4 ms·

Not all tech, of course, but a lot of tech (eg, HFT and adtech) is zero-sum, so it's low impact by definition.

by sudosysgen 3y ago

s1artibartfast 3y ago
I think I see. You are framing impact ( I think) as something similar to net positive utils in a utilitarian framework
- sudosysgen 3y ago
  If I am an HFT engineer, making a faster trading platform literally has very little impact. It doesn't create any value, it just moves value from the slower guy to the faster guy. Then all the HFT companies hire smarter engineers and all improve their latency just as much, and there is literally zero impact to society.
  - s1artibartfast 3y ago
    I get the zero sum part, it is the other part that was less clear to me. Eg having an iq 120 dmv teller or server at brunch or But then again, I disagree with your premise that these other careers are intelligence starved - some of my smartest friends do them. I think rather those careers are simply more stifling and less creative. That is to say your HFT seems flashy because they are doing novel work in a developing area. If they were in the civil service, they probably wouldn't be reinventing the wheel, but just an efficient bureaucrat.
    - ethanbond 3y ago
      Obviously people sort on intelligence into different industries. You’re looking at a paper on similar/probably-mostly-identical sorting happening between specific firms. I don’t believe everyone working in e.g. food service is dumb (nor that everyone working in adtech or HFT is smart), only that the distribution is probably quite extremely uneven. There is plenty of space for creativity and novel thinking within bureaucracies, trades, service industries, etc., and that is exactly the reason it’d be great to have a more even distribution of smart people into them. An efficient bureaucrat can navigate themselves to the right place to make a huge amount of impact on people’s actual daily lives at scale. That’s a good thing!
  - mewpmewp2 3y ago
    HFT provides liquidity, makes markets more efficient, prices more accurate and narrows spreads. Moving money around as accurately as possible increases the accuracy of allocating funds where they should be. Everything in the World and economy is connected. Accurate prices, efficient trading conditions allow for better allocation of funds on targets worthwhile pursuing. On the whole you can think that there's Company A and Company B, where both produce different amounts of value. There's an optimal state of how much each company should be funded in order for optimal value production from both together. The more accurate you are with allocating funds the best are the gains. So in that sense it's not zero sum.
JumpCrisscross 3y ago
> a lot of tech (eg, HFT and adtech) is zero-sum It's zero sum within the system. The existence of HFT as a whole is positive sum, though I agree that the compensation differential attributed to it outclasses that sum by a large measure.
- sudosysgen 3y ago
  The existence of HFT definitely is a net positive, but I think it's a tough argument to say that further latency reduction still has any real positive offshoot, so I'd say that it's now as a whole in a zero-sum regime - anyone winning, on the margin, only makes them profit as much as someone else is losing, whereas earlier you could say that tighter spreads and more liquidity helps other market participants.
  - JumpCrisscross 3y ago
    > it's a tough argument to say that further latency reduction still has any real positive offshoot There’s an element of Feinman’s spinning, wobbling plates [1]. The problems one must solve to push latency ever lower have knock-on effects. > whereas earlier you could say that tighter spreads and more liquidity helps other market participants The guarantee of being nipped curtails entire classes of dirty behaviour. Chunking out an order is smart, but it wasn’t always done. HFT, in this model, serves an adversarial regulatory function. We also have the basic fact that, in a Ricardian sense, folks are trading with that HFT shop over their slower competitors. At some infinitesimal level, the counterparty chose (and was delivered on) that preference. [1] https://pubs.aip.org/aapt/ajp/article-abstract/75/3/240/1056339/Feynman-s-wobbling-plate?redirectedFrom=fulltext https://pubs.aip.org/aapt/ajp/article-abstract/75/3/240/1056...
    - ethanbond 3y ago
      I don’t think someone choosing an available preference is de facto validation that the preference was made available by anything close to a globally optimal allocation of resources. Of course I’m suspicious of any system that claims to be able to globally optimize in this way, but I do think it’d be better for almost everyone if we could.

Not all tech, of course, but a lot of tech (eg, HFT and adtech) is zero-sum, so it's low impact by definition.

s1artibartfast 3y ago

I think I see. You are framing impact ( I think) as something similar to net positive utils in a utilitarian framework

sudosysgen 3y ago
If I am an HFT engineer, making a faster trading platform literally has very little impact. It doesn't create any value, it just moves value from the slower guy to the faster guy. Then all the HFT companies hire smarter engineers and all improve their latency just as much, and there is literally zero impact to society.
- s1artibartfast 3y ago
  I get the zero sum part, it is the other part that was less clear to me. Eg having an iq 120 dmv teller or server at brunch or But then again, I disagree with your premise that these other careers are intelligence starved - some of my smartest friends do them. I think rather those careers are simply more stifling and less creative. That is to say your HFT seems flashy because they are doing novel work in a developing area. If they were in the civil service, they probably wouldn't be reinventing the wheel, but just an efficient bureaucrat.
  - ethanbond 3y ago
    Obviously people sort on intelligence into different industries. You’re looking at a paper on similar/probably-mostly-identical sorting happening between specific firms. I don’t believe everyone working in e.g. food service is dumb (nor that everyone working in adtech or HFT is smart), only that the distribution is probably quite extremely uneven. There is plenty of space for creativity and novel thinking within bureaucracies, trades, service industries, etc., and that is exactly the reason it’d be great to have a more even distribution of smart people into them. An efficient bureaucrat can navigate themselves to the right place to make a huge amount of impact on people’s actual daily lives at scale. That’s a good thing!
- mewpmewp2 3y ago
  HFT provides liquidity, makes markets more efficient, prices more accurate and narrows spreads. Moving money around as accurately as possible increases the accuracy of allocating funds where they should be. Everything in the World and economy is connected. Accurate prices, efficient trading conditions allow for better allocation of funds on targets worthwhile pursuing. On the whole you can think that there's Company A and Company B, where both produce different amounts of value. There's an optimal state of how much each company should be funded in order for optimal value production from both together. The more accurate you are with allocating funds the best are the gains. So in that sense it's not zero sum.

JumpCrisscross 3y ago

> a lot of tech (eg, HFT and adtech) is zero-sum It's zero sum within the system. The existence of HFT as a whole is positive sum, though I agree that the compensation differential attributed to it outclasses that sum by a large measure.

sudosysgen 3y ago
The existence of HFT definitely is a net positive, but I think it's a tough argument to say that further latency reduction still has any real positive offshoot, so I'd say that it's now as a whole in a zero-sum regime - anyone winning, on the margin, only makes them profit as much as someone else is losing, whereas earlier you could say that tighter spreads and more liquidity helps other market participants.
- JumpCrisscross 3y ago
  > it's a tough argument to say that further latency reduction still has any real positive offshoot There’s an element of Feinman’s spinning, wobbling plates [1]. The problems one must solve to push latency ever lower have knock-on effects. > whereas earlier you could say that tighter spreads and more liquidity helps other market participants The guarantee of being nipped curtails entire classes of dirty behaviour. Chunking out an order is smart, but it wasn’t always done. HFT, in this model, serves an adversarial regulatory function. We also have the basic fact that, in a Ricardian sense, folks are trading with that HFT shop over their slower competitors. At some infinitesimal level, the counterparty chose (and was delivered on) that preference. [1] https://pubs.aip.org/aapt/ajp/article-abstract/75/3/240/1056339/Feynman-s-wobbling-plate?redirectedFrom=fulltext https://pubs.aip.org/aapt/ajp/article-abstract/75/3/240/1056...
  - ethanbond 3y ago
    I don’t think someone choosing an available preference is de facto validation that the preference was made available by anything close to a globally optimal allocation of resources. Of course I’m suspicious of any system that claims to be able to globally optimize in this way, but I do think it’d be better for almost everyone if we could.