HNHacker News
TopNewBestAskShowJobs

TrailMixRaisin

114 karma · joined July 25, 2019

submissionscomments
TrailMixRaisin··on Hospitals are selling troves of medical data
I am sorry, but your answer is nearly the exact textbook example how wrong most people are with their intuition. In 2000 Latanya Sweeney showed that 87% of Americans can be identified by Sex, Birthday and Zip-Code. As other commenters point out, the connection to external databases is extremely powerful and dangerous to privacy.

There are other concept in research to still use this data in research like differential privacy or using KI to synthesize data according to training data provided. But so far all concepts that tried to alter and then publish a datasets directly for research failed miserably.

TrailMixRaisin··on Hospitals are selling troves of medical data
There is a (famous) quote from Cynthia Dwork, who is most likely the key researcher behind differential privacy: "De-identified data isn’t".

You can either re-identify the people behind the data or you alter it that strong that it becomes useless for meaningful applications.

TrailMixRaisin··on How to Criticize Computer Scientists (2001)
In my understanding this comes from the fact that many ML is more engineering than science but wants to be seen (or is reviewed as) theoretical research. So it fits the situation in the article even better.
TrailMixRaisin··on Ketchup is a non-Newtonian fluid
There is a paper where the authors use ketchup for polishing high precision lenses. The especially want to have the property of the shear-thinning for best results: https://www.spiedigitallibrary.org/conference-proceedings-of...

To be clear, there are better polishing agents in the scenarion than ketchup. On of the authors mentions that they wanted to do some research that could get them the IG Nobelprice.

← PreviousPage 2 of 2