Does anyone else take issue with, or have profound concerns about, platform science?
For clarity, it refers to the tendency the last decade for science to be organized into top down meta level networks (which at times, or more often than not, share a political and economic reality not to dissimilar to market based platform economics).
@koen_hufkens@academicchatter Yeah, I have to agree, I am not sure what you mean by "platform science" or "meta level networks". I have never encountered these terms before.
@DrEvanGowan@academicchatter It's a borrowed (new) term I use. I equate the economic and political fallout of these similar to online "platform" economics of for example Facebook or Uber.
Basically, setups where little new is created, on the platform, aside from a way to aggregate labour - AND - where the benefits of this labour generally don't flow back in equal measure to the contributors (users).
@DrEvanGowan@academicchatter With this I mean that, when done correctly, networks can be great. But this requires radical openness in governance and the underlying process(ing). The latter is rarely the case.
This results in these networks having similar dynamics than say Uber, where the contributions to a network disproportionately benefit the network owner (rarely the contributors).
@koen_hufkens@academicchatter I am still not sure if I fully understand, but in my field (Earth sciences), most work is still very decentralized. Generally speaking, even organized activities (e.g. climate model intercomparisons) are still done in a very ad hoc way, giving large degrees of freedom to each research group on how they approach the problem.
@DrEvanGowan@academicchatter I'll give an example. In movement ecology you have Movebank. Which provides certain services to people who contribute their data (i.e. it pulls in remote sensing and climate data for locations - mostly).
This service is "free", but by and large this service is not reproducible by others (their processing is not open). i.e. you do not teach people to do this on their own. Thereby, sucking the oxygen out of the room when it comes to funding general infrastructure.
@DrEvanGowan@academicchatter Then there is the issue of "god mode" referring to the abuse of Uber spying on its users. And I've seen this, as I had preferential access to data in the past. In short, those who own the network can pre-empt large studies because they can see / and leverage the value of the full dataset before anyone else does.
This gives network owners an incredible advantage in getting things "first out of the door", as well as explore research questions and data gaps.
@DrEvanGowan@academicchatter The latter then reflects onto the potential to write and acquire funding for said network.
This is further compounded by the fact that the more people contribute to large the incentive there is to fund it - especially when the "secret sauce" is not shared. In many ways, things become too big to fail.
@DrEvanGowan@academicchatter This is my general criticism, that networks are fine. Unless they do not share their inner workflows and governance structures. And especially in ecology there is a lot of that out there.
@koen_hufkens@academicchatter I think that in my field this is not really an issue, because there is such little funding in general that no one has the resources consolidate everything into one place like that. For example, there is no centralized database of geological evidence of past sea level variations (something that is very much needed to assess future potential sea level rise) - I am having to compile it myself.
@DrEvanGowan@academicchatter So the way to make this into a platform would be to setup a database, where you gatekeep all aspects (only sharing a subset of the data back to the contributors - or a transformed "product").
You then leverage this into soliciting more contributions (the growth phase) from external collaborators. This will give you the opportunity to posit that it becomes indispensable (while not sharing all data or the underlying processing) - increasing personal funding chances.
@DrEvanGowan@academicchatter Now I've been the devil's advocate here. Don't do this, as it limits reproducibility and in the long-term negatively impacts the open data discourse IMO.
@koen_hufkens@academicchatter Haha, I haven't even been able to convince anyone to give me a job to do this, let alone make it something that could be monetized. Even though my database is fully open on Github and described in a highly cited paper that I published 3 years ago, I do not know of anyone else using it.
@DrEvanGowan@academicchatter Often monetizing this require a certain momentum (critical mass), and being open about things works against this. Again, not saying you should do this - but it seems to be a pattern in certain fields.
Add comment