Smartphone Location Data Is Worth Big Money to Wall Street | Hacker News Reader

Smartphone Location Data Is Worth Big Money to Wall Street

wsj.com

54 pointsSREinSF7 comments

This is such a dark use of tech. Involuntary using people's phones and fta it says data goes mostly to hedge funds. Then the kicker was at the bottom:

"In September, as Hurricane Florence churned toward the Carolina coast Thasos watched evacuation zones and found that in well-to-do census blocks, 65% of the people fled, while only 39% left poor areas. Such information could inform disaster response or infrastructure spending, yet it might also have commercial value. “You might look at that and say, ‘Gosh, I could price insurance differently,’ ” Mr. Pentland said."

So they're not even interested in savings lives so much as pricing insurance. Fantastic.

Err isn't there social benefit to having better priced insurance? I don't understand your position here.

Assuming that insurance is already priced as accurately as it can be according to the available data, then discovering that wealthy people tend to evacuate disaster areas while the poorer residents remain would suggest that you charge greater premiums for the impoverished and give a price break to the rich. As it stands, the poor are effectively getting subsidized by otherwise similar individuals with more wealth.

Insofar as the utility associated with money is roughly logarithmic, then the net utility from a pricing change will likely decrease. For example, a wealthy person will be pleased with a $500 reduction in premiums (that's a new pair of shoes or some decent concert tickets); a $500 increase might induce a poor person to forgo renewing their life insurance (with disastrous effects for their dependents should they die) or if they pay it, then other worthy expenses might get cut to make up the difference (home repairs, sending your kids to summer camp, retirement savings).

Even assuming that everyone values money equivalently despite their net worth, it still leaves a bad taste in my mouth. Here's one wealthy elite technocrat, already invading people's privacy on an Orwellian scale, and he's using that access to explicitly prey on poor people to make money for other people who are already rich.

Or, maybe you can charge both less because now you can replace the flees-disasters-rate conservative guess with less conservative data. And make a more competitive product. And reduce costs to end consumers. Not everything has to be sinister when someone breaks out that different groups of people behave differently.

In the article it mentions that the data comes from apps on the phone, so insofar as the user has clicked through the EULA, Privacy Policy, and permissions screens[0], it's voluntary in theory but obviously not in practice. The interesting thing is that the data source comes from "hundreds" of innocuous apps (weather, directions to ATMs, social media add-ons) that send the user's location data to a company that aggregates and then sells it.

This in turn is only possible because for useful but easily coded apps, you'll always have a race to the bottom on price, and once you hit free then the only way to make a return on your investment is by taking advantage of the user, by ads or by tracking. And it will be this way for the forseeable future, or at least until the harvested data indicates that there's uncaptured profit in making apps that aren't free, but promise to respect your privacy.

-----

0. It sure would be nice if someone could use natural language processing to parse the legalese and spit out a summary of what you're really agreeing to. Maybe combined with some sort of phone-farm (like they use for testing builds on different devices) to download various apps in order to acquire the corpus and maybe test the what servers the apps communicate with, and how often. I know that there are sites dedicated to interpreting EULAs from major companies/apps, but probably the greater threat is the myriad free utility apps that get used once and then forgotten.

Bypass paywall:

* DDG: https://duckduckgo.com/?q=Your+Smartphone%E2%80%99s+Location...

* Google: https://www.google.com/search?q=Your+Smartphone%27s+Location...

If Google link fails... https://outline.com/yPWTZR

Reply on news.ycombinator.com