...and it sounds like they initially suppressed and injected a lot more stories, too, so it's conceivable that this person was pissed off about these suppressio

...and it sounds like they initially suppressed and injected a lot more stories, too, so it's conceivable that this person was pissed off about these suppressions, but not about others that happened to agree with their own prejudices. And then there's the issue that's these kind of algorithms need tuning - they're not intrinsically neutral. It's unlikely that conservative and liberal social networks/stories have the same kind of structures, and details of how the network is connected can influence what's trending. This makes the complaint a little absurd. Not only are algorithmic results also biased, these manual corrections might conceivably be useful to correct that bias. There's this naive assumption that there exists the One True Fair and Neutral ranking algorithm that's unbiased. But the whole notion is ridiculous: we want that bias, because we want to discriminate in favor of "trending" content - and "trending" isn't even rigorously defined. And even if we had a rigorous definition, that doesn't mean it'd be the most useful definition. Elections have strictly defined protocols, yet that doesn't mean an elected US president has the majority of votes. And lest you believe majority voting is the answer, a majority elected president might nevertheless be inferior to a competitor by majority opinion (i.e. it's not stable)... and of course regardless of election result there's no guarantee that current opinions are a good predictor of future opinions let alone achieved results. In short: not only is this just one person's opinion, the standard they hold Facebook to (no bias) is impossibly high. Perhaps there's some amount of unwanted (by whom) bias, but dealing with that isn't as black and white as the complainer suggests.