4 ms·

Well, I think that you're wrong because you're focusing only on the "census" area - this reflects in my opinion the typical "black or white"-tendencies/extremes

Well, I think that you're wrong because you're focusing only on the "census" area - this reflects in my opinion the typical "black or white"-tendencies/extremes or today's times. I was trying to post a simple example - the event mentioned by the article is not the only occurrence and I guess that city halls had as well records about e.g. votes (which would relate a person to a specific party/movement/way-of-thinking), which would not be related directly to census but which had to be destructed as well. I do not blame the census (and I agree that the pros of census overweight the cons). It's just that the infos that we can release today are not gathered and administered by the national-census-authority and that the their detail go in some areas well beyond what the census would gather.

slg 8y ago
I was speaking about the census because that was your example, but the concept applies to all data. To be as general as possible, people focus too much on the negative value of data without considering the positive value. If the positive value is definite and the negative value is remote, we shouldn't let the existence of that negative value prevent us from receiving the positive value.
- zepearl 8y ago
  (wow, I just had one of those "cooldown-enforced"-moments here on YC that did not provide the "reply"-link...) > people focus too much on the negative value of data without considering the positive value. If the positive value is definite and the negative value is remote, we shouldn't let the existence of that negative value prevent us from receiving the positive value. Mmmhhh, I have personally the opposite feeling (people focus too much on the positive side of data/app and underestimate the negative side, which is how Facebook & Google & others can continue to thrive). > If the positive value is definite and the negative value is remote, we shouldn't let the existence of that negative value prevent us from receiving the positive value. E.g. those poor people in Nigeria definitely got a definite positive value while actively sucking oil from the ruptured pipeline, ignoring negative remote outcome (of blowing up), until it blew up. I cannot ignore remote potential outcomes... .
  - slg 8y ago
    It is an open secret that you can click on a post's age to open a comment and reply immediately to avoid the cool down periods. I am not suggesting that remote potential outcomes should be ignored (which is what probably happened in your two examples). I am suggesting that remote outcomes should be kept in perspective. Let's make numbers up for that second example. Let's say that the pipeline is guaranteed to generate $1 billion over its lifetime but has a .1% chance of failure that causes $10 billion in damage. Some people might view that $10 billion potential damage as disqualifying. However if you do the multiplication the expected negative value is only $10 million. Therefore the expected value of the pipeline is $990 million and the pipeline is probably still a good idea (I admit I am oversimplifying this issue). This is the same reason why a nuclear power plant is generally a better option over a coal plant even though the nuclear plant has the chance of a much greater catastrophic failure. There is also always the question of who receives the positive and negative value. The census example was good because the positive and negative values both impact the general population. The Google/Facebook and Nigeria examples are a little different because the people bearing the brunt of the negative value likely don't benefit as much from the positive value as they should. That distorts incentives in a different way, but that is another debate entirely.

slg 8y ago

I was speaking about the census because that was your example, but the concept applies to all data. To be as general as possible, people focus too much on the negative value of data without considering the positive value. If the positive value is definite and the negative value is remote, we shouldn't let the existence of that negative value prevent us from receiving the positive value.

zepearl 8y ago
(wow, I just had one of those "cooldown-enforced"-moments here on YC that did not provide the "reply"-link...) > people focus too much on the negative value of data without considering the positive value. If the positive value is definite and the negative value is remote, we shouldn't let the existence of that negative value prevent us from receiving the positive value. Mmmhhh, I have personally the opposite feeling (people focus too much on the positive side of data/app and underestimate the negative side, which is how Facebook & Google & others can continue to thrive). > If the positive value is definite and the negative value is remote, we shouldn't let the existence of that negative value prevent us from receiving the positive value. E.g. those poor people in Nigeria definitely got a definite positive value while actively sucking oil from the ruptured pipeline, ignoring negative remote outcome (of blowing up), until it blew up. I cannot ignore remote potential outcomes... .
- slg 8y ago
  It is an open secret that you can click on a post's age to open a comment and reply immediately to avoid the cool down periods. I am not suggesting that remote potential outcomes should be ignored (which is what probably happened in your two examples). I am suggesting that remote outcomes should be kept in perspective. Let's make numbers up for that second example. Let's say that the pipeline is guaranteed to generate $1 billion over its lifetime but has a .1% chance of failure that causes $10 billion in damage. Some people might view that $10 billion potential damage as disqualifying. However if you do the multiplication the expected negative value is only $10 million. Therefore the expected value of the pipeline is $990 million and the pipeline is probably still a good idea (I admit I am oversimplifying this issue). This is the same reason why a nuclear power plant is generally a better option over a coal plant even though the nuclear plant has the chance of a much greater catastrophic failure. There is also always the question of who receives the positive and negative value. The census example was good because the positive and negative values both impact the general population. The Google/Facebook and Nigeria examples are a little different because the people bearing the brunt of the negative value likely don't benefit as much from the positive value as they should. That distorts incentives in a different way, but that is another debate entirely.