5 ms·

> I hate it that simplistic naratives have won in public discourse, and we're stuck with overuse of analogies. At least they have a "simplistic narrative". You

> I hate it that simplistic naratives have won in public discourse, and we're stuck with overuse of analogies. At least they have a "simplistic narrative". You just vaguely gesture towards what you think should happen without any reasoning or argument to support your point. > Data is not like anything. Every model is wrong, some are useful.

mihaic 2y ago
> Every model is wrong, some are useful. Yes, and I'm saying this one isn't useful, it's worse than no model at all. I can still be correct in pointing out that something is bad even if I don't have something better.
- amelius 2y ago
  For you the model may be bad. But for 90% of US population it may be useful. Or, it may be useful to some if 90% of the US population believed the model.
- josefritzishere 2y ago
  Is that quote from George Box or does it go back further? I've always liked it.
- krisoft 2y ago
  > Yes, and I'm saying this one isn't useful I hear you say it, but you provide still no reasoning. Not even a handwave. Do you expect us to intuit your thinking? Why is it not usefull? What do you find disagreable with the article’s reasoning? > I can still be correct in pointing out that something is bad even if I don't have something better. You could be. You are not convincing. The article is. So who is winning here? “Simplistic narrative” vs “believe it because mihaic says so”?
  - mihaic 2y ago
    > You are not convincing. The article is. So who is winning here? “Simplistic narrative” vs “believe it because mihaic says so”? Yeah, and I was just expressing my opinion on disliking that simple narratives are winning. If you were honestly trying to have a conversation, and were genuinely curios of my arguments, I'd put in the effort to explain them, like hakunin for instance posted slightly above. I could for instance talk about how I feel like the overuse of analogies is dumbing the audience, like how movie scenes of "hacking" get removed from reality, but the medium develops a language of its own. Data is also hard to explain properly since a difference in scale change the nature of what abuses you can make with it. I could here introduce the analogy to the problems of solving a random crime (a data point) versus organized crime, that reached a new dimension simply from scale. But I would also make a point of where this analogy breaks down -- the article merely mentions it's not a perfect analogy and then just continues. Again, I'm not trying to convince anyone about what I think about data, but rather of more responsibility in properly communicating concepts, and correcting when someone oversimplified concepts.
    - krisoft 2y ago
      > If you were honestly trying to have a conversation, and were genuinely curios of my arguments I am actually. > I'd put in the effort to explain them Thank you! > I could for instance talk about how I feel like the overuse of analogies is dumbing the audience, like how movie scenes of "hacking" get removed from reality, but the medium develops a language of its own. Now that is an interesting example. The thing is a real hacker doesn't look like much. Visually they can be anybody. They probably have a computer. They probably type in commands (but not always) to a command line. They probably type into a text editor and save it. Visually me sitting in this chair and writing this comment to you, and a master hacker infiltrating a secure network are almost indistinguishable. We know that the hacking involves mental puzzle solving, subterfuge, probing and discovering systems, the danger of being found, dealing with active and passive defences etc. So they illustrate these elements visually. Because from the storytelling perspective what matters is not how it is done (person typing in commands into the command line) but what action do they represent. The emotional content. How would you want it handled? Why would that be better? In what sense would it be better? (Would it be better story telling? Or would it be better for society at large?) I tell you a bigger problem with movie language: corrupt police officers. Certain kind of crime genre requires that the police is either corrupt or incompetent. Because if they were neither the whole premise of the movie would be solved by the hero walking to the police and making a report and the baddies would be caught, end of story. But that is bad story telling. It is not interesting to watch. So in these kind of movies corrupt and/or incompetent police is a requirement. Which over represents police corruption/incompetence in crime movies which has real-world implications. It slowly corrodes society where people expect less from police, less surprised when corruption or incompetence is reported, and makes people less likely to turn to the police with their problems. That's very bad. I'm not sure how stylised "hacker battles" are bad in the same sense, but maybe you thought more about that than me? > correcting when someone oversimplified concepts. But what is the oversimplification? We exchanged multiple comments over this and I still don't understand your point. The article doesn't say that "private data" is literally "human organs". It is not recommending that you should do an immunological compatibility test before you hear someone's private data, or that you should transport private data in a chilled container. That would be clear nonsense. Instead it recommends that legislature could ban the trade of private data. Like how we ban the trade of human organs. That is what the article says. Which of this is "oversimplified"?

mihaic 2y ago

> Every model is wrong, some are useful. Yes, and I'm saying this one isn't useful, it's worse than no model at all. I can still be correct in pointing out that something is bad even if I don't have something better.

amelius 2y ago
For you the model may be bad. But for 90% of US population it may be useful. Or, it may be useful to some if 90% of the US population believed the model.
josefritzishere 2y ago
Is that quote from George Box or does it go back further? I've always liked it.
krisoft 2y ago
> Yes, and I'm saying this one isn't useful I hear you say it, but you provide still no reasoning. Not even a handwave. Do you expect us to intuit your thinking? Why is it not usefull? What do you find disagreable with the article’s reasoning? > I can still be correct in pointing out that something is bad even if I don't have something better. You could be. You are not convincing. The article is. So who is winning here? “Simplistic narrative” vs “believe it because mihaic says so”?
- mihaic 2y ago
  > You are not convincing. The article is. So who is winning here? “Simplistic narrative” vs “believe it because mihaic says so”? Yeah, and I was just expressing my opinion on disliking that simple narratives are winning. If you were honestly trying to have a conversation, and were genuinely curios of my arguments, I'd put in the effort to explain them, like hakunin for instance posted slightly above. I could for instance talk about how I feel like the overuse of analogies is dumbing the audience, like how movie scenes of "hacking" get removed from reality, but the medium develops a language of its own. Data is also hard to explain properly since a difference in scale change the nature of what abuses you can make with it. I could here introduce the analogy to the problems of solving a random crime (a data point) versus organized crime, that reached a new dimension simply from scale. But I would also make a point of where this analogy breaks down -- the article merely mentions it's not a perfect analogy and then just continues. Again, I'm not trying to convince anyone about what I think about data, but rather of more responsibility in properly communicating concepts, and correcting when someone oversimplified concepts.
  - krisoft 2y ago
    > If you were honestly trying to have a conversation, and were genuinely curios of my arguments I am actually. > I'd put in the effort to explain them Thank you! > I could for instance talk about how I feel like the overuse of analogies is dumbing the audience, like how movie scenes of "hacking" get removed from reality, but the medium develops a language of its own. Now that is an interesting example. The thing is a real hacker doesn't look like much. Visually they can be anybody. They probably have a computer. They probably type in commands (but not always) to a command line. They probably type into a text editor and save it. Visually me sitting in this chair and writing this comment to you, and a master hacker infiltrating a secure network are almost indistinguishable. We know that the hacking involves mental puzzle solving, subterfuge, probing and discovering systems, the danger of being found, dealing with active and passive defences etc. So they illustrate these elements visually. Because from the storytelling perspective what matters is not how it is done (person typing in commands into the command line) but what action do they represent. The emotional content. How would you want it handled? Why would that be better? In what sense would it be better? (Would it be better story telling? Or would it be better for society at large?) I tell you a bigger problem with movie language: corrupt police officers. Certain kind of crime genre requires that the police is either corrupt or incompetent. Because if they were neither the whole premise of the movie would be solved by the hero walking to the police and making a report and the baddies would be caught, end of story. But that is bad story telling. It is not interesting to watch. So in these kind of movies corrupt and/or incompetent police is a requirement. Which over represents police corruption/incompetence in crime movies which has real-world implications. It slowly corrodes society where people expect less from police, less surprised when corruption or incompetence is reported, and makes people less likely to turn to the police with their problems. That's very bad. I'm not sure how stylised "hacker battles" are bad in the same sense, but maybe you thought more about that than me? > correcting when someone oversimplified concepts. But what is the oversimplification? We exchanged multiple comments over this and I still don't understand your point. The article doesn't say that "private data" is literally "human organs". It is not recommending that you should do an immunological compatibility test before you hear someone's private data, or that you should transport private data in a chilled container. That would be clear nonsense. Instead it recommends that legislature could ban the trade of private data. Like how we ban the trade of human organs. That is what the article says. Which of this is "oversimplified"?