5 ms·

In 1891, the first electromechanical telephone exchange was invented by Almon Strowger, an undertaker. He was concerned that the human operators were diverting

by ajb 6y ago

In 1891, the first electromechanical telephone exchange was invented by Almon Strowger, an undertaker. He was concerned that the human operators were diverting his customers (specifically, the wife of one of his competitors was an operator). If you think about it, every business interaction before the twentieth century was mediated by 1:1 interactions with humans, who brought their own prejudices and self-interest to it. The Stowger exchange was the start of an era of "mechanical honesty" - machines, businesses, and even government departments that could only act in one way, because any bespoke deviation was too inefficient to exist/be profitable, and so ordinary citizens could rely on them. We are coming to the end of that era. Computing power has reached the point where bespoke dishonesty and manipulation can be implemented efficiently. The public still retains the expectations of the mechanical honesty era, and is an easy mark. That has to change... [edited for punctuation]

castratikron 6y ago
I would say there's no moral difference between a human operator and a mechanical operator whose behavior is controlled by a human. Configuring the operator to drop calls from a particular area code would be an example in the case you provided.
- ajb 6y ago
  Sure. It just used to be harder to build in, and to do so precisely. In your example, dropping calls from an area code, that's a relatively coarse filter. There's a whole strand of academic research [1] arguing about whether people can build prejudice into technological artifacts, eg one disputed example was an underpass with low clearance, supposedly so that white middle class people in cars could pass, but black poor people in buses could not [1, p123]. These days that dispute is ridiculous - obviously prejudice and bias can be built into technological artifacts. It's one "if statement" away. [1] "Do Artifacts have politics" https://www.cc.gatech.edu/~beki/cs4001/Winner.pdf https://www.cc.gatech.edu/~beki/cs4001/Winner.pdf
  - krisgee 6y ago
    >These days that dispute is ridiculous - obviously prejudice and bias can be built into technological artifacts. It's one "if statement" away. And with machine learning we're building our biases into the algorithms explicitly; see all those "resume parsers are racist" articles. This is clearly an issue that needs to be dealth with but the problem is that it's not a problem we can engineer our way of, we need to confront the biases that we have and try and build ways to expose biases we don't even realize we have. This is obviously quite hard and if I had a solution to it I'd probably be implementing it instead of waxing poetic on HN.
    - AnthonyMouse 6y ago
      > And with machine learning we're building our biases into the algorithms explicitly; see all those "resume parsers are racist" articles. It's even worse than that, because it's so hard to tell if this is even happening. Suppose guidance counselors in predominantly black schools are telling kids to focus on athletics and the ones in predominantly white schools tell them to focus on intellectual extracurriculars. Then a resume parser sorts people with athletics listed into physical jobs and people with intellectual pursuits listed into intellectual jobs, which of course results in the black applicants getting callbacks for the lower paying jobs. This is pretty clearly the guidance counselors causing the disparity rather than the algorithm, but we only know that because it was stipulated in the hypothetical. In real life you may not have enough data to be able to discern the underlying cause. In other words, you don't know what the baseline racial disparity is based on all of the non-racial factors that correlate with race, so you don't know if the problem is in the algorithm or was caused somewhere upstream and the algorithm is only producing the accurate outputs for its inputs. In theory you can evaluate this by checking up on how the candidates hired do compared to how the algorithm predicted they would do, but that's a noisy signal (what does "doing well" mean?), you might not have a large enough sample size to get meaningful results, and it has a lag time of potentially several years, by which point you may already be using a different algorithm. It's a hard problem.
      - abfan1127 6y ago
jjeaff 6y ago
At the same time, technology offers some solutions to this like ssl certificates. Perhaps there is a business model for bespoke extended validation ssl certs where, for example, they are tied to a physical address.
- falcolas 6y ago
  SSL certificates, which require you to trust someone you’ve never met with the entirety of your digital life. From your social connections to your bank account and everything in between, you’re putting your faith not only in their honesty, but in their ability to limit mistakes. Not that there’s a better solution yet, but technical solutions like SSL certificates still boil down to trust, and in whom you place your.
  - zxcmx 6y ago
    This is also true for your computer hardware, the software you run, the medicines you take, the food you eat, the water you drink... I guess my point is that that human trust is so pervasive and fundamental that there are diminishing returns to eliminating it. What will always remain in style I think is accountability. So trust, but have supply chain integrity such that you know who it is that you're trusting and who is responsible when things go wrong.
selfhoster11 6y ago
"Mechanical honesty" worked because it was much harder to have automated "if statement" mechanisms with non-electronic machinery. And then electronic computers made it much, much easier.
sxzxs 6y ago
Can you elaborate more on (or give an example of) "mechanical honesty"? From the outside looking in, it doesn't seem unfeasible to implement biases in mechanical systems (especially government departments!), but I haven't given it a lot of thought.
- samatman 6y ago
  The phone book. It's a public mapping of people and businesses with how to contact them. Any dishonesty would be noticed, because everyone has the same phone book. Contrast with getting a phone number from $large-company, be it Yelp, Google or otherwise. The only thing stopping them from selling this privilege to an intermediary is their reputation, and clearly where Yelp is concerned, this isn't enough. One could argue that Google has been exploiting this since the creation of AdWords: they will happily sell the right to advertise on a search for your exact company name to a competitor, a nice income stream which verges on extortion. It's even possible to deliver personalized dishonest results; applications including search engines know who you are, and can lie to you and only you. It's certainly possible to be 'mechanically dishonest', but no question that contemporary technology makes it easier and more lucrative.
  - Gibbon1 6y ago
    Another example is targeting job ads to particular demographics. You can put out an ad that only white men between the ages of 22 and 30 who come from an upper class background will see. And women, minorities and people from poor backgrounds, not only won't see them, they won't know they exist.
  - Qwertious 6y ago
    I don't think Yelp has done anything other than "lie in the phone book" so far, so it's a bad example. A better example might be where Uber made their app turn off a bunch of shady privacy a violating settings, but only if the phone was geographically located at Apple HQ.
hedora 6y ago
A few people seemed to have missed an important step in the argument: Mechanical systems used to be simple enough to be (mostly) auditable. For instance, if you have a wind up clock on the wall and a chalkboard schedule in an office, you can be sure it tells the same time and the same schedule, regardless of who looks at it. Like an evil secretary, meeting room scheduling tablets could be configured to gaslight particular employee by telling them they have a room reservation when they don’t, making them late/early, steering them to embarrassing interactions with upper management, etc. until they’re fired.
- harpiaharpyja 6y ago
  I would argue that this auditability is a bigger factor than GP's "computing power" claim. It's the result of hiding the process from users under the name of convenience.
CyberDildonics 6y ago
This is nonsense, computing power has nothing to do with it switching out phone numbers.
1vuio0pswjnm7 6y ago
"The public still retains the expectations of the mechanical honesty era, and is an easy mark." Is this statement suggesting that the "work" of the businesses exploiting these unrealistic expectations is "easy". When a country's most successful companies are "middlemen" built upon a foundation of easily exploiting misplaced trust (misplaced belief in mechanical honesty) and governments actively seek to encourage more such "entrepreneurship", what does that mean for the future.
kens 6y ago
> the start of an era of "mechanical honesty" I'd like to mention the under-appreciated development of the cash register in the 1880s. Prior to the cash register, the owner had to trust cashiers not to pocket money. Cash registers became enormously popular and revolutionized sales since they kept everyone honest, as well as letting business owners know what was going on. (The book "Before the Computer: IBM, NCR, Burroughs, & Remington Rand & the Industry They Created" goes into a lot of detail on this.)
- frogpelt 6y ago
  The cash register was originally called Ritty's Incorruptible Cashier. Source: https://en.wikipedia.org/wiki/Cash_register https://en.wikipedia.org/wiki/Cash_register
- m463 6y ago
  And remember that receipts aren't for you. In fact, it might cost them money because it facilitates returns. Businesses don't care if they give you a receipt. They care if a receipt was generated by the machine, and are a check on the employee. those signs "if we didn't give you a receipt your meal is free" take on a different meaning in that light.
  - rualca 6y ago
    > They care if a receipt was generated by the machine, and are a check on the employee. Aren't receipts used to track the store's revenue, as in the numbers reported to your local tax agency? As far as I know, it's common for cashiers to have a daily/weekly/monthly slack value that they are allowed to report as a difference between the number reported by cash registers and the amount of cash delivered, because SNAFUs do happen during a day's work.
    - hnick 6y ago
      This could be stored on a computer now, though. But numbers can be fudged. Taiwan had an interesting way around this with a receipt lottery. Ask for a receipt, it has a number on it that enters you in a lottery. Presumably the generation of this number means the sale was reported. https://guidetotaipei.com/article/taiwan-receipt-lottery-%E7%B5%B1%E4%B8%80%E7%99%BC%E7%A5%A8%E4%B8%AD%E7%8D%8E%E8%99%9F%E7%A2%BC https://guidetotaipei.com/article/taiwan-receipt-lottery-%E7...
redis_mlc 6y ago
> He was concerned that the human operators were diverting his customers (specifically, the wife of one of his competitors was an operator). So like Dominion voting machines diverting votes from Trump to Biden. How apt.
rconti 6y ago
I find it fascinating when someone finds a historical parallel to a situation that we think is novel and unprecedented, reminding us that nothing is truly new, and it's not as bad as we think. Then takes a hard left turn into "but this time it's different because reasons" and gets us back on the expected "sky is falling" path.
durnygbur 6y ago
> exchange was the start of an era of "mechanical honesty" This is only the early stage. With more parties and money involved "mechanical honesty" turns into betting race and the "machanical exchange" redirects to where the highest bidders desires.
eru 6y ago
Mail and telegraph were perhaps an earlier move in that direction. A paper trail keeps people honest to some degree, too.
tim333 6y ago
>The public still retains the expectations of the mechanical honesty era As part of the public I more expect almost everything internet to be biased by commercial interests. Best you can do is seek out the less corrupted.

by ajb 6y ago

castratikron 6y ago

I would say there's no moral difference between a human operator and a mechanical operator whose behavior is controlled by a human. Configuring the operator to drop calls from a particular area code would be an example in the case you provided.

ajb 6y ago
Sure. It just used to be harder to build in, and to do so precisely. In your example, dropping calls from an area code, that's a relatively coarse filter. There's a whole strand of academic research [1] arguing about whether people can build prejudice into technological artifacts, eg one disputed example was an underpass with low clearance, supposedly so that white middle class people in cars could pass, but black poor people in buses could not [1, p123]. These days that dispute is ridiculous - obviously prejudice and bias can be built into technological artifacts. It's one "if statement" away. [1] "Do Artifacts have politics" https://www.cc.gatech.edu/~beki/cs4001/Winner.pdf https://www.cc.gatech.edu/~beki/cs4001/Winner.pdf
- krisgee 6y ago
  >These days that dispute is ridiculous - obviously prejudice and bias can be built into technological artifacts. It's one "if statement" away. And with machine learning we're building our biases into the algorithms explicitly; see all those "resume parsers are racist" articles. This is clearly an issue that needs to be dealth with but the problem is that it's not a problem we can engineer our way of, we need to confront the biases that we have and try and build ways to expose biases we don't even realize we have. This is obviously quite hard and if I had a solution to it I'd probably be implementing it instead of waxing poetic on HN.
  - AnthonyMouse 6y ago
    > And with machine learning we're building our biases into the algorithms explicitly; see all those "resume parsers are racist" articles. It's even worse than that, because it's so hard to tell if this is even happening. Suppose guidance counselors in predominantly black schools are telling kids to focus on athletics and the ones in predominantly white schools tell them to focus on intellectual extracurriculars. Then a resume parser sorts people with athletics listed into physical jobs and people with intellectual pursuits listed into intellectual jobs, which of course results in the black applicants getting callbacks for the lower paying jobs. This is pretty clearly the guidance counselors causing the disparity rather than the algorithm, but we only know that because it was stipulated in the hypothetical. In real life you may not have enough data to be able to discern the underlying cause. In other words, you don't know what the baseline racial disparity is based on all of the non-racial factors that correlate with race, so you don't know if the problem is in the algorithm or was caused somewhere upstream and the algorithm is only producing the accurate outputs for its inputs. In theory you can evaluate this by checking up on how the candidates hired do compared to how the algorithm predicted they would do, but that's a noisy signal (what does "doing well" mean?), you might not have a large enough sample size to get meaningful results, and it has a lag time of potentially several years, by which point you may already be using a different algorithm. It's a hard problem.
    - abfan1127 6y ago

jjeaff 6y ago

At the same time, technology offers some solutions to this like ssl certificates. Perhaps there is a business model for bespoke extended validation ssl certs where, for example, they are tied to a physical address.

falcolas 6y ago
SSL certificates, which require you to trust someone you’ve never met with the entirety of your digital life. From your social connections to your bank account and everything in between, you’re putting your faith not only in their honesty, but in their ability to limit mistakes. Not that there’s a better solution yet, but technical solutions like SSL certificates still boil down to trust, and in whom you place your.
- zxcmx 6y ago
  This is also true for your computer hardware, the software you run, the medicines you take, the food you eat, the water you drink... I guess my point is that that human trust is so pervasive and fundamental that there are diminishing returns to eliminating it. What will always remain in style I think is accountability. So trust, but have supply chain integrity such that you know who it is that you're trusting and who is responsible when things go wrong.

selfhoster11 6y ago

"Mechanical honesty" worked because it was much harder to have automated "if statement" mechanisms with non-electronic machinery. And then electronic computers made it much, much easier.

sxzxs 6y ago

Can you elaborate more on (or give an example of) "mechanical honesty"? From the outside looking in, it doesn't seem unfeasible to implement biases in mechanical systems (especially government departments!), but I haven't given it a lot of thought.

samatman 6y ago
The phone book. It's a public mapping of people and businesses with how to contact them. Any dishonesty would be noticed, because everyone has the same phone book. Contrast with getting a phone number from $large-company, be it Yelp, Google or otherwise. The only thing stopping them from selling this privilege to an intermediary is their reputation, and clearly where Yelp is concerned, this isn't enough. One could argue that Google has been exploiting this since the creation of AdWords: they will happily sell the right to advertise on a search for your exact company name to a competitor, a nice income stream which verges on extortion. It's even possible to deliver personalized dishonest results; applications including search engines know who you are, and can lie to you and only you. It's certainly possible to be 'mechanically dishonest', but no question that contemporary technology makes it easier and more lucrative.
- Gibbon1 6y ago
  Another example is targeting job ads to particular demographics. You can put out an ad that only white men between the ages of 22 and 30 who come from an upper class background will see. And women, minorities and people from poor backgrounds, not only won't see them, they won't know they exist.
- Qwertious 6y ago
  I don't think Yelp has done anything other than "lie in the phone book" so far, so it's a bad example. A better example might be where Uber made their app turn off a bunch of shady privacy a violating settings, but only if the phone was geographically located at Apple HQ.

hedora 6y ago

A few people seemed to have missed an important step in the argument: Mechanical systems used to be simple enough to be (mostly) auditable. For instance, if you have a wind up clock on the wall and a chalkboard schedule in an office, you can be sure it tells the same time and the same schedule, regardless of who looks at it. Like an evil secretary, meeting room scheduling tablets could be configured to gaslight particular employee by telling them they have a room reservation when they don’t, making them late/early, steering them to embarrassing interactions with upper management, etc. until they’re fired.

harpiaharpyja 6y ago
I would argue that this auditability is a bigger factor than GP's "computing power" claim. It's the result of hiding the process from users under the name of convenience.

CyberDildonics 6y ago

This is nonsense, computing power has nothing to do with it switching out phone numbers.

1vuio0pswjnm7 6y ago

"The public still retains the expectations of the mechanical honesty era, and is an easy mark." Is this statement suggesting that the "work" of the businesses exploiting these unrealistic expectations is "easy". When a country's most successful companies are "middlemen" built upon a foundation of easily exploiting misplaced trust (misplaced belief in mechanical honesty) and governments actively seek to encourage more such "entrepreneurship", what does that mean for the future.

kens 6y ago

> the start of an era of "mechanical honesty" I'd like to mention the under-appreciated development of the cash register in the 1880s. Prior to the cash register, the owner had to trust cashiers not to pocket money. Cash registers became enormously popular and revolutionized sales since they kept everyone honest, as well as letting business owners know what was going on. (The book "Before the Computer: IBM, NCR, Burroughs, & Remington Rand & the Industry They Created" goes into a lot of detail on this.)

frogpelt 6y ago
The cash register was originally called Ritty's Incorruptible Cashier. Source: https://en.wikipedia.org/wiki/Cash_register https://en.wikipedia.org/wiki/Cash_register
m463 6y ago
And remember that receipts aren't for you. In fact, it might cost them money because it facilitates returns. Businesses don't care if they give you a receipt. They care if a receipt was generated by the machine, and are a check on the employee. those signs "if we didn't give you a receipt your meal is free" take on a different meaning in that light.
- rualca 6y ago
  > They care if a receipt was generated by the machine, and are a check on the employee. Aren't receipts used to track the store's revenue, as in the numbers reported to your local tax agency? As far as I know, it's common for cashiers to have a daily/weekly/monthly slack value that they are allowed to report as a difference between the number reported by cash registers and the amount of cash delivered, because SNAFUs do happen during a day's work.
  - hnick 6y ago
    This could be stored on a computer now, though. But numbers can be fudged. Taiwan had an interesting way around this with a receipt lottery. Ask for a receipt, it has a number on it that enters you in a lottery. Presumably the generation of this number means the sale was reported. https://guidetotaipei.com/article/taiwan-receipt-lottery-%E7%B5%B1%E4%B8%80%E7%99%BC%E7%A5%A8%E4%B8%AD%E7%8D%8E%E8%99%9F%E7%A2%BC https://guidetotaipei.com/article/taiwan-receipt-lottery-%E7...

redis_mlc 6y ago

> He was concerned that the human operators were diverting his customers (specifically, the wife of one of his competitors was an operator). So like Dominion voting machines diverting votes from Trump to Biden. How apt.

rconti 6y ago

I find it fascinating when someone finds a historical parallel to a situation that we think is novel and unprecedented, reminding us that nothing is truly new, and it's not as bad as we think. Then takes a hard left turn into "but this time it's different because reasons" and gets us back on the expected "sky is falling" path.

durnygbur 6y ago

> exchange was the start of an era of "mechanical honesty" This is only the early stage. With more parties and money involved "mechanical honesty" turns into betting race and the "machanical exchange" redirects to where the highest bidders desires.

eru 6y ago

Mail and telegraph were perhaps an earlier move in that direction. A paper trail keeps people honest to some degree, too.

tim333 6y ago

>The public still retains the expectations of the mechanical honesty era As part of the public I more expect almost everything internet to be biased by commercial interests. Best you can do is seek out the less corrupted.