4 ms·
The Data Delusion
- yawnxyz 4y agohttps://archive.is/5j1RE https://archive.is/5j1RE
- NKosmatos 4y agoNice article and very informative, but I don’t think it’s the data we should be afraid of, but the usage/interpretation/analysis of them for ill-intentioned purposes.
- tgv 4y agoI haven't read the article, but while your answer may be technically right in some sense, it ignores the issue. Shouldn't we be afraid of guns? Cigarettes? Atomic weapons? Data provides the reputation to an analysis or interpretation that it is grounded enough to be considered correct. If the data is bad or doesn't warrant the interpretation, nobody will know. It stays hidden with the data.
- asteroidz 4y agoThe usual caution applies here. The danger of data isn't in the data itself, but in how it's used and interpreted. In this particular case, a relatable comparison would be a book - it's just a bunch of words on a page, but those words can be used to inspire or to manipulate. Pretty important to be mindful of the context and the intent behind the data. Something important I was thinking of the other day is the power dynamics that come into play with data as the new currency. Those with access to more data (money) and better tools for analysis (investment) of it end up having an unfair advantage.
- Jgrubb 4y agoYou're right about resources lending an unfair advantage of course - when haven't they? - but the real danger to me isn't that those with greater access to resources will have the capability to produce better analyses, it's that they'll have the capability to better push their agenda. I remember watching Colin Powell on TV in 2003 explaining to congress, "You see these trucks here? These are clearly carrying weapons of mass destruction. Only bad trucks would leave tire marks and move around like this." Clearly this was a bad read, and the whole of humanity has suffered for this agenda being sold to us as an irrefutable, data-backed truth.
- jongjong 4y agoData is especially dangerous when it comes to economics because financial ideas have a self-fulfilling tendency. For example, if a lot of investors are using the same data and same system/approach to figure out which investments they should make this year, they may end up being led towards making the same investments and this would create a lot of demand which would increase the price of the investment throughout the year... It doesn't matter if the investment was actually horrible, it's the shared belief that it was good which made it grow in value in the short or medium term. When certain people make important decisions on a global scale, it doesn't matter what the data says. The decisions may be so impactful that they will make all historical data worthless as that data was only ever relevant within a specific environment. When you fundamentally change the environment, you invalidate the data which was gathered in that environment. Decisions on a global scale tend to change the environment in fundamental ways. Common sense is the most important factor, not data.
- walkhour 4y agoMarket madness is one the best if not the best example of self fulfilling prophecy there is. But there are many use cases of data that don't result in anything like that, e.g. the example at the end of the article: car insurances.
- disgruntledphd2 4y agoReally, all car insurance models is propensity to claim. By definition, this will be lower for richer individuals, leading to poor social outcomes.
- walkhour 4y agoSure, that's quite a leap though, that higher car insurances for poor people are a cause of poverty. Surely they contribute something, but how much? Is it 0.001%? Is it whatever they contribute to big enough to not get lost in noise? You would need a prove for the statement that higher car insurances are themselves a cause of poverty. Maybe higher car insurances make poor people less likely to get a car, and they end up richer because they don't waste time in useless holidays that would've made them poorer. I'm not claiming this is what happens, just that your logic even if it makes sense in some abstract sense, may be separated from reality.
- schnitzelstoat 4y agoOne of the books they mention looks pretty cool - The Declassification Engine by Matthew Connelly. Has anyone read it?
- fissure 4y ago[flagged]
- rektide 4y agoI really enjoyed Cory Doctorow's outlook & his critique of Shoshanna Zuboff's Age of Survelliance Capitalism, which is largely that these are not near magical superpowered systems which can with incredible precision reprogram human behavior & reshape the world. These companies want to convince us of their superhuman insight & abilities, and rather than sustain a skeptical counter-outlook, the prevailing counter-narrative is to drastically over-subscribe their hype & to radically over-project their capabilities, is freaking the frell out & panicking about it. https://www.cigionline.org/articles/doctorow-versus-zuboff/ https://www.cigionline.org/articles/doctorow-versus-zuboff/