17 September 2021

A new data ethics curriculum, and fellows to implement it

On Friday I briefly talked about a call for data ethics fellows that’ll be coming out soon, promising more information on Monday. well, it’s Monday Wednesday Friday. My bad. At the bottom of this page is a link to the draft call for applicants as it exists today, but I wanted to make as much of the info behind what we were thinking available as I could, and I want to hear feedback if there are parts of this that are unclear or ambiguous, and similarly if there are questions that I can answer or incorporate into the final version.

Anyway, here are the bullet points, with expanded information, and the link to the actual draft call for fellows further below:

  • Full time
  • Appointment 1-1.5 years (extension possible depending on funding)
  • Salary: $75,000
  • Relocation to SF not a requirement, but is supported if desired
  • Academic background not required

Okay, let’s work through this:

CADE will be taking applications for a data ethics fellowship, seeking fellows who will help me with something I’ve been thinking about for some time - a plan to redesign USF’s curriculum for data ethics to focus on 3 things:

  1. to more deeply engage with ethics theory;
  2. to meaningfully introduce students to methodological tools to identify and analyze causes of harms; and finally
  3. to teach students about a rotating set of domains where harms are mediated by algorithmic or data systems.

I’ve been thinking about this a lot, especially since teaching an intro to data ethics course at USF, and the gist of my thinking is that I desperately needed a lot more time with students. And like not even 20% more time, but 200 or 300% more time. At the kind of disparity that I felt, one class just didn’t feel like enough. I needed a lot more time, and more importantly I felt like I just needed more people who knew about the domains that I was cursorily mentioning.

When I got appointed to this role as interim director of CADE (oh right I haven’t blogged about that; I’m the interim director at CADE now), I came to realize that I was in a position to support the work I value; I’ve been incrementally exploring what that new latitude and capacity, as the director of CADE, can do.

Someone asked what kind of fellows I’m looking for, and I’m copying/pasting mostly what I told them:

I want fellows to design and teach courses (as in, involved deep dives) covering the broader area of their work. so I want people who can articulate a clear group of people, or a clear domain of AI/data harms, in their work.

So like if someone’s work has been about anti-Blackness in AI, that feels like something they could design a class around for sure; or if their work has been about AI in clinical settings, that (or maybe a slightly broader topic of health) could be a great topic. There’d be overlap between those two examples, but that’s just to illustrate that the through-lines don’t have to be all parallel to each other.

I really want students to take a course on theoretical foundations so they understand the vocabulary that we use in AI/data justice, and a course on methods so they have some exposure to things like coding interview data, writing as a reflexive practice, what qualitative data offer that quant data can’t, etc… and then to take 2-3 classes about like… “special topics in AI/data ethics”. And the thinking is these offerings would just change as fellows came and left, because the through-line that each person brings is fairly unique to them.

One year we might have someone who’s done a lot of work studying electronic health records, and then a year or two later CADE might have a fellow who studies clinical decision support algorithms. Those aren’t precisely the same thing, but the point isn’t to produce graduating cohorts every year who have some depth on medicine/health and technology; the point is for graduating students to have a very real appreciation of how complicated the story gets when we try to characterize a group of people or a domain.

Okay, enough reflection-bordering-on-solipsism. Here are the bullet points from above with some expansion:

Full time

Someone asked if this position would allow for an existing affiliation with another institution. Courtesy appointments of course are fine, but I don’t think the work we’re asking for lends itself to having multiple appointments each representing a full-time commitment. I hope the duration of the appointment, the salary, and the benefits make it possible for people to make this their main professional commitment.

For what it’s worth, if you’re affiliated with a place that’ll let you take a leave of absence for the time you’ll be at USF, that would be fantastic and I’d have no problem whatsoever with that arrangement.

1/1.5-year appointment (extension possible contingent on funding)

This is trickier to explain, but hopefully makes sense. The appointment is nominally for 1 year. Fellows are guaranteed an additional 6 months if they’ve either taught the domain class described earlier in this post, or if they’re “on track” to teach it (ie a syllabus has been finalized; better still if the class is scheduled for an upcoming term, but I have no idea what the future will look like, so maybe that’s an unreasonable ask, and I don’t want to put up that kind of requirement).

Extensions are possible, but they depend on what our funding situation is, which I’ll describe a bit further down this post.

fellowship salary: $75,000

The low bar for me is that if I ask you to work at a place based in some city, I ought to pay you enough to live in that city.

More on this in the next section, though.

relocation to SF not a requirement, but is supported if desired

As far as I know, USF staff are allowed to be remote for the foreseeable future. How far is “foreseeable” is a problem that I’m trying to iron out, but that has less to do with bureaucratic opacity and more to do with the difficulty forecasting COVID variants.

Even outside of the current ongoing pandemic with no end on the horizon, some people can only safely work from home or otherwise remote from our office in downtown SF; or have family they have to take care of; or any number of reasons. I don’t want a requirement to be in SF to be a reason someone doesn’t apply.

All that being said, if you want to be in or near SF, we can reimburse some of your relocation expenses. The status quo is to reimburse for up to $1500 in relocation costs, which… just isn’t enough. We’re working on increasing that amount substantially, and when the formal draft is ready this will be finalized. We’re still working on the details and I don’t know how much I can say, but it won’t be marginal.

The document we’re trying to bring our policy into parity with imposes a few requirements, the most relevant ones being that you have to be moving from more than 50 miles away, and you have to be moving to within 50 miles of USF’s main campus. I’m assuming that means “50 miles as the crow flies” (which may not be the right reading, so ask me to double check if you might want to relocate). That radius from USF’s main campus (2130 Fulton Street) includes Santa Rosa, Napa, Vacaville, Antioch, and about 95% of San Jose. All this is to say that you don’t need to move to within the city of San Francisco to get the relocation reimbursement.

Is it clear that I’ve been thinking about this a bit? Let’s move on.

academic background not required

People have asked and I’ve tried to answer in correspondence, but for people who don’t know they can DM me or email me or whatever: you don’t have to have an academic background. Journalists, organizers, and activists from all backgrounds have precisely the experience and knowledge that I want students to come to value and appreciate, and aspire to develop in their own work going forward. It would be self-sabotage to restrict this position to people with academic backgrounds.

The things I care about are whether you can do the work - the research, writing, and public work - and whether you’re able to design and teach a class. Academia offers a few shorthands for whether someone can do independent research and writing, but you can have those skills any number of other ways.

I hope it’s clear that I’ve been thinking about what CADE can do to foster a form of centrally and peripherally academic work that’s ethically sustainable given the precarity and often extractive nature of short-term academic appointments. My hope is that the work I’ve described in this post would be mutually beneficial - helping fellows convey their work and the space in which they work in ways that are not only illuminating, but also persuasive of the importance and complexity of the issues they study and document.

Something I appreciate a great deal about CADE is that we’ve never taken money from corporate entities or interests, and we plan never to change that. It’s given me space to think about my discomfort with surveillance and algorithmic technologies without even the briefest thought that our work undermines the business interests of the people or groups who fund us. We’re principally supported by Craig Newmark’s philanthropy. If you join CADE, I’ll do my best never to surprise you with new financial ethical entanglements, and I’ll never pursue financial support that I think would undermine or compromise your independence either in appearance or practice. If you have any questions about that or anything else in this post, please feel free to contact me. Otherwise, look forward for the formal call for applicants.

Okay, you’ve read to the end, here’s the draft call

If you have something to say about this post, email me or tweet at me.