In Brussels to speak about a mix of things I know very well (platform liability laws, filters, global takedown orders, fundamental rights, TOS issues) and things I am trying to figure out (the path to meaningful transparency, is there such a thing as smart ranking regulation). 1/
I am probably more useful to other people when I talk about that first batch. But I’m more excited about the second. 2/
On transparency: We operate in the dark re content moderation as long as researchers & policymakers can only guess/rely on anecdata about *actual* content removal decisions. (@PJLeerssen and I have a pending article on sources of real data, and what we should want) 3/
@PJLeerssen But the barriers, including GDPR barriers, to sharing anything but aggregate data are real. Particularly where the content was removed because platforms thought it violated a law. The real work to navigate this is only beginning. 4/
@PJLeerssen On ranking: It seems to me that
1. Laws regulating platforms’ algorithmic ranking, curation, and recommendations are coming.
2. We very far from knowing what a smart version of those laws would look like.
3. Half-baked laws will get passed anyway. 5/
1. Laws regulating platforms’ algorithmic ranking, curation, and recommendations are coming.
2. We very far from knowing what a smart version of those laws would look like.
3. Half-baked laws will get passed anyway. 5/
@PJLeerssen In particular, as @PJLeerssen & others discuss here, there is tension between asking platforms to surface *diverse* content (as in “media diversity”) and asking them to surface *quality* content as defined by e.g. the Code of Practice on Disinformation. blogs.lse.ac.uk/medialse/2019/… 6/
@PJLeerssen (BTW - you should read that article not only because it is smart, but because it describes a remarkable, and under- discussed, pending German law that regulates ranking.) 7/
@PJLeerssen Is “fair” ranking defined by viewpoint diversity? Speaker diversity (national background, gender, etc.)? Competition? Any of these framings may yield different results. 8/
@PJLeerssen To what extent should these goals override giving users what they “want” (or appear to want based on clicks and behavior), in order to give them what we think they should want? I think of this as the “eat your kale” problem? 9/
@PJLeerssen I think this is a deep question about when the law shifts from wanting a private company to respond to market forces to wanting that company to serve other, more state-like goals. People with telecoms/media reg backgrounds have a lot of insight (and disagreement!) on this. 10/
@PJLeerssen If ranking regulation is coming, is starting with a half-baked law that fails to answer these questions a good thing, letting us iterate to better laws? Or a guarantee of expensive legal chaos and uncertainty, leading to platform ranking that still pleases no one? 11/
@PJLeerssen See, like I said, I’m more excited about the questions I can’t answer yet :) But I am looking forward to talking about my old standby topics (hello, Glawischnig-Piesczek ruling!) too. 12/
