What I Really Think About inBloom

inBloom clearly want everyone's data on their servers, even though they've spent a lot of money to make that not technically necessary. Writing an installable open source server of the complexity of inBloom requires a significant ongoing expense compared to just writing something to run on your own kit. Yet, they don't promote this feature.

They don't promote the fact that running an inBloom server does not require a relationship with inBloom, despite the fact that quite reasonable mistrust of inBloom is threatening to kill the whole non-profit, philanthropically funded project.

That is, their website could say -- tomorrow, accurately, if they wanted -- "Hi, Bill Gates paid us to write a state of the art data integration platform for school districts and states and give it away. You can pay us to host it, host it yourself, pay someone else to do it, and freely switch between those if you're not happy."

Essentially, this would negate 90% of the concern about the project (Bill would still be griping). Up to this point, every indication is they'd rather run the whole thing into the ground than actually publicize the full implications of their own licensing and development strategy. Needless to say, I find this deeply puzzling.

There are two likely explanations for this that I can think of.

The first is that it is simply more convenient for vendors to have a single data hub to deal with. Even with standard protocols, connecting to more individual local inBloom servers would still be more overhead for vendors to deal with than just flipping a few settings on inBloom Central when Pearson or Amplify gets a new district account (and yes, permission from the district). This would be easier for district's too, but I think less of an advantage and with a more obvious cost (parents getting freaked out about their data going to inBloom).

One corollary to this point is that if districts are running their own servers, they might be more likely to decide to sell their district data, or at least try to get a discount in return giving it up to a vendor. "Selling" student data sounds odious, but it might be better than simply giving it away -- it is a valuable commodity generated at taxpayer expense, after all.

The more important point is that I believe they want access to a single, large data-set that they can analyse. People are understandably concerned about their PII -- Personally Identifiable Information -- as they should be, but at the end of the day, your kid's fourth grade IEP isn't a valuable commodity. Its cash value is negligible. What's valuable is all the data in aggregate. You know, BIG DATA. The inBloom privacy & security policy isn't reassuring in this regard, as it focuses (understandably) on PII. Post-Snowden, I don't know how to parse any claims such as "Vendors have no access to student records." Is this like how collecting phone metadata isn't like tapping my phone? Do vendors have access to aggregate reports derived from student records?

Even if it is played tight and strictly down the line at the district level, how much will the often quite-unaccountable state authorities be able to give away? To be sure, this isn't a problem unique to inBloom, but inBloom certainly smooths the process of getting enough state secretaries to sign on the dotted line and flip a few switches on their data dashboard to give Gates or Pearson the keys to the whole kingdom.

On one hand this all boils down to trust, and the justifiable lack thereof. Data has been used as a weapon against my community in the past decade.

On the other hand, let's just close with a reminder that this entire premise is ridiculous. Turning over the private records of citizens, generated in public institutions to a single, national, privately-held database for at best the sake of the convenience of not having to pay a few competent sys admins at the state or local level is an idea that never should have even made it off the whiteboard. You shouldn't need to know any details at all to understand that.