3 ms·
I’ve been advocating similar feature before, while I was in and since I left the company and I’m happy that Facebook is trying (because it wasn’t always the cas
by bertil 8y ago
I’ve been advocating similar feature before, while I was in and since I left the company and I’m happy that Facebook is trying (because it wasn’t always the case, even when the problem was clearly identified internally). My version was “the smartest person who disagree with this” (which was essentially my way of having a permanent link to remind me of George Will’s name; can’t get it to stick somehow).
A leak to the WSJ is interesting: I can’t imagine that who opposed what feature could possibly come from anyone not inside (and high up) the company. Even internally, those arguments rarely leave the product review, and that’s at most a third of the executive committee, the product owner, maybe a couple of members of the team and some directors in between. Leaking features before release, or worst, cancelled features, that was extremely frowned upon years ago. With the current tone in the media, I can’t imagine it’s imaginable. Either someone committed career suicide, or it came from higher up. I don’t know how to read the media to understand if that could help Facebook within an “us Facebook vs. Kaplan” narrative.
On the merit of the case, Joel Kaplan probably had a reasonable point: if the feature mostly flags conservative sources and not the NYTimes, it could incense the current pro-Trump movement further — even if one source is factual and the Daily Stormer is… well, far less so. I can’t imagine there were no ways around that, always offering a “correction” that might not contradict one but would clearly challenge far more point in the other, trigger it progressively, trying to understand the impact on far-right and more reasonable people etc. (Facebook recently published paper able to detect and automatically adapt a model to subtle controls like that.) A product like that would not be scrapped, not after the first product review, but asked to tweak some aspects. Either Kaplan has been consistently pushing against any version of it (and that would be mentioned more clearly) or there’s more to it.
- dlo 8y ago> even if one source is factual and the Daily Stormer is… I think it's disingenuous to use the Daily Stormer here as an example. I have definitely come across publications on the left that play loose with the facts or have unlikely explanations for directly observable facts. It would be more fair to use something like PragerU: https://www.bbc.com/news/technology-45247302 https://www.bbc.com/news/technology-45247302
- bertil 8y agoOh, I certainly don’t think it’s a black & white situation. I was just taking the point of view of someone who needs to make a decision from a situation where it’s unbalanced. Actually, pretty much every product Facebook has launched had a well-known unbalanced impact: “get out the vote” efforts get more young people out and they vote more to the left; detailing right to vote drives a subset of potential voters (younger, recent citizens) which isn’t balanced either; same for informing about candidates’ Page, or their program. I suspect that getting a Hans-Rosling/Steven-Pinker-like effort to counter-act spurious arguments with actual statistics would also not be neutral. All are on principle laudable efforts; all don’t have a neutral impact on the vote. I am able to name a right-wing rag because that one is in the press; in general, I stay away from the nastier stuff. As I mentioned in the first paragraph: I care about the smartest of the people who disagree with me and contextualised, nuanced, balanced information. Neither PragerU, the Daily Stormer or other come close to that.