3 ms·

>I don't think the problem with the modern internet is scale so much as a failure to build some constraints into its design. Can you speak more concretely wha

by jasode 7y ago

>I don't think the problem with the modern internet is scale so much as a failure to build some constraints into its design. Can you speak more concretely what you mean here? Are you talking about technical "constraints" into protocols such as "http" or "TCPIP"? Or constraints into DNS? Or constraints on HTML markup language? What would an "internet technical architecture designed to prevent negativity" actually look like? Is there an example repo on github or a computer science research paper showing the algorithms that would satisfy this ideal?

anyonecancode 7y ago
I'm not sure -- in the physical space, a lot of the constraints are social, but they emerged in the context of physical constraints. You could, for example, walk into a restaurant and start haranguing the people there, but generally people don't. Without even raising the possibility of the police being called, there's a whole set of inputs the would-be haranguer can see and respond to -- the facial expressions and other body language of disapproval by the others in the restaurant, the sudden change in noise patterns in the room as all the private conversations ceased and everyone shifted their attention to the disruption, etc. Such social signals are lacking or extremely muted in non-physical settings. Is there a way to bring similar social signals to the online world? Maybe some equivalent will emerge as we grow accustomed to being online (though, given how poorly people drive despite cars being a thing for several generations, I'm skeptical of our ability to fully adapt our social systems to some kinds of technology). Or to take another example of constraints -- if someone in a bar spreads a false rumor, that misinformation can quickly spread to all the patrons in the bar, but it's reach beyond that bar will be slow. By that time more factual information might also be circulating, and the damage of the false information blunted. Online, by contrast, misinformation spreads so much faster than factual information that it is often nearly impossible to counter. I don't know what the solution exactly is here, but I feel that public spaces need to have more speedbumps. In the same way that people are jerks when they drive and the answer is often "less driving, and slower," I suspect that the answer to bad online social spaces is "less online, and slower," but I'm not sure what that looks like.
- ryandrake 7y ago
  > I suspect that the answer to bad online social spaces is "less online, and slower," but I'm not sure what that looks like. Kind of like how HN limits us to a handful of posts per day before the “you’re posting too much” roadblock, causing you to really think about what you have to say and whether you want to burn one of your budget on it. Usenet could have used something like that.
- JoshTriplett 7y ago
  > Is there a way to bring similar social signals to the online world? You're assuming that people will react to those signals by stopping (or not starting in the first place). Many of the most disruptive people look for those signals as their goal, because they want to be disruptive. Think about people who will interpret any variation of "ugh" signals as "oooh, I've found the buttons to poke to get fun noises". If you want to solve this problem, that's a large part of the threat model.
- antepodius 7y ago
  >if someone in a bar spreads a false rumor, that misinformation can quickly spread to all the patrons in the bar, but it's reach beyond that bar will be slow. By that time more factual information might also be circulating, and the damage of the false information blunted [...] I think you're being a bit optimistic about physical spaces here. Rumor mills are as old as the hills. News spreads more slowly IRL, but that goes for truth as well as falsehood. You claim that misinformation spreads faster online than truth and imply this is different from IRL info, but I don't see why this should be the case. The same underlying reasons that favourite one message over another operate in both domains.
anyonecancode 7y ago
Another angle on this is the economic models. Physical social spaces are generally social as a side effect -- it's not how they make money. People go to a restaurant or a bar because they want to be around other people, but bars an restaurants make their money by selling food and drink, not by selling a "social experience." The economic models we've seen so far online are different -- the product is "be social here" and I think that's problematic. Few people want to pay just to hang out, online or off. But those selling this space have to make money somehow -- so if you're Facebook, you make your money by advertising to the people hanging out at your place, meaning to make more money you need to get more people to come be advertised to, or convince advertisers they're getting more value per ad (and so you start intrusively data mining your advertising targets).

by jasode 7y ago

anyonecancode 7y ago

I'm not sure -- in the physical space, a lot of the constraints are social, but they emerged in the context of physical constraints. You could, for example, walk into a restaurant and start haranguing the people there, but generally people don't. Without even raising the possibility of the police being called, there's a whole set of inputs the would-be haranguer can see and respond to -- the facial expressions and other body language of disapproval by the others in the restaurant, the sudden change in noise patterns in the room as all the private conversations ceased and everyone shifted their attention to the disruption, etc. Such social signals are lacking or extremely muted in non-physical settings. Is there a way to bring similar social signals to the online world? Maybe some equivalent will emerge as we grow accustomed to being online (though, given how poorly people drive despite cars being a thing for several generations, I'm skeptical of our ability to fully adapt our social systems to some kinds of technology). Or to take another example of constraints -- if someone in a bar spreads a false rumor, that misinformation can quickly spread to all the patrons in the bar, but it's reach beyond that bar will be slow. By that time more factual information might also be circulating, and the damage of the false information blunted. Online, by contrast, misinformation spreads so much faster than factual information that it is often nearly impossible to counter. I don't know what the solution exactly is here, but I feel that public spaces need to have more speedbumps. In the same way that people are jerks when they drive and the answer is often "less driving, and slower," I suspect that the answer to bad online social spaces is "less online, and slower," but I'm not sure what that looks like.

ryandrake 7y ago
> I suspect that the answer to bad online social spaces is "less online, and slower," but I'm not sure what that looks like. Kind of like how HN limits us to a handful of posts per day before the “you’re posting too much” roadblock, causing you to really think about what you have to say and whether you want to burn one of your budget on it. Usenet could have used something like that.
JoshTriplett 7y ago
> Is there a way to bring similar social signals to the online world? You're assuming that people will react to those signals by stopping (or not starting in the first place). Many of the most disruptive people look for those signals as their goal, because they want to be disruptive. Think about people who will interpret any variation of "ugh" signals as "oooh, I've found the buttons to poke to get fun noises". If you want to solve this problem, that's a large part of the threat model.
antepodius 7y ago
>if someone in a bar spreads a false rumor, that misinformation can quickly spread to all the patrons in the bar, but it's reach beyond that bar will be slow. By that time more factual information might also be circulating, and the damage of the false information blunted [...] I think you're being a bit optimistic about physical spaces here. Rumor mills are as old as the hills. News spreads more slowly IRL, but that goes for truth as well as falsehood. You claim that misinformation spreads faster online than truth and imply this is different from IRL info, but I don't see why this should be the case. The same underlying reasons that favourite one message over another operate in both domains.

anyonecancode 7y ago

Another angle on this is the economic models. Physical social spaces are generally social as a side effect -- it's not how they make money. People go to a restaurant or a bar because they want to be around other people, but bars an restaurants make their money by selling food and drink, not by selling a "social experience." The economic models we've seen so far online are different -- the product is "be social here" and I think that's problematic. Few people want to pay just to hang out, online or off. But those selling this space have to make money somehow -- so if you're Facebook, you make your money by advertising to the people hanging out at your place, meaning to make more money you need to get more people to come be advertised to, or convince advertisers they're getting more value per ad (and so you start intrusively data mining your advertising targets).