[lj3]

> That's different than cherry-picking on purpose in order to "win" an argument, which is indeed a human flaw and not a technological one.

As much as I want to believe that, it has not been my experience. I've found that if there's a possible way to misinterpret a statement, somebody on the internet will misinterpret it. They don't do it on purpose to win an argument most of the time, either. It happens everywhere on the internet (HN, Reddit, twitter, even blog comment sections) because communication is difficult and internet increases the exposed surface area of a statement. The more people to read a thing, the more likely it is somebody will interpret something that wasn't intended.

[the_af]

Maybe. But wouldn't you say this flaw is more likely with a tech platform that breaks down a debate into 140 character long pieces and makes it difficult to see the whole context?

Look at the three tweets Antirez mentions. The second one starts

"the 99% percentile is bad [...]"

Antirez mentions someone took this to literally mean that percentiles are a bad metric. To me, this doesn't particularly require bad faith or malicious cherry-picking; it just requires a bit of carelessness, enabled by the lack of context and no indication this was part #2 of a 3-long series of tweets. And apparently, once the "rebuttals" start piling up, they generate a snowball effect. Not entirely Twitter's fault, but definitely made worse by it, which is what (I think) Antirez is arguing:

> "Once upon a time, people used to argue for days on usenet, but at least there was, most of the times, an argument against a new argument and so forth, with enough text and context to have a normal condition. This instead is just amplification of hate and engineering rules 101 together."