Every week we like to highlight some of the most insightful comments from our readers. And just like last week, Switchers had a lot to say about the FDA and 23andMe -- especially after Brian Fung reported that 23andMe would stop issuing interpretative health-related genetic reports, but continue to provide ancestry information and raw health information. Switch commenter micron26 explains why he is on board with that move:
To call this the right decision is an understatement.
Anyone who basing medical outcomes on most of the genetic data uncovered by the antiquated 23andMe SNP array was making decisions based on interpretations we now know to be erroneous in many cases.
Science moves on. 23andme did not...and had the hubris to pretend that their product was providing something of value in cases where it can easily be proven that it was not.
For those who view this as governmental overreach, whose job is to take action when a company is making false claims that could directly impact health decisions? Also, keep in mind that you can still buy the product and get all the genealogical and raw data you want. The only real change is that 23andMe can't continue to make unvalidated medical claims.
But not everyone agreed about that, with user GFRGFRGFR arguing "23andMe was providing INFORMATION, not therapy. As in all things the end user is responsible for how that information is used (or misused)." Micron26 countered:
There's nothing wrong with 23andme giving this type of information: individual A has a rare heterozygous SNV (call it nucleotide Z) at position B on chromosome C, and giving that result a quality score.
There is something wrong with 23andme making the claim that nucleotide Z is associated with a predisposition to condition D based on a 6-year GWAS publication result (from some random authors in a third-tier genetics journal) that has never been replicated.
Microsoft declared the U.S. government an "advanced persistent threat" this week based on the revelations about NSA activity over the past few months. Reader WhamboMPS2 thinks that designation "spot on," saying:
I have no issue with the government engaging in espionage. It's unseemly and unpleasant but is a part of the interactions between nation-states that goes back millennia. However, much of the NSA's actions have swept up the data of innocent Americans along the way. My data -- YOUR data. All with minimal oversight by a secretive court that acts outside the reach or view of the Americans that are supposedly being protected.
The 4th Amendment is pretty clear: "The right of the people to be secure in their... papers, and effects... shall not be violated, and no Warrants shall issue, but upon probable cause..." There is (I believe!) no probable cause for the government to believe that my "papers and effects" are worthy of being violated. The people around me who use the services of Microsoft (or Apple, or Google, or Yahoo, or...) may fail the same test. And the government has the right (and even the duty) to go after that data. But it is not reasonable that in gathering the data of the "evil-doers" that the government should scoop up literally billions of records of Americans simply going about their daily lives.
But reader OscarMayer2 believes there might be some ulterior motives to Microsoft's designation, saying:
Microsoft, like Yahoo, Google and others have understood the real threat to their overseas businesses and loss of overseas markets from the recent Snowden revelations" and that the tech sector's
The Tech Sector's recent indignation and criticism of Washington are primarily to reassure foreign customers that the tech sector is not an extension of US surveillance sector. Time will tell how well the tech sector succeeds in walking away from current perceptions of collusion, real or imagined.
When the Manhattan Project concluded, Richard Feynman sat down and made a list of all the things you could do with atomic power - which was really just a list of things that needed power plants, plus the statement "you could use an atomic power plant for that." He wound up the holder of a patent on atomic airplanes.
Many software patents are like that - they take a known "something" and do it with a computer instead. That may once have been novel, but surely it no longer is. Note though that distinction is about novelty, not about the nature of software as math or not. There are things that can only be done with computers, which are being thought of for the first time, and which create genuinely useful outcomes and breakthroughs.