Hacker News new | ask | show | jobs
by candiodari 2809 days ago
Meaning that keeping data away from people who would use it for their own purposes, from selling you something to impounding your car if you're behind on payments cannot realistically be accompolished.

For each use, judge for yourself if that's good or bad, effective or ineffective, but the purpose doesn't change whether the data is available or not.
1 comments
Individualist 2809 days ago
Bad data is bad data. It's true it gets averaged out due to the masses. Still it does provide a somewhat accurate picture, but has a degree of error.

Best means I can argue this are colleagues in plant life and marine biology using the same tools are still attempting to get concrete predictive data for the environment. It's simply not there yet.

link
candiodari 2809 days ago
Nope. What I'm saying is that YOU're average (and me, and everyone). Which means that given a few facts about you, the rest can be correctly inferred. Which means that you cannot prevent systems from inferring how you can, say, be induced to buy something, because you cannot keep those few trivial facts it needs for that from the system.
link
Individualist 2809 days ago
I keep getting ads for things I do not want to buy due to the average. For example a championship belt that recently has a tacked on women in my feed. By no means is that something that interests me but can understand the average may be tempted.

In regards to the younger generations I definitely understand the fear. But we get bored as a species, otherwise networks would never of gone down the outrage path of targeted ads had an absolute effect. And even then more and more people are setting aside their smart phones and jumping off of social media or just using it for the bare minimum.

What you are arguing is happening, but whether or not we are truly on the last leg of the data argument is not true. https://en.m.wikipedia.org/wiki/Carpenter_v._United_States

link
candiodari 2809 days ago
That just means the algorithms don't work well enough or don't have enough data yet. Also there's a little bit of bad incentives here: these companies also get paid for showing you something at all.
link