[Air-L] Twitter Data Access Update

[Shulman, Stu]

There is an important account of the impact of the loss of access to the
Twitter API in the paper today:
https://www.washingtonpost.com/technology/2023/06/20/twitter-policy-elon-musk-api/

The restrictions on the free API are a problem, though the idea of free in
this context is misleading. It was never free. Shareholders, lenders, and
advertisers paid to make it possible to expose the data via APIs in ways
that diverted attention from the real cost of engineers, buildings,
electricity, bandwidth, lawyers, HR, sales people, and many other expenses
that pay for large scale platform operations. Somebody always paid for all
of that or else it went into the Twitter debt pile, which was large and
growing at the time of the acquisition. The current owner has laid bare the
finances of Twitter in stark and bizarre ways while undermining
advertising.

I have spent the last few months talking to a wide variety of stakeholders
about what is possible now. There are third party data vendors that have
deals with Twitter that predate and supersede the $42K/month options widely
reported in the press. We are closer to launching a new multi-university
consortium to ensure that at least some Twitter data remains accessible
going into the fall 2023 semester and 2024. It will not be free or fully
unlimited. It will not be "big data" on the scale of 10s or 100s of
millions of Tweets. Nor will it arrive via the API, at least not at first.
However, it will be well curated and stored in a manner where teaching and
collaborative multi-institution research will take place with fresh data
samples pulled weekly from the full Twitter firehose going back up to 15
months. Once the cornerstones of the data access consortium are
established, new short-term and one-time options will come into play for
students and other researchers that are not a part of the founding group of
universities. Barring a mercurial violation of existing contracts, Twitter
data will be accessible along with blogs, YouTube, and some other media
sources as well.

I have heard a number of compelling reasons to move on from Twitter while
doing this consortium-building work. My concern is the loss of transparency
at a critical juncture in history. There is a surging flow of openly
problematic hate speech. Networks dedicated to election and COVID denialism
are publishing massive amounts of content. A full-time post-J6 insurrection
industry is ramping up powered by trolls and bots. The QAnons and
anti-vaxxers are in the open telling preposterous stories and demanding
violent revenge. If we move on from studying Twitter, or let one person
determine the framework of possibility for thousands of academics, we
may look back in the near future with some regret and wonder whether
anything could have been done.

History is path dependent. This is a proposed alternate path where we keep
academic eyes focused on Twitter. The new owner may prefer to have
academics give up, protest via workshops and special issues, and then move
on. I am loath to comply with that planned diversion of attention. It is
not safe for teetering democracies. My undergraduate professor of American
political history, Howard Zinn, would likely urge us all to stay vigilant
via collective action if he were here today. This is a moment where
research, scholarship, and activism necessarily collide. You cannot stand
still on a moving train. Anyone watching the RFK Jr Express leave the
station, backed and boosted by the Twitter platform, knows the US election
in 2024 is going to be an epic information war.

Please contact me for a data demo or more information about having your
institution join the consortium:
https://calendly.com/discovertext


-- 
Dr. Stuart W. Shulman
Founder and CEO, Texifter
Editor Emeritus, *Journal of Information Technology & Politics*