Thread by @emilymbender

A few thoughts on citational practice and scams in the #ethicalAI space, inspired by something we discovered during my #ethNLP class today:

>>

Today's topic was "language variation and emergent bias", i.e. what happens when the training data isn't representative of the language varieties the system will be used with.

The course syllabus is here, for those following along:
faculty.washington.edu/ebender/2021_5…

>>

Week by week, we've been setting our reading questions/discussion points for the following week as we go, so that's where the questions listed for this week come from.

>>

Anyway, the definition of the term "emergent bias" was highly relevant, of course, so I was pulling up Friedman & Nissenbaum 1996:

dl.acm.org/doi/abs/10.114…

>>

In parallel, one of the students did a web search for the term, and landed on this page:

technologicalethics.org/three-kinds-of…

Some of that text looked oddly familiar! But a thorough search of the site turned up exactly zero citations to Nissenbaum & Friedman 1996 (nor any other citations, for that matter).

>>

We poked around the website a bit more, trying to figure out who is behind it, and could only find a first name ("Henry") as well as this helpful invitation to call ... but no phone number:

>>

On the other hand, "Henry" and colleagues are apparently selling ethics seminars, for $495-$1295, depending on the length of the seminar.

>>

And the listed partners include @MonashUni (who might prefer not to be associated with such a shady organization).

>>

Another student turned up this old tweet (from March 2020) advertising a keynote by Henry, whose last name appears to be Dobson:

https://twitter.com/WSAoffice/status/1227152403113369601

>>

And from there, the Medium page, apparently by the same person:

henrydobson.medium.com

(with the tagline "tech ethicist")

I guess it's in vogue to be a "tech ethicist" these days, and of course anyone can hang out their shingle. But someone doing serious work in this space will have good citational practice, relating their work to others'.

>>

It's really easy to see that Henry Dobson is not someone to take seriously---he's flat out plagiarized Friedman & Nissenbaum!---but even if he were presenting his own original work, the lack of citations would be a red flag.

>>

In particular, if someone working in this space (and selling services no less!) appears to be unaware of the work of:

@LatanyaSweeney @jovialjoy @rajiinio @timnitGebru @merbroussard @mer__edith @ruha9 @mmitchell_ai or @safiyanoble

... I will approach their wares w/skepticism.