[specproc]

The thing with social media platforms is that this data is user-generated, so you've got the company "owning" user content.

This data is often of great public value. I track conversations around a social issue as part of my work for a non-profit.

I'd counter it's unethical to prevent people from accessing this data.

[jnsaff2]

I’m not disagreeing with your comment but

> great public value

Having been to twitter mostly through the most recent prominent war, man the signal to noise ratio is really low even when being careful about who to follow and who to block. There is so much disinformation, bad takes, uninformed opinions presented as facts, pure evil, etc.

So I guess it could be used for training very specific things or cataloging the underbelly of humanity but for general human knowledge it’s a frigging cesspool.

[specproc]

OK, not gonna argue with that. There is, I guess, a perception that it matters because policy-makers, and the wonks and hacks that influence them are hooked. The value for me (and ergo the public, some classic NGO thinking there for you) lies in understanding those dynamics.

I do not use the Twitters myself, and actively discourage others from doing so. Sends people bonkers.

[lmeyerov]

I mean, we have found election manipulations like large-scale inauthentic activity of out-of-staters explicitly targeting African Americans, and projects here even to the extent of the perpetrators getting indicted. Other projects were tracking vaccine side-effect self-reports faster than the CDC and other disaster intelligence.

We were actually gearing up to switch to paid accounts as we found use cases that could subsidize these efforts... And then the starting price for reasonably small volumes shot up to like $500k/yr.