4 ms·

> Society should discuss if recommendation algorithms and targeted advertisement must be regulated. It is known that people have more probability of sharing con

by SllX 3y ago

> Society should discuss if recommendation algorithms and targeted advertisement must be regulated. It is known that people have more probability of sharing content that shocks them more. We can have that discussion now: “No.” To the extent that you want to apply any law, it’s to content, not the recommendation algorithm: copyright infringing stuff[1], pornography, snuff films, and other already illegal stuff. I don’t like people like say, Tucker Carlson, their whole schtick is making themselves viral by making a very shocking mountain out of a mole hill and riling people up, but the government also has no place passing legislation against or “regulating” that. [1]: I’m setting aside my own issues with extant copyright laws for the sake of a more productive discussion on the proposed topic, but I’m not entirely happy with them either.

smt88 3y ago
> I don’t like people like say, Tucker Carlson, their whole schtick is making themselves viral by making a very shocking mountain out of a mole hill and riling people up, but the government also has no place passing legislation against or “regulating” that. You can regulate recommendation algorithms (by making them transparent, opt-in, etc.) without infringing on Carlson's right to speech. My proposed regulations would be that all automated recommendation algorithms are: 1. clearly labelled as automated (not "editor's picks" or "other people are watching," which connote social validation that may not exist) 2. opt-in (no recommendations shown unless you request them) 3. auditable (for X days after a recommendation is made, it should be possible to see each step involved in it being recommended to that user) None of these infringe on anyone's speech. They just make it harder for massive, irresponsible companies to do psychological experiments on billions of people without transparency or culpability.
- SllX 3y ago
  My concerns aren’t purely first amendment related. Some of this is just product decisions and business progresses. 1. What does this labeling get you? Taking YouTube as an example, it is basically some combination of what other people watch, what channels I Subscribe to, what’s picked out and even in those exceptional circumstances where I’m like 1 of maybe 40 views ever on a video, someone uploaded something similar to something I watched before. That’s the point of the algorithm: it takes multiple inputs, and spits them out in an automated fashion. I think most people also understand that these are machine generated recommendations, even if they might phrase it differently. So again, what does it get you? 2. Why opt-in? That’s a product decision. If you’re using the service, you’re opting into how the service functions as is, and the service provider may provide additional options for how it works beyond that, but why does the government get to make that kind of product decision? Then, if YouTube isn’t providing a list of recommendations when you load it, what are you seeing instead? Is that going to be decided by the government as well? 3. Why are users entitled to see an audit of what’s basically proprietary product information about how the website works, effectively making that information available to competitors as well? > None of these infringe on anyone's speech. They just make it harder for massive, irresponsible companies to do psychological experiments on billions of people without transparency or culpability. This experimentation is a genuine ethics concern, and also from a practical standpoint I can’t walk up to SaaS that I paid for and expect it to work the same way as I had been using it every time. I would not support your proposed remedies though, but there may be a productive route in governing psychological experimentation.
  - smt88 3y ago
    > 1. What does this labeling get you? It tells you that the recommendation is blind and wasn't reviewed by a human. For example, much of the right-wing US media audience primarily consumes Fox News, while non-right audiences are spread across outlets. So let's say your algorithm is based on what other people view. The entire feed would be Fox (and this was a common issue with Google News before I stopped using it). A minority of people are making it look like a majority of people are consuming that media, which legitimizes it. "Google" (a legitimate company) might be pushing a conspiracy theory as the #1 trending news story. > 2. Why opt-in? That’s a product decision. All experiments should be opt-in. We don't know the effects of algorithmic recommendations, especially because they constantly evolve. > why does the government get to make that kind of product decision? Because the government's job is to make it illegal to do things that are dangerous. Why does the government get to decide that cars have published safety ratings or that food has lists of ingredients? If (for example) Facebook wants to be responsible for its algorithm's recommendations, then sure, take the government out of it. It becomes Facebook's speech, which is protected. But Facebook is hiding behind Section 230, and that should come with additional oversight and restrictions. > 3. Why are users entitled to see an audit of what’s basically proprietary product information about how the website works, effectively making that information available to competitors as well? Seeing how something worked doesn't tell you how to build it yourself. If TikTok tells me how it recommended a video to me, that information isn't nearly deterministic enough for me to reverse-engineer their algorithm. In fact, their algorithm is impossible to replicate without all of their training data (which remains proprietary). I agree that some "secrets" would be exposed, but they wouldn't be useful to a competitor. They'd be useful to consumers who want to choose the most competitive product.
    - nradov 3y ago
      Facebook is not hiding behind Section 230 for their recommendation algorithm. They are hiding behind the First Amendment. By making recommendations, Facebook is expressing an opinion. US courts take a dim view of government attempts to restrict the expression of opinions. There is an extensive body of case law on this issue.
- andrei_says_ 3y ago
  > massive, irresponsible companies to do psychological experiments on billions of people without transparency or culpability Exactly. While measuring and optimizing for monetization, completely ignoring and willing to sacrifice individual mental health, and society in general. The division and mass brainwashing we are experiencing do have a cost - that happens to be an external used one from the perspective of the content distribution / advertising platforms. At this point people are openly being lied that a significant part of the population are groomers and pedophiles which 100% targets blind destructive unconditional rage and unreconcileable division. For the benefit of distraction, political division and engagement. We need a force to limit the effects of the infinite cynicism of the people doing the brainwashing. We could at least not auto-amplify their voices.
rsynnott 3y ago
> but the government also has no place passing legislation against or “regulating” that. Banning recommendations would probably be fairly excessive, at least if the recommendation is consensual (that is, if the site has the user's informed consent to profile them for recommendation purposes). However, it would arguably be pretty reasonable, and in line with how, say, 'native' advertising is generally treated, to require sites to include a warning to the effect that this is a machine recommendation which is vulnerable to manipulation. As is, users may read recommendations as "this is what my friends are watching" or "this is what a human has chosen as recommendations for everyone", whereas in fact they are generally "this is the platform's attempt at maximising the ad revenue they can squeeze out of me, but also it may have been manipulated by malicious actors and the platform knows that". This should probably be made clearer to users.
- SllX 3y ago
  Is that really the case though? I think most people do understand that it is largely a non-human automated process that figures out what to show them. I asked someone else what the labeling really gets you, but your argument seems to be that the transparency is an end in and of itself. So what about just a general disclosure that this service relies on automated recommendations?

by SllX 3y ago

smt88 3y ago

> I don’t like people like say, Tucker Carlson, their whole schtick is making themselves viral by making a very shocking mountain out of a mole hill and riling people up, but the government also has no place passing legislation against or “regulating” that. You can regulate recommendation algorithms (by making them transparent, opt-in, etc.) without infringing on Carlson's right to speech. My proposed regulations would be that all automated recommendation algorithms are: 1. clearly labelled as automated (not "editor's picks" or "other people are watching," which connote social validation that may not exist) 2. opt-in (no recommendations shown unless you request them) 3. auditable (for X days after a recommendation is made, it should be possible to see each step involved in it being recommended to that user) None of these infringe on anyone's speech. They just make it harder for massive, irresponsible companies to do psychological experiments on billions of people without transparency or culpability.

SllX 3y ago
My concerns aren’t purely first amendment related. Some of this is just product decisions and business progresses. 1. What does this labeling get you? Taking YouTube as an example, it is basically some combination of what other people watch, what channels I Subscribe to, what’s picked out and even in those exceptional circumstances where I’m like 1 of maybe 40 views ever on a video, someone uploaded something similar to something I watched before. That’s the point of the algorithm: it takes multiple inputs, and spits them out in an automated fashion. I think most people also understand that these are machine generated recommendations, even if they might phrase it differently. So again, what does it get you? 2. Why opt-in? That’s a product decision. If you’re using the service, you’re opting into how the service functions as is, and the service provider may provide additional options for how it works beyond that, but why does the government get to make that kind of product decision? Then, if YouTube isn’t providing a list of recommendations when you load it, what are you seeing instead? Is that going to be decided by the government as well? 3. Why are users entitled to see an audit of what’s basically proprietary product information about how the website works, effectively making that information available to competitors as well? > None of these infringe on anyone's speech. They just make it harder for massive, irresponsible companies to do psychological experiments on billions of people without transparency or culpability. This experimentation is a genuine ethics concern, and also from a practical standpoint I can’t walk up to SaaS that I paid for and expect it to work the same way as I had been using it every time. I would not support your proposed remedies though, but there may be a productive route in governing psychological experimentation.
- smt88 3y ago
  > 1. What does this labeling get you? It tells you that the recommendation is blind and wasn't reviewed by a human. For example, much of the right-wing US media audience primarily consumes Fox News, while non-right audiences are spread across outlets. So let's say your algorithm is based on what other people view. The entire feed would be Fox (and this was a common issue with Google News before I stopped using it). A minority of people are making it look like a majority of people are consuming that media, which legitimizes it. "Google" (a legitimate company) might be pushing a conspiracy theory as the #1 trending news story. > 2. Why opt-in? That’s a product decision. All experiments should be opt-in. We don't know the effects of algorithmic recommendations, especially because they constantly evolve. > why does the government get to make that kind of product decision? Because the government's job is to make it illegal to do things that are dangerous. Why does the government get to decide that cars have published safety ratings or that food has lists of ingredients? If (for example) Facebook wants to be responsible for its algorithm's recommendations, then sure, take the government out of it. It becomes Facebook's speech, which is protected. But Facebook is hiding behind Section 230, and that should come with additional oversight and restrictions. > 3. Why are users entitled to see an audit of what’s basically proprietary product information about how the website works, effectively making that information available to competitors as well? Seeing how something worked doesn't tell you how to build it yourself. If TikTok tells me how it recommended a video to me, that information isn't nearly deterministic enough for me to reverse-engineer their algorithm. In fact, their algorithm is impossible to replicate without all of their training data (which remains proprietary). I agree that some "secrets" would be exposed, but they wouldn't be useful to a competitor. They'd be useful to consumers who want to choose the most competitive product.
  - nradov 3y ago
    Facebook is not hiding behind Section 230 for their recommendation algorithm. They are hiding behind the First Amendment. By making recommendations, Facebook is expressing an opinion. US courts take a dim view of government attempts to restrict the expression of opinions. There is an extensive body of case law on this issue.
andrei_says_ 3y ago
> massive, irresponsible companies to do psychological experiments on billions of people without transparency or culpability Exactly. While measuring and optimizing for monetization, completely ignoring and willing to sacrifice individual mental health, and society in general. The division and mass brainwashing we are experiencing do have a cost - that happens to be an external used one from the perspective of the content distribution / advertising platforms. At this point people are openly being lied that a significant part of the population are groomers and pedophiles which 100% targets blind destructive unconditional rage and unreconcileable division. For the benefit of distraction, political division and engagement. We need a force to limit the effects of the infinite cynicism of the people doing the brainwashing. We could at least not auto-amplify their voices.

rsynnott 3y ago

> but the government also has no place passing legislation against or “regulating” that. Banning recommendations would probably be fairly excessive, at least if the recommendation is consensual (that is, if the site has the user's informed consent to profile them for recommendation purposes). However, it would arguably be pretty reasonable, and in line with how, say, 'native' advertising is generally treated, to require sites to include a warning to the effect that this is a machine recommendation which is vulnerable to manipulation. As is, users may read recommendations as "this is what my friends are watching" or "this is what a human has chosen as recommendations for everyone", whereas in fact they are generally "this is the platform's attempt at maximising the ad revenue they can squeeze out of me, but also it may have been manipulated by malicious actors and the platform knows that". This should probably be made clearer to users.

SllX 3y ago
Is that really the case though? I think most people do understand that it is largely a non-human automated process that figures out what to show them. I asked someone else what the labeling really gets you, but your argument seems to be that the transparency is an end in and of itself. So what about just a general disclosure that this service relies on automated recommendations?