Thank you.
Going back to what you said about transparency, Professor Sandvig, because this is something we've heard a lot about in this committee, if people know where the data is coming from and they understand how the algorithms work, this would allow a certain amount of oversight, as it does in many other areas.
You're suggesting that transparency alone would not actually have that effect. In order to have audits and in order to have a regulator, obviously the information needs to be available, even if you were to audit that information. Are you saying that we need transparency but in such a way that we know who the “who” is in terms of who is actually going to be reviewing? Or is it the public in general—civil society—that would have to do that?