Database Reference
In-Depth Information
what he has done over the last twenty years of his career. So when doctors
see a new patient, they are harnessing their experience and what cases they
have read—maybe even some from a medical journal. What would be better
is to be able to combine the particular instance of a person, their condition,
medication, and symptoms in a privacy-protecting way, to then be fed into a
system. Then we could have a system tell the doctor, “We have seen this about
15 times. If you do treatment X, the guy will do really poorly, so that is not
really what you want to do. However, if you do treatment Y, the guy will get
better.”
So being able to have the collective experience of medical conditions, symp-
toms, diagnosis, medications, treatments, and results would be tremendous.
After all, that is really what data is—a collective experience of what happened
being brought together. This is important because clinical trials are problem-
atic for two reasons: (a) they never have real-life environmental settings where
people take all of the stuff, because people are screened very clearly; and
(b) the other part is that I do have my concerns, and they have been voiced
repeatedly, about how reliable some clinical trials are. I want to leave it at
that.
I think that from observational data, we could learn so much more as to what
are effective solutions and what are ineffective solutions. Along with more
observational data, we also need better forms of data regulation. Regulating
the data and privacy implications have to be dealt with. I think this is true for
advertising and privacy in general. Not just in medical applications and social
data people share on Facebook, which are used for all kinds of purposes.
I think the medical application is the best case where you can make a state-
ment of, “Here is what should really be done.” We need to start being able to
compare the benefits to the costs without sensationalism.
I get upset when I hear of singleton sensational examples like the Target case
where the dad found out his daughter was pregnant. Dads get to know that
their precious daughters are pregnant every single day, through all kinds of
means. Whether it is the daughter's friend who tells them, or a nosy neighbor
who lets it slip out, or something else—these things happen all the time. That
is part of life. I do not want to necessarily push this as a call for more targeting.
Rather I think we have to confront the fact that once in a while, privacy might
be violated. After all, it happens in real life every single day.
So I get upset at the boulevard newspaper atmosphere that exploits this situ-
ation to instigate a huge outcry. The fact is that you cannot make privacy
airtight. That cannot be the goal. We need to have a really meaningful con-
versation of what the tradeoffs are. Sadly, I do not really see that occurring in
the near future because often privacy advocates sit on one side with a lot of
complaints and no proposal as to how to do it better, while on the other side,
the general attitude is more along the lines: “The sky is the limit. Let's just do
whatever we want.”
