A quick list of sociology of information issues that come up here:
- If datasets contain "information" that could serve the public interest but that might not be extracted by private actors should that data always be made available?
- How should we think about the disincentive to collect data to begin with that will likely emerge if corporations have to always consider the downside of a third party discovering in it information that is contrary to the corporate interest?
- Where do we put the inevitable disputes that will arise over which experts extracted information from data in the correct manner? Does the "more info is always better" stance merely push social uncertainty up one level?
- The article also notes that "Roughly a decade ago, some experts raised concerns that doctors were not getting the full picture about a drug’s risks and benefits because they tended to hear or read about only those trials in which the medication showed a benefit." This raises two issues. First there is the obvious one: "deliberate" bias in terms of what information gets passed on. But there's a second one too -- we need to look at the social networks and channels through which information passes. The more these structures are institutionalized, the more likely, I suspect, there is a built-in bias to what information gets through. Public disclosure might disrupt this just enough to make it a little more likely that the benefits of "marketplace of ideas" and "truth will out" will be realized.