4 ms·

I wish I could upvote this 10x though I see atm that is not necessarily. A very nice summary of priors and predicates that need to be grokked! To these I myse

I wish I could upvote this 10x though I see atm that is not necessarily. A very nice summary of priors and predicates that need to be grokked! To these I myself often add several aspects looking more specifically what "social graph" means and what the so-called algorithm means, with lemmas about "engagement" and its industry... ...and the underbelly of how the algorithm works, the constellation of surveillance tracking entity resolution and targeting that make up the actual functioning of "teh algorithm." These regularly lead me back to trying to find pithy summations of specific paths through this mire, such as "deplatforming is not censorship, it is failure to provide or sell a megaphone," and, "we need to reason about 1A the way we do about 2A," (ie in terms of understood liberty vs aggregate societal harms). Another such summation is simply to observe that the subset of "the information out there" that any individual sees, and which is amplified within their own social circles online and off-, is the most hotly contested property in the market today. In many markets. Against this individual integrity and good faith often appears to be but a reed in the wind.

bcrl 5y ago
Exactly. The fundamental difference in advertising today compared to 30 years ago is that for the entire of history until the last 2 decades, it was expensive to target individuals. Ads on TV and newspapers have to adhere to standards established by organizations like the Advertising Standards Council. Malicious ads or propaganda that target individuals with misinformation or malware or phishing can be bought on modern platforms for pennies or mere fractions of a penny. Nobody is being held to account for the damage they are doing to our society. In the past at least the cost of postage was a barrier to entry, but today? This desperately needs to be fixed.
- FFRefresh 5y ago
  What misinformation have you been convinced of because you saw a targeted advertisement? Or a slightly different version: What views do you hold that originated from a targeted ad? Or for some more specific examples: In elections, do you start out looking to elect one candidate and then you see a targeted ad and you change your vote because of it? For COVID vaccines, did you start off with a stance and then you see a targeted ad and you change your stance because of that ad?
  - bcrl 5y ago
    As I post every so often here on Hacker News, have a look at this ad my elderly father got a couple of years ago just before COVID hit: https://imgur.com/a/Lk85lST https://imgur.com/a/Lk85lST . It's a perfectly innocent search by someone looking for help with a problem that directs towards an ad for a phishing site. The worst part is that it's highly targeted. I tried doing EXACTLY the same search from my own phone and laptop and got completely different results despite coming in from the same wifi network at the same location and the same time. This isn't acceptable. If malicious ads couldn't be targeted like that so that more people saw the same unacceptable content, it sure as hell would be reported a lot more. Instead, someone has made a decision that nobody should be held accountable. It generates a profit, so it's OK, right?
    - FFRefresh 5y ago
      Oh totally, that ad seems malicious. I was just curious about the misinformation angle that you mentioned and how you in particular have been duped or had your opinions changed because of a targeted ad.
      - bcrl 5y ago
        I don't see those ads. My filter bubble is completely different. Personally, I think it would be beneficial for society if all of us were exposed to information that falls outside of what our individual filter bubble normally shows us from time to time. Another example that came up in my life yesterday relates to the trucker protest here in Ottawa. One of the news stories from last month about restrictions is being heavily promoted by people sympathetic to the cause. However, in the intervening 4 weeks since this particular story was published, those restrictions were lifted and are no longer being enforced. The people promoting the old article do so because it benefits their narrative, but they are effectively spreading misinformation now by failing to acknowledge that the content is no longer true. That's a problem, and that kind of behaviour is not easy to solve. Maybe someday our web browsers will have the ability to flag content that is obsolete, but that's a pipe dream at present.
tekknik 5y ago
> These regularly lead me back to trying to find pithy summations of specific paths through this mire, such as "deplatforming is not censorship, it is failure to provide or sell a megaphone," and, "we need to reason about 1A the way we do about 2A," (ie in terms of understood liberty vs aggregate societal harms). This is one of the scariest paragraphs I’ve seen on HN in some time. Who exactly gets to reason about these changes? How do we change 1A and no cause censorship? Our political parties have thrown the entire nation into division and one side wants to limit speech while the other doesn’t. So what do we do? Implement a dictatorship? While we’re here, 2A has not cause societal harms, the humans did. Separate them already. The amount of gun owners vs those shooting up places have a very very very wide margin.
- aaroninsf 5y ago
  By "we" I mean, we as a society. Me, you, all of us. By "reason" I mean, we need to have open debate about where the appropriate line today is in various domains between competing goods and aims. The conflict is often as here, between collective interest, and individual liberty. Not the time or place reopen/rehash well-trod arguments, but I will observe that from the perspective of someone who believes in gun control, and to my point about the analogy, our collective problem actually is precisely that humans are regardless of good faith intentions, defined by predictably acting out of accord with those intentions, and hence we as society have reason to seek to regulate and mitigate the impact of force multipliers. In the US, guns are the most well-known force multiplier in the domain of interpersonal conflict. But today the force multiplier which arguably has orders of magnitude more impact, is "speech." The distinction between "people" and their tools is IMO specious. Humans have agency; what is at issue is that given that agency, what do we agree about tool use? Reasoning about "free speech" does not === censorship. It means, we need to educate people to have a better mental model of what our industry (here on HN) is capable of doing and is doing, in the feedback system which marries surveillance and the construction of personalized information streams. The short-hand way for describing this is a good start: it amounts to stochastic mind control. No individual may be "programmed," but our industry and its clients put billions of dollars and literal and political fortunes among other resources, into the presumption that our current tooling is capable of producing discernible changes in collective behavior. We argue here and elsewhere regularly as professionals about the ethics and consequences of using this system to e.g. provoke consumption. When I say we need to reason about speech, I am myself most interested in us coming to terms with, and finding a way to survive as a society which e.g. values individual agency (call it liberty if you prefer), when that same tooling over the last decade especially has increasingly been used and with exponentially increasing power, to drive social and political behavior. In the US this is most evident and most clearly a present danger wrt the fomentation of tribalism, in specific tribes which are largely defined by their anti-democratic, anti-communitarian, authoratarianism, married by convenience and inclination to racism, homophobia, misogyny, and the exhausting catalog of other permanent social horrors. Twitter, or the online ad market, are no more "free speech" in an Originalist sense than a contemporary semi-automatic weapon is a musket. In bother cases, we need to understand and apply Constitutional law to find the appropriate balance between collective and individual interests, and what I saying is, we need to do so with an honest discussion of evolving power dynamics given that individuals now have at least hypothetical access to force multipliers. To make this concrete: the 2016 and 2020 elections were subject to asymmetrical information warfare. Asymmetry in the terrorist sense: a small but motivated group, such as the GRU, is capable of provoking catastrophic disruption. I do not posit any particular answers. But we need to have the hard discussion, the sooner the better.

bcrl 5y ago

Exactly. The fundamental difference in advertising today compared to 30 years ago is that for the entire of history until the last 2 decades, it was expensive to target individuals. Ads on TV and newspapers have to adhere to standards established by organizations like the Advertising Standards Council. Malicious ads or propaganda that target individuals with misinformation or malware or phishing can be bought on modern platforms for pennies or mere fractions of a penny. Nobody is being held to account for the damage they are doing to our society. In the past at least the cost of postage was a barrier to entry, but today? This desperately needs to be fixed.

FFRefresh 5y ago
What misinformation have you been convinced of because you saw a targeted advertisement? Or a slightly different version: What views do you hold that originated from a targeted ad? Or for some more specific examples: In elections, do you start out looking to elect one candidate and then you see a targeted ad and you change your vote because of it? For COVID vaccines, did you start off with a stance and then you see a targeted ad and you change your stance because of that ad?
- bcrl 5y ago
  As I post every so often here on Hacker News, have a look at this ad my elderly father got a couple of years ago just before COVID hit: https://imgur.com/a/Lk85lST https://imgur.com/a/Lk85lST . It's a perfectly innocent search by someone looking for help with a problem that directs towards an ad for a phishing site. The worst part is that it's highly targeted. I tried doing EXACTLY the same search from my own phone and laptop and got completely different results despite coming in from the same wifi network at the same location and the same time. This isn't acceptable. If malicious ads couldn't be targeted like that so that more people saw the same unacceptable content, it sure as hell would be reported a lot more. Instead, someone has made a decision that nobody should be held accountable. It generates a profit, so it's OK, right?
  - FFRefresh 5y ago
    Oh totally, that ad seems malicious. I was just curious about the misinformation angle that you mentioned and how you in particular have been duped or had your opinions changed because of a targeted ad.
    - bcrl 5y ago
      I don't see those ads. My filter bubble is completely different. Personally, I think it would be beneficial for society if all of us were exposed to information that falls outside of what our individual filter bubble normally shows us from time to time. Another example that came up in my life yesterday relates to the trucker protest here in Ottawa. One of the news stories from last month about restrictions is being heavily promoted by people sympathetic to the cause. However, in the intervening 4 weeks since this particular story was published, those restrictions were lifted and are no longer being enforced. The people promoting the old article do so because it benefits their narrative, but they are effectively spreading misinformation now by failing to acknowledge that the content is no longer true. That's a problem, and that kind of behaviour is not easy to solve. Maybe someday our web browsers will have the ability to flag content that is obsolete, but that's a pipe dream at present.

tekknik 5y ago

> These regularly lead me back to trying to find pithy summations of specific paths through this mire, such as "deplatforming is not censorship, it is failure to provide or sell a megaphone," and, "we need to reason about 1A the way we do about 2A," (ie in terms of understood liberty vs aggregate societal harms). This is one of the scariest paragraphs I’ve seen on HN in some time. Who exactly gets to reason about these changes? How do we change 1A and no cause censorship? Our political parties have thrown the entire nation into division and one side wants to limit speech while the other doesn’t. So what do we do? Implement a dictatorship? While we’re here, 2A has not cause societal harms, the humans did. Separate them already. The amount of gun owners vs those shooting up places have a very very very wide margin.

aaroninsf 5y ago
By "we" I mean, we as a society. Me, you, all of us. By "reason" I mean, we need to have open debate about where the appropriate line today is in various domains between competing goods and aims. The conflict is often as here, between collective interest, and individual liberty. Not the time or place reopen/rehash well-trod arguments, but I will observe that from the perspective of someone who believes in gun control, and to my point about the analogy, our collective problem actually is precisely that humans are regardless of good faith intentions, defined by predictably acting out of accord with those intentions, and hence we as society have reason to seek to regulate and mitigate the impact of force multipliers. In the US, guns are the most well-known force multiplier in the domain of interpersonal conflict. But today the force multiplier which arguably has orders of magnitude more impact, is "speech." The distinction between "people" and their tools is IMO specious. Humans have agency; what is at issue is that given that agency, what do we agree about tool use? Reasoning about "free speech" does not === censorship. It means, we need to educate people to have a better mental model of what our industry (here on HN) is capable of doing and is doing, in the feedback system which marries surveillance and the construction of personalized information streams. The short-hand way for describing this is a good start: it amounts to stochastic mind control. No individual may be "programmed," but our industry and its clients put billions of dollars and literal and political fortunes among other resources, into the presumption that our current tooling is capable of producing discernible changes in collective behavior. We argue here and elsewhere regularly as professionals about the ethics and consequences of using this system to e.g. provoke consumption. When I say we need to reason about speech, I am myself most interested in us coming to terms with, and finding a way to survive as a society which e.g. values individual agency (call it liberty if you prefer), when that same tooling over the last decade especially has increasingly been used and with exponentially increasing power, to drive social and political behavior. In the US this is most evident and most clearly a present danger wrt the fomentation of tribalism, in specific tribes which are largely defined by their anti-democratic, anti-communitarian, authoratarianism, married by convenience and inclination to racism, homophobia, misogyny, and the exhausting catalog of other permanent social horrors. Twitter, or the online ad market, are no more "free speech" in an Originalist sense than a contemporary semi-automatic weapon is a musket. In bother cases, we need to understand and apply Constitutional law to find the appropriate balance between collective and individual interests, and what I saying is, we need to do so with an honest discussion of evolving power dynamics given that individuals now have at least hypothetical access to force multipliers. To make this concrete: the 2016 and 2020 elections were subject to asymmetrical information warfare. Asymmetry in the terrorist sense: a small but motivated group, such as the GRU, is capable of provoking catastrophic disruption. I do not posit any particular answers. But we need to have the hard discussion, the sooner the better.