Originally published at: https://boingboing.net/2019/06/20/sum-count-average.html
…
2 Likes
Please don’t fuck this up Google, there is so many useful conclusions to be extracted out of the shit load of data we have but privacy has to be top of the concerns.
6 Likes
I look forward to Google neglecting this project for 36 months after it reaches maturity, then abruptly canceling it, leaving hundreds of applications that rely on it in the lurch.
6 Likes
Either that or more likely what they’ve done with Chrome. Reap the benefits of their supposed open source project but locking down its use so only they can truly benefit.
2 Likes
My concerns:
-
Why does the city have identifiable data about train trips in the first place? That’s starting from a problematic position.
-
If both sides can keep their data secret, then it seems it would be hard/impossible to verify that the other party is providing actual, realistic data. If the payment provider has an interest in getting the extra train service approved (or not), they can bring a fudged dataset to the party and there’s no way for the city to audit it. (Maybe there’s math here to solve the problem.)
-
With all the data remaining secret, it’s not science. If others can’t examine the data and try to reproduce your experiment or analysis independently, then we shouldn’t be making public policy decisions based on the results. If this were two commercial organizations trying to figure something out, well, then maybe it’s OK. But a municipal transit agency? Nope. Not if their constituency cannot independently verify the analysis.
-
I can’t think of many cases where both parties would find the result valuable. In the common case, I’d expect the party that cares about the result would have to offer an incentive to the other party to get them to participate. So, although the computation is symmetric, the relationship is not. This amplifies the implicit trust issue.
4 Likes
3 Likes
This topic was automatically closed after 5 days. New replies are no longer allowed.
