4 ms·

Nah, I don't think Spolsky is responsible. What is really responsible is that computers themselves influence people to think algorithmically. I know that sounds

Nah, I don't think Spolsky is responsible. What is really responsible is that computers themselves influence people to think algorithmically. I know that sounds weird, but I observed after doing a degree in math: after a lot of work in mathematics, I started to think much more analytically and algorithmically. I think the same thing happens at tech companies, and that's why Spolsky was so inclined to make the process more efficient: tech companies are themselves reactors to grow the algorithmic way of thinking in people, and that in turn causes people to seek for algorithmic solutions such as the hiring system we have today. Yes, Spolsky might have created some new hiring practice, but if it wasn't for him, it would be someone else because it is a natural consequence of a large group of people thinking in an algorithmic way for years and years. If cults can influence poeple to commit mass suicide, surely ten years of math and programming can influence people to become more like machines.

mmcdermott 3y ago
I've certainly seen the phenomena you describe, but it doesn't seem like that belief is completely universal. "Humanists often contend that machines tend to dehumanize people by forcing them to have rigid personalities, but really, the contrary is true. Because the machines are rigid, the people who use them must–if they are to be successful–supply more than their share of flexibility." – Gerald Weinberg, The Psychology of Computer Programming, chapter 11.
- vouaobrasil 3y ago
  I think the flexibility that Weinberg was talking about was the expansion and expression of flexibility within a narrow domain. And I think we have an immense proclivity in modern society to approach problems like climate change with a very specialized, algorithmic mind (optimize along one variable like reduction of CO2, design specialized batteries), all within a given system. If you look at some indigenous cultures who had a more holistic viewpoint without exposure to machinery, they are much more about relationships between all living organisms and the emotionality behind those relationships, rather than using analytic methods to solve things. So, while I understand where Weinberg is coming from, I do not think his view applies to the greater discussion. Moreover, I think he was confused about what others meant by dehumanization, especially when he talked about personalities. It is not the personalities that become rigid, but the problem-solving approaches. My 2 cents....
- organsnyder 3y ago
  I think both can be true. The need to conform to something rigid requires flexibility. However, you're effectively adopting the rigidity of that thing, so you are no longer able to employ your flexibility.
adamc 3y ago
While I kind of agree, I think it's an extension of thinking that started at the dawn of the industrial age, where people had to contend with the rather limited options for how the technology could work, and build systems around it. The rise of engineering was one of the consequences. But the options (mechanical devices) of the industrial age were obviously much less flexible than what we have now in software, and you're probably right that that flexibility has had a big effect on how many people think.
- temporarely 3y ago
  Excuse me for interjecting some facts here: software development did not start in the late 90s. The group of people who were building software when it was very much an art were super sharp and analytical. It is beyond arrogance (it is ignorance) to think that we are subjecting ourselves to unreasonable "mathematical rigor" while the earlier generations were apparently clueless cavemen of IT. Utter nonsense. Go see what IBM -- just one example -- did in its heyday and just who were the people who were making and writing software for those machines. Joel is being Joel -- self satisfied -- and according himself the position of leading light of an entire industry's recruitment dilemma and direction after the internet (doh) exploded demand for software developers. The internet also caused another problem (technical blogs): all of a sudden people who would normally never even cross paths with the elites of their industry (because they were and are not in the same caliber to be quite frank) were reading their whitepaper tealeaves and writing blogs on how everyone had to do x, y and z. Recruitment is just one item. Can we discuss stack and architecture and team structure and methodology as well. So Joel is right. Blowhards with blogs are quite responsible for the mess that is the profession of software development today.
  - hkon 3y ago
    Many a time I wanted to just write up changes I wanted in the teams in anonymous blog posts and provide links to team members. Would have saved me a lot of time.
    - temporarely 3y ago
Macha 3y ago
I also wonder how long of a lasting impact he's even had on his founded companies. Trello is obviously absorbed into the Atlassian blob, Stack Overflow's current owners are clearly in the extraction phase, and while Fog Creek always seemed a cool place to work, my understanding is its current incarnation of Glitch is just your typical mid-sized tech co.
- diggan 3y ago
  > Fog Creek always seemed a cool place to work, my understanding is its current incarnation of Glitch is just your typical mid-sized tech co. I don't know neither Fog Creek or Glitch, but what you're writing sounds like you're putting a negative spin on it. To me it sounds wonderful that it's not yet another startup/company that basically boils down to "take over the world with VC funding or die trying".
  - Macha 3y ago
    I don't even mean in terms of the product. Their old main product was a bug tracker and their current iteration is a low code dev product. I mean Fog Creek used to be a strong proponent of developer facing decisions like private offices for developers, and (back when it was a novelty) spending time easy onboarding and builds.
    - FireBeyond 3y ago
      > I mean Fog Creek used to be a strong proponent of developer facing decisions like private offices for developers People like to rag on MSFT, but when I was there (around mid 2000s), this was very much the norm. Meeting rooms and shared spaces, but generally every engineer had their own office (at least in my BU).
- rmccue 3y ago
  > my understanding is its current incarnation of Glitch is just your typical mid-sized tech co Glitch was acquired by Fastly, so it's not a fully independent mid-sized tech co, but it does seem to be doing well and seems reasonably independent still.
tarkin2 3y ago
Professional deformation is a term I’ve heard used to describe this
itronitron 3y ago
I disagree. You could just as easily say that 95% of tech is reusing other people's stuff and so tech companies just copy, sorry reuse, the hiring methods of other tech companies. Larger companies are more likely to have an engineered hiring process and they are also more visible than smaller companies, so their hiring methods are the ones that get copied.
thomastjeffery 3y ago
I think he wasn't responsible in the sense that he was simply present for a naturally emergent phenomenon; but at the same time, he was responsible for actively participating in and promoting that phenomenon. Responsibility is relative to authority, and Spolsky ended up with a very significant amount of authority. Whether or not his behavior was good or bad, it's pretty clear that he had more responsibility than any one person really should. I wouldn't consider that his fault: he didn't really seek it out, or intentionally abuse it. I would, however, consider it what happened. If we tack a step back, and look at this scenario as a system, we can see that it's the structure of the system itself that lead to these problems, not the intentions of any of its participants. Now that we have the power of hindsight and objectivity, we can use that to design a better system. The big question is, will we ever get around to implementing it?
- nradov 3y ago
  Emergent systems can't really be intentionally designed.
  - thomastjeffery 3y ago
    Why not? We are able to design systems, and we are able to objectively recognize and understand emergent behaviors in arbitrary systems. Sure, we can't exhaustively find every emergent behavior, but we are not 100% blind to them, either. The only way to guarantee an emergent behavior continues is to preserve the structure of its system. That means the only strategy that can change emergent behaviors is a strategy that changes systems.
- robertlagrant 3y ago
  You can use it to design a different system, but it's pretty unlikely to be better.
  - thomastjeffery 3y ago
    You can choose to keep the system you've got, and that is guaranteed to be as bad as it is. The only other option is anarchy. Out of the three, I advocate intentionally designing a system.
BoiledCabbage 3y ago
> Nah, I don't think Spolsky is responsible. It's hard to read his guide now with recent eyes and not completely agree with him. [1] . Software had been hiring for a while and never hit it. Plenty of other professions have very technical or algorithmic thinking without it. I'm not convinced it was inevitable as you say. 1. https://www.joelonsoftware.com/2006/10/25/the-guerrilla-guide-to-interviewing-version-30/ https://www.joelonsoftware.com/2006/10/25/the-guerrilla-guid...

mmcdermott 3y ago

I've certainly seen the phenomena you describe, but it doesn't seem like that belief is completely universal. "Humanists often contend that machines tend to dehumanize people by forcing them to have rigid personalities, but really, the contrary is true. Because the machines are rigid, the people who use them must–if they are to be successful–supply more than their share of flexibility." – Gerald Weinberg, The Psychology of Computer Programming, chapter 11.

vouaobrasil 3y ago
I think the flexibility that Weinberg was talking about was the expansion and expression of flexibility within a narrow domain. And I think we have an immense proclivity in modern society to approach problems like climate change with a very specialized, algorithmic mind (optimize along one variable like reduction of CO2, design specialized batteries), all within a given system. If you look at some indigenous cultures who had a more holistic viewpoint without exposure to machinery, they are much more about relationships between all living organisms and the emotionality behind those relationships, rather than using analytic methods to solve things. So, while I understand where Weinberg is coming from, I do not think his view applies to the greater discussion. Moreover, I think he was confused about what others meant by dehumanization, especially when he talked about personalities. It is not the personalities that become rigid, but the problem-solving approaches. My 2 cents....
organsnyder 3y ago
I think both can be true. The need to conform to something rigid requires flexibility. However, you're effectively adopting the rigidity of that thing, so you are no longer able to employ your flexibility.

adamc 3y ago

While I kind of agree, I think it's an extension of thinking that started at the dawn of the industrial age, where people had to contend with the rather limited options for how the technology could work, and build systems around it. The rise of engineering was one of the consequences. But the options (mechanical devices) of the industrial age were obviously much less flexible than what we have now in software, and you're probably right that that flexibility has had a big effect on how many people think.

temporarely 3y ago
Excuse me for interjecting some facts here: software development did not start in the late 90s. The group of people who were building software when it was very much an art were super sharp and analytical. It is beyond arrogance (it is ignorance) to think that we are subjecting ourselves to unreasonable "mathematical rigor" while the earlier generations were apparently clueless cavemen of IT. Utter nonsense. Go see what IBM -- just one example -- did in its heyday and just who were the people who were making and writing software for those machines. Joel is being Joel -- self satisfied -- and according himself the position of leading light of an entire industry's recruitment dilemma and direction after the internet (doh) exploded demand for software developers. The internet also caused another problem (technical blogs): all of a sudden people who would normally never even cross paths with the elites of their industry (because they were and are not in the same caliber to be quite frank) were reading their whitepaper tealeaves and writing blogs on how everyone had to do x, y and z. Recruitment is just one item. Can we discuss stack and architecture and team structure and methodology as well. So Joel is right. Blowhards with blogs are quite responsible for the mess that is the profession of software development today.
- hkon 3y ago
  Many a time I wanted to just write up changes I wanted in the teams in anonymous blog posts and provide links to team members. Would have saved me a lot of time.
  - temporarely 3y ago

Macha 3y ago

I also wonder how long of a lasting impact he's even had on his founded companies. Trello is obviously absorbed into the Atlassian blob, Stack Overflow's current owners are clearly in the extraction phase, and while Fog Creek always seemed a cool place to work, my understanding is its current incarnation of Glitch is just your typical mid-sized tech co.

diggan 3y ago
> Fog Creek always seemed a cool place to work, my understanding is its current incarnation of Glitch is just your typical mid-sized tech co. I don't know neither Fog Creek or Glitch, but what you're writing sounds like you're putting a negative spin on it. To me it sounds wonderful that it's not yet another startup/company that basically boils down to "take over the world with VC funding or die trying".
- Macha 3y ago
  I don't even mean in terms of the product. Their old main product was a bug tracker and their current iteration is a low code dev product. I mean Fog Creek used to be a strong proponent of developer facing decisions like private offices for developers, and (back when it was a novelty) spending time easy onboarding and builds.
  - FireBeyond 3y ago
    > I mean Fog Creek used to be a strong proponent of developer facing decisions like private offices for developers People like to rag on MSFT, but when I was there (around mid 2000s), this was very much the norm. Meeting rooms and shared spaces, but generally every engineer had their own office (at least in my BU).
rmccue 3y ago
> my understanding is its current incarnation of Glitch is just your typical mid-sized tech co Glitch was acquired by Fastly, so it's not a fully independent mid-sized tech co, but it does seem to be doing well and seems reasonably independent still.

tarkin2 3y ago

Professional deformation is a term I’ve heard used to describe this

itronitron 3y ago

I disagree. You could just as easily say that 95% of tech is reusing other people's stuff and so tech companies just copy, sorry reuse, the hiring methods of other tech companies. Larger companies are more likely to have an engineered hiring process and they are also more visible than smaller companies, so their hiring methods are the ones that get copied.

thomastjeffery 3y ago

I think he wasn't responsible in the sense that he was simply present for a naturally emergent phenomenon; but at the same time, he was responsible for actively participating in and promoting that phenomenon. Responsibility is relative to authority, and Spolsky ended up with a very significant amount of authority. Whether or not his behavior was good or bad, it's pretty clear that he had more responsibility than any one person really should. I wouldn't consider that his fault: he didn't really seek it out, or intentionally abuse it. I would, however, consider it what happened. If we tack a step back, and look at this scenario as a system, we can see that it's the structure of the system itself that lead to these problems, not the intentions of any of its participants. Now that we have the power of hindsight and objectivity, we can use that to design a better system. The big question is, will we ever get around to implementing it?

nradov 3y ago
Emergent systems can't really be intentionally designed.
- thomastjeffery 3y ago
  Why not? We are able to design systems, and we are able to objectively recognize and understand emergent behaviors in arbitrary systems. Sure, we can't exhaustively find every emergent behavior, but we are not 100% blind to them, either. The only way to guarantee an emergent behavior continues is to preserve the structure of its system. That means the only strategy that can change emergent behaviors is a strategy that changes systems.
robertlagrant 3y ago
You can use it to design a different system, but it's pretty unlikely to be better.
- thomastjeffery 3y ago
  You can choose to keep the system you've got, and that is guaranteed to be as bad as it is. The only other option is anarchy. Out of the three, I advocate intentionally designing a system.

BoiledCabbage 3y ago

> Nah, I don't think Spolsky is responsible. It's hard to read his guide now with recent eyes and not completely agree with him. [1] . Software had been hiring for a while and never hit it. Plenty of other professions have very technical or algorithmic thinking without it. I'm not convinced it was inevitable as you say. 1. https://www.joelonsoftware.com/2006/10/25/the-guerrilla-guide-to-interviewing-version-30/ https://www.joelonsoftware.com/2006/10/25/the-guerrilla-guid...