"playing on a seashore"
"Both the victor and the vanquished are but drops of dew, but bolts of lightning —thus should we view the world." -Ouchi Yoshitaka (1507/1551).
|20141027 151434 Z||Isaac Asimov Asks, “How Do People Get New Ideas?”||www.technolo … e-get-new-ideas/||Cool, Inspiring, Mind, Psychology, Reading, Text||"On Creativity" (1959) by Isaac Asimov, just published. A small group of creative, relaxed, jovial folks having a cerebration session sounds like a lot of fun!|
|20141021 182145 Z||Hendo Hoverboards - World's first REAL hoverboard||/www.kicksta … -real-hoverboard||Cool, Gadget, Physics, TECH||Hendo Hoverboards should be all over the place in a few years. Go, go Lenz's Law!|
|20141010 152809 Z||The Quest to Build an Elevator to Space||gizmodo.com/ … space-1638192427||Space Elevator, TECH||Whoot! Companies that are trying to build space elevators, like Obayashi and LiftPort, have been getting more coverage lately.|
|20141010 152038 Z||Batteries Included: A Solar Cell that Stores its Own Power||news.osu.edu … s-its-own-power/||Solar, Sustainability, TECH||The combination solar cell and battery by Yiying Wu at Ohio State University is one of the more exciting developments in solar tech.|
|20141003 153202 Z||New 'Diamond nanothreads' may make space elevator a reality||www.cnet.com … vator-a-reality/||Chemistry, Nanotechnology, Space Elevator, TECH||The new diamond nanothreads may be stronger than carbon nanotubes. Possibly another step closer to space elevators.|
|20140812 145734 Z||RIP Robin Williams||Celebrity, News, Obituaries, Psychology, Show Biz||RIP Robin Williams (Chicago 1951-07-21/2014-08-11). For your wonderful work --making us laugh and cry-- I thank you. For your advocacy for those with depression, nihilism, or pain --even with your death-- I thank you.|
My family and I have thoroughly enjoyed watching "Cosmos: A Spacetime Odyssey", the recently concluded 13 episode science documentary TV series. A sincere thank you to Neil deGrasse Tyson, Seth MacFarlane, Ann Druyan, Fox, the National Geographic Channel, and the many who worked on Cosmos.
The show was thoroughly aligned with content my wife and I encourage for our kids. A sense of curiosity and wonder. Openness to different ideas and people. Awareness of the danger and responsibility of climate change: A climate change of a few degrees (or a few ppb of CO2) leads to an Ice Age or a Heat Age. The beauty and explanatory power of evolution. The importance of epistemology and the scientific method, and how it can stray.
I'll conclude with five simple rules that Neil mentioned in the last episode.
We've been distracted by the stunning advances that information technology has brought us, to the exclusion of very deeply held needs that we have in society.
In college I could barely do chemical engineering because I was so distracted by computers! The Web wasn't even up then, so people have even more distractions now.
Cosmos has, as its mission statement, the effort to convey to you why science matters. That is a different motivating factor than "Here's all this science I want to teach you."
There is so much mind in science that people forget that it has so much heart in it too.
Last night I attended the inaugral meeting [http://www.meetup.com/Health-Data-Liberation/events/156222932/] of the Meetup group "Health Data Liberation" [http://www.meetup.com/Health-Data-Liberation/]. The meeting was run by Dr Rebecca Wurtz, an Epidemiologist at Northwestern. 2-3 dozen attended including doctors, folks in BI & analytics, several in the open data movement, a fellow from payers, etc. There was a vigorous and enthusiastic discussion, that was not dominated by any one individual.
There was much discussed, and I just want to hit on a few key points.
Issue: Trust. Different systems already have our data. But who do we trust to merge data? How can we trust them? The data merging entity would have to promise to be the end-of-line as far as identified data, if they want to share the merged data, then they could only share de-identified data. The entity would have to have strong security and impoorting arrangements. Each person would have the right to their own data and to pull out at any time.
Issue: De-identifying data. There was much involved in suspending disbelief that we could actually de-identify data --especially once a data set has been merged with other data sets. The HIPAA standard for de-identification (45 CFR 164.514) does not stop people from using their Sherlock skills to identify people in merged data sets without the name, gender, address, date of birth, etc.
Issue: Name. It also became apparent that since we want the data to be as wide and complete as possible. The term "health data" is too restrictive. Perhaps "life data"?
Issue: Falsified data. We may like the concept of the quantified self, but did you really do 1000 push ups in 30 minutes today? It seems that we need external parties to provide some degree of validation of the data.
Issue: Select which data sets to share. People may be OK sharing most health data but not genomic data or book purchases. The People should be able to choose which data set (hospital, pharmacy, payers, labs, medical & exercise devices, Google, Amazon, genome, banks, utilities, etc) they want merged into the open data set. They should be able to do this in batches or stream, scheduled or one-offs.
Issue: Sharing data. I'm guessing that this will probably go the Semantic Web route with RDF, but in the meantime we have to deal with HL7 v2, HL7 v3, HL7 FHIR, EDI 837s, XDS, etc, etc. And that's not even talking about the non-health data sources.
Issue: Analysis. Open life data is in the twilight zone between collecting data and analyzing it. While systems already collect and analyze internal & proprietary data, no one is up to sharing the data for de-identification and open analysis yet. We didn't discuss the kinds of open analysis that could be done, but perhaps that would need to be a separate discussion because the other issues overshadow this one.
Issue: Starting. My guess is we'd have to start with small set of people who volunteer to make their life data open (but de-identified!) at some site X. Then they can choose which data sets to share their identified data with X. Site X would merge different identified data set but only share a de-identified data set to anyone. If successful, then there will probably be ways to incentivize people to participate in the data liberation.
Open data, open source. For the tldnr crowd: