to predict how software and the computerization of everything will pan out then electrification is a better example to consider.
- nugget 9y ago
  The revolutions that occurred in electrification (up to and including sophisticated home appliances), mass production (clothing, food, shelter), and transportation (from horses to cars) brought a lot more people into the workforce and into the same connected market economy. Prior to these inventions, people were spread out and almost entirely disconnected from a shared economy: in the 1870s about 75% of American families were rural and very time-intensively grew/raised most of what they needed to survive themselves. Over the next hundred or so years a bunch of wonderful inventions moved everyone together into the same very tightly-connected market economy, and reduced non-leisure time commitments at home by 80%+ (notably allowing women to join and remain in the workforce in greater numbers). There is a strong argument that this connective effort is what led to massive GDP growth - that these were a series of "one trick" ponies, by definition - and that while there are amazing inventions on the horizon, no doubt, they will have much more impact on quality of life than on traditional GDP and productivity, at least in terms of Wall Street's expectations. Even if autonomous robots come and reduce non-leisure domestic commitments to absolute zero, it wouldn't be nearly as impactful as the advancements that we've already experienced, from a productivity standpoint. I tend to agree that "it's different this time" is one of the most dangerous phrases in the English language but at some point, maybe it's true.

knodi123 9y ago

It's not different. But all of those things involved massive crashes in certain sectors of the economy.

tbirrell 9y ago

Ease of scale basically.