We have been talking about "ethics" in tech for a while and I am not saying that that's irrelevant. But I feel like talking about "politics" is more relevant: Ethics tells you not to do or to do a thing. Politics explains why someone wants to build a thing and allows questioning that, challenging that. But we let tech hide behind "being neutral" or whatever BS they call it.
@tante I mean this is the classic "do artefacts have politics" thing right (yes). It's been a part of the discussions of ethics for a long time (was my introduction to it!). I don't think you can divorce ethics from politics quite so easily. I think ethics also requires that sort of reflective questioning. Otherwise how do you get to the "do or do not"?
@CatherineFlick my feeling right now is that we kinda settled on a set of "ethical princinples" whose politics are often largely unexamined (like privacy/data minimization or whatever). That's why the EFF still gets to shape so much of digital discourse while mostly being a libertarian org who doesn't share too many values with many of the organizations basing their thinking on their work.
@CatherineFlick (EFF just being one example here, it's not specifically about them)
@tante I mean I don't disagree with you, but I'm not sure what you mean by "largely unexamined" principles. Having written quite a few of them (e.g. ACM Code of Ethics) there was quite a lot of examination at the time of writing! But I think the issue is that it's sort of considered a "solved problem" rather than the start of an ongoing conversation. That's why you get big companies with useless "Responsible Innovation" teams or "trust and safety" teams that have no teeth.
@CatherineFlick @tante "industry unbound" is a really really good book by @ariezra on how compliance teams and regimes even those staffed by really motivated individuals become reduced to box-checking and decentered through decision-making by "engineering fiat"
@hipsterelectron @tante @ariezra I'll add it to the list! thanks for the rec
@CatherineFlick @tante the ACM is a fairly niche organization though. Wikipedia says 110k members, as opposed to ~20 million developers worldwide. So the top 0.4% of ethics enjoyers probably considered this quite seriously (as did many academic ethicists) but the broader audience of median HN readers, it seems to me, internalized something like "encryption good, DRM bad, regulation bad" and stopped thinking about it for a decade or so.
@glyph @CatherineFlick I think this basically nails it. Sure there are very informed debates on anything, but do they reach mainstream? The application of "ethics" has become a ritual (for people who want to feel good), it's a set of learned processes to go through, not a tool to structure and guide actual, critical thought.
@tante @CatherineFlick I suspect you've both already read this one but for any onlookers to this discussion, I think this article is highly revealing about the structural dynamics that make our field apparently resistant to a robust ethical immune response: https://siderea.dreamwidth.org/1817931.html
@glyph @tante I haven’t read this! But it’s an argument we’ve been having for decades in the computer ethics/tech ethics field. And sadly until governments require it, it won’t happen. And and and I can go on with all the ands. I think most of us have largely given up on trying to get licensing on the books. They have it in one though. There’s two levels of discussion here - one is the aspirational “how should we be” and the other is “if you do/don’t do this you’re bad”
@CatherineFlick @tante Oh it's a really fascinating read if you haven't, I definitely recommend it. Some very uncomfortable truths in there.
Personally what this convinced me on was less "licensure is the only path forward" and more "we need basic liability rules for software". Licensure is a sophisticated and complex way of transferring liability to individual practitioners, but if the service provider isn't even liable in the first place, I don't think it's the bottleneck
@CatherineFlick @tante and emphatically yes, "governments need to do something" is the core problem here. it's *possible* (although vanishingly unlikely) that corporations could get together self-regulate, but it's not even hypothetically possible that individual practitioners could make any real change (without industry-wide unions, at least, which is an unrealistic lift without even a tiny fraction of shops being union yet)
@CatherineFlick @tante in a way this is why I feel a sense of weltschmerz when looking at e.g. <https://opensourcepledge.com>. It's a good effort! Companies should participate! Good on Sentry for doing it! But it's also the exception that proves the rule. It's an industry-defining norm-shattering radical thing to propose that companies pay a relative pittance to mitigate existential threats *to their own business* simply because they *might* be able to continue to free ride.
@CatherineFlick @tante but the possibility of these companies engaging in collective Odyssean bargains establishing self-restrictions to mandate ethical standards of behavior or product development is light-years away from this, we can't even crawl but society needs us to fly
@glyph @tante @CatherineFlick thanks for sharing! I think tying ethics to (usually personal) certification. Ultimately I would argue the hard part is distinguishing what actions require the code of ethics/the certification. Is it pushing to prod? Whose prod? Is it sending PRs? Reviewing PRs? What about open source software? What about "low code" platforms?
I'm not saying we shouldn't try, but I think a lack of clarity where to draw the line is part of why many reject professional licensing.
@Namnatulco @glyph @tante well arguably it should be well before you’re pushing anything to prod :) and really, all of those things! The ACM Code (which I had a hand in) has some good guidance for use, but the hard part is more about who takes responsibility in a complex system/large team. It’s very different from eg medicine where there is usually a very clear line of responsibility.
@CatherineFlick @glyph @tante yeah, "Push to prod" was supposed to be my "we can all agree on this one" example, though I'm aware many would fight me on this ^^
As for the ACM guidance, I think it has a very strong academic focus (for good reason), but I don't think I know anyone outside of academia that knows what ACM is
@Namnatulco @glyph @tante that’s interesting! It was written with our industry members taking a large part so it is aimed at all professionals, not just academia. By nature these things need to be a bit academic to ground them but the intention is for it to be far broader than that. And sure, it’s not the biggest organisation, but it is the biggest organisation of computing professionals. We gotta start somewhere (we def have a large proportion of industry members!)
@CatherineFlick @glyph @tante yeah, I think this is also a USA/Europe divide. In Germany specifically, the Gesellschaft für Informatik (which I think is in part also the German ACM chapter, though they are separate entities) also fulfills a similar role, though I think they are slightly more well known, especially among those that have supporting or compliance roles. But I think if you ask random programmers or IT admins, most will not be familiar with either.
@CatherineFlick @glyph @tante (disclaimer: in this split I am more on the compliance end of the spectrum)
@tante @CatherineFlick not an expert but it sounds like the definition of ethical analysis might have been watered down into technosolutionist terms from the more expansive forms of ethics that could be discussed. i recall vividly how my cs major used to have a very good ethics requirement from the philosophy department which my year was converted into a course taught by a terrible cs professor widely renowned as an easy A instead of allowing students to interrogate and form conclusions
@tante @CatherineFlick i think legal orgs like the EFF/ACLU/IA are very prone to this sort of thinking and the litigatory approach of ACLU/IA tends to platform less lengthy discussions of what should be vs a cauterization of the status quo and a specific concept of free speech. to their credit the EFF also advocates on policy and not just litigation although i'm not familiar with their efforts to craft vs shoot down policy
@hipsterelectron @tante I think they're also required to act like this - they are funded by people who want noisy results, so there's an incentive to have a hard line and stick to it.
@CatherineFlick @tante that's a much better way of phrasing it thank you