5 ms·

This is not the first time, it won't be the last time. The next hype round is only a matter of time because rich people need somewhere to put their money. Unti

This is not the first time, it won't be the last time. The next hype round is only a matter of time because rich people need somewhere to put their money. Until you solve the problem of middle management status being based on the number of people you manage, you will keep getting bloated organizations. Until an organization figures out a better way to manage their dependencies on each other, every new thing an organization wants to do will require an n log n number of people. (In bad organizations, it's n^2.) Organizations are just committing on "not growing" in terms of product offerings in the next period of time. The stock market, after some time, will reward growth and the cycle will start again.

l33t233372 4y ago
> will require an n log n number of people. (In bad organizations, it's n^2.) I understand this is signaling that the author understands technical concepts like big-O, but why use technical terms imprecisely when non-technical terms work just fine?
- baremetal 4y ago
  Because it's fun, and well suited to the audience.
- margorczynski 4y ago
  Well it isn't fully clear to me what "n" here would be - seems a bit hand-wavy.
- marcosdumay 4y ago
  What non-technical term would communicate the same idea?
- astroalex 4y ago
  Considering the audience, this language is more vivid and precise than saying something like "middle management causes bloat." It puts a picture in your head of how that bloat grows and what causes the bloat (the tree structure of the organization).
- ebiester 4y ago
  It's not precise, but it's pretty close when you look at the number of management and support positions to individual contributors in many techinal organizations. It's beyond the scope of this conversation, but the short is that it seems to become a network effect problem. Every team does not work with every other team, but it does end up in a hub-and-spoke system with matrixes over the system. Oversight (ops, security, PMO) do not scale linearly because the competing priorities make scheduling more difficult. I've read others who have spoken about it, but I'd have to go back and find it.
- MrStonedOne 4y ago
  [dead]
debacle 4y ago
> rich people need somewhere to put their money. With more typical interest rates, rich people will make more typical investments.
- pdntspa 4y ago
  Which is great but the fed previously stated that it planned on lowering interest rates once inflation was under control
  - debacle 4y ago
    That's going to take a long time.
    - pdntspa 4y ago
      2024, according to JPow speeches from a few months ago. Though now I don't know...
    - substation13 4y ago
      Rate hikes may be cancelled if there is a debt crisis.
- ebiester 4y ago
  What will these typical investments be? Consider capital gains taxation versus dividend taxation in your analysis.
coldtea 4y ago
>The next hype round is only a matter of time because rich people need somewhere to put their money. They put their money in hype when the money is free or close, and putting it in those "investments" make sense. Like when you have QE and near-zero interest rates...
- kypro 4y ago
  Expect this doesn't explain past hype bubbles when there were non-zero interest rates. During the dot-com bubble rates were above 5%, for example.
  - coldtea 4y ago
    Tons of money funnelled into banking turned to cheap loans for VC money, however was in great supply, and this proped up stocks artificially out of any pragmatic evaluation, and hyped the public for them (with the cooperation of the media), which in turn provided further supply for the extravagance: "even the tactical errors committed by investors were insufficient to create the bubble that burst so dramatically last year. For that, we needed a perverse private-public partnership led by Wall Street and the Federal Reserve. The Fed mistakenly created too much money in the fall of 1999 in order to fend off the expected deflationary impact of the Y2K bug—effectively pouring gasoline on a smoldering fire in the stock markets—and banks and brokerage houses used some of the excess liquidity to fund the share price bubble. In the fourth quarter of 1999, the Fed expanded the money supply at an annual rate of 22% (9.6% after seasonal adjustment). In comparison, the rate of growth of the money supply in the fourth quarter of 2000 was 9.2% (–2.8% after seasonal adjustment). Fed Chairman Alan Greenspan, seeing the flood of money into risky start-ups, pricked the dot-com balloon with monetary tightening in the spring of 2000. It was the dot-coms’ misfortune to be first the beneficiaries and then the victims of these larger economic forces". In that case they didn't need "near zero interest" rates, because they were artifically promised much larger returns than the actual rates. That is the actual key driver: easy handed loans for BS uses (to VCs, and even better the general public), and hyped returns above the rates. The "near zero" rates is not necessary, it's just the more extreme case. https://hbr.org/2001/05/whos-to-blame-for-the-bubble#:~:text=Venture%20capitalists%20bear%20a%20marked,that%20the%20dot%2Dcoms%20represented https://hbr.org/2001/05/whos-to-blame-for-the-bubble#:~:text....
cmrdporcupine 4y ago
In the end, a few decades on, managers still have no idea how to make software project delivery reliable and quantifiable. And too many of these organizations still act like they think throwing more engineers at a problem makes things go faster. And the $$ and prestige of working in tech still attracts too many people whose primary interest is not just in getting things done, but being seen doing it, and, as, you say, improving their status. The tide is going out and we're seeing who isn't wearing bathing suits. But it might be the whole beach. ... Still though the primary thing driving investment in tech and software is the same thing that drove investment in industrialized cotton mills or Ford Model T assembly lines: mechanization makes profits, and... computerized mechanization is the most powerful cost reducing system ever known to mankind.
- ResearchCode 4y ago
  Software is not reliable in the same sense that mathematics delivery is not reliable. If you're doing something routine and reliable you should be automating it.
- thedougd 4y ago
  Managers usually over staff because it’s an easy lever available to them to cover their ass. If the project fails or is late, a simple criticism is, “Why didn’t you scale up the team?” They can avoid this criticism and shift the focus to externalities such as shifting requirements, late dependencies, etc.

l33t233372 4y ago

> will require an n log n number of people. (In bad organizations, it's n^2.) I understand this is signaling that the author understands technical concepts like big-O, but why use technical terms imprecisely when non-technical terms work just fine?

baremetal 4y ago
Because it's fun, and well suited to the audience.
margorczynski 4y ago
Well it isn't fully clear to me what "n" here would be - seems a bit hand-wavy.
marcosdumay 4y ago
What non-technical term would communicate the same idea?
astroalex 4y ago
Considering the audience, this language is more vivid and precise than saying something like "middle management causes bloat." It puts a picture in your head of how that bloat grows and what causes the bloat (the tree structure of the organization).
ebiester 4y ago
It's not precise, but it's pretty close when you look at the number of management and support positions to individual contributors in many techinal organizations. It's beyond the scope of this conversation, but the short is that it seems to become a network effect problem. Every team does not work with every other team, but it does end up in a hub-and-spoke system with matrixes over the system. Oversight (ops, security, PMO) do not scale linearly because the competing priorities make scheduling more difficult. I've read others who have spoken about it, but I'd have to go back and find it.
MrStonedOne 4y ago
[dead]

debacle 4y ago

> rich people need somewhere to put their money. With more typical interest rates, rich people will make more typical investments.

pdntspa 4y ago
Which is great but the fed previously stated that it planned on lowering interest rates once inflation was under control
- debacle 4y ago
  That's going to take a long time.
  - pdntspa 4y ago
    2024, according to JPow speeches from a few months ago. Though now I don't know...
  - substation13 4y ago
    Rate hikes may be cancelled if there is a debt crisis.
ebiester 4y ago
What will these typical investments be? Consider capital gains taxation versus dividend taxation in your analysis.

coldtea 4y ago

>The next hype round is only a matter of time because rich people need somewhere to put their money. They put their money in hype when the money is free or close, and putting it in those "investments" make sense. Like when you have QE and near-zero interest rates...

kypro 4y ago
Expect this doesn't explain past hype bubbles when there were non-zero interest rates. During the dot-com bubble rates were above 5%, for example.
- coldtea 4y ago
  Tons of money funnelled into banking turned to cheap loans for VC money, however was in great supply, and this proped up stocks artificially out of any pragmatic evaluation, and hyped the public for them (with the cooperation of the media), which in turn provided further supply for the extravagance: "even the tactical errors committed by investors were insufficient to create the bubble that burst so dramatically last year. For that, we needed a perverse private-public partnership led by Wall Street and the Federal Reserve. The Fed mistakenly created too much money in the fall of 1999 in order to fend off the expected deflationary impact of the Y2K bug—effectively pouring gasoline on a smoldering fire in the stock markets—and banks and brokerage houses used some of the excess liquidity to fund the share price bubble. In the fourth quarter of 1999, the Fed expanded the money supply at an annual rate of 22% (9.6% after seasonal adjustment). In comparison, the rate of growth of the money supply in the fourth quarter of 2000 was 9.2% (–2.8% after seasonal adjustment). Fed Chairman Alan Greenspan, seeing the flood of money into risky start-ups, pricked the dot-com balloon with monetary tightening in the spring of 2000. It was the dot-coms’ misfortune to be first the beneficiaries and then the victims of these larger economic forces". In that case they didn't need "near zero interest" rates, because they were artifically promised much larger returns than the actual rates. That is the actual key driver: easy handed loans for BS uses (to VCs, and even better the general public), and hyped returns above the rates. The "near zero" rates is not necessary, it's just the more extreme case. https://hbr.org/2001/05/whos-to-blame-for-the-bubble#:~:text=Venture%20capitalists%20bear%20a%20marked,that%20the%20dot%2Dcoms%20represented https://hbr.org/2001/05/whos-to-blame-for-the-bubble#:~:text....

cmrdporcupine 4y ago

In the end, a few decades on, managers still have no idea how to make software project delivery reliable and quantifiable. And too many of these organizations still act like they think throwing more engineers at a problem makes things go faster. And the $$ and prestige of working in tech still attracts too many people whose primary interest is not just in getting things done, but being seen doing it, and, as, you say, improving their status. The tide is going out and we're seeing who isn't wearing bathing suits. But it might be the whole beach. ... Still though the primary thing driving investment in tech and software is the same thing that drove investment in industrialized cotton mills or Ford Model T assembly lines: mechanization makes profits, and... computerized mechanization is the most powerful cost reducing system ever known to mankind.

ResearchCode 4y ago
Software is not reliable in the same sense that mathematics delivery is not reliable. If you're doing something routine and reliable you should be automating it.
thedougd 4y ago
Managers usually over staff because it’s an easy lever available to them to cover their ass. If the project fails or is late, a simple criticism is, “Why didn’t you scale up the team?” They can avoid this criticism and shift the focus to externalities such as shifting requirements, late dependencies, etc.