I'm increasingly frustrated with "I don't know enough about ethics to include it in my class."
My take: If you e.g. teach an ML class & you actually don't know ANYTHING about ML ethics... learn. Ethics is part of ML. I'm sorry your education was lacking, but now time to learn.
Like, professors should be constantly continuing to learn. I learn new things so that I can teach them all the time. If you say "I don't know enough about ethics" what you probably mean is "I don't care enough about ethics."
I'm not suggesting that if you teach an ML class you need to go read Kant. I'm suggesting go read about ethics IN YOUR FIELD. Ethics is part of that field. So go learn it in the same way you keep up to date on ANYTHING that is new or you don't know!
So I am really really glad that a lot of people are trying to create resources to help with integrating ethics into classes. And maybe that will make it easier. But I also feel like it's not unreasonable to ask instructors of technical classes to do some work to learn. 🤷♀️
To clarify, I am NOT suggesting that this is a substitution for actual ethics coursework from experts trained in it. Just that along with that coursework, ethics about a specific thing should be at least touched on in the course for that thing.
Also a moment of optimism: ethics education for technologists is really improving! So our next generation of computing professors will be much better equipped to touch on ethics in their classes. 🥰
It's a bit scattered across multiple threads but if you read this also check out conversation with @allergyPhD and @Greene_DM about the importance of not teaching ethics as part of CS at the expense of e.g. humanities departments with deep expertise.
In honor of PhD application season winding down, here's a tweet-thread-that-should-probably-be-a-blog-post on things I have observed through hearing from a LOT of PhD applicants in many different fields over the past eight months. TL;DR this process can be better. 🧵
Consider things admissions committees or faculty might expect to see in an application: LORs from certain types of people that mention certain things, statements of purpose with certain elements (faculty mentioned, why this program, etc.)
Are applicants TOLD these expectations?
As one critical example, are you in a field where it is common or there is even an expectation that applicants reach out to potential advisors before they submit an application? Is there any reason applicants would know this if they don't already have mentorship in that field?
This thread is for live-tweeting the ethics session at #SIGCSE2021. FIVE papers at @SIGCSE_TS this year about ethics in computer science education! 🧵
First up: "How Students in Computing-Related Majors Distinguish Social Implications of Technology" by Diandra Prioleau et al. at University of Florida.
They presented students with scenarios about AI technology (e.g. recidivism algorithms)...
... and found that their participants could spot social implications, but frequently missed issues of systemic discrimination. But surprisingly: About half of students had never heard of these issues, which points to a gap in computing curriculum. dl.acm.org/doi/10.1145/34…
As you know, I am a fan of @tiktok_us these days, but I need to put them on blast for a bad design choice. Folks interested in content moderation/platform safety, buckle up. This is a story about bad people exploiting a loophole for harassment. We can learn from this. [Thread 🧵]
TikTok has a "block" feature that works similarly to Twitter. If you block someone, you can't see them and they can't see you. This includes comments.
So now we have A (person being harassed) and B (awful person who thinks it's fun to e.g. leave death threats in comments)...
B has figured out that they can comment on A's post and then immediately block A, which then means that A can't see that comment - and in fact doesn't even know it's there since it doesn't show up in their notifications.
Some thoughts on why this news - the potential for next gen watches (both Samsung Galaxy and Apple) to provide blood glucose readings - could be game-changing (not so much for #T1D folks but for everyone else). macrumors.com/2021/03/05/app…
This tech would almost certainly not be an improvement over existing continuous glucose monitors like what I use, but (I think?) it's rare for people with Type 2 to have insurance coverage for CGMs, especially if you're not on insulin and don't have to worry about lows.
The beauty of continuous monitoring over finger sticks is that you can get DATA. Unless you waste a lot of (expensive!) test strips to try to experiment, you're not going to know e.g. exactly when and how much your blood sugar spikes after meals.
Ever thought about how messed up it is from a harm vs benefit perspective that copyright infringement is more heavily moderated/enforced than, say, hate speech and harassment? I was reminded of this by re-listening to this @ThisAmerLife episode. [Thread🧵] thisamericanlife.org/670/beware-the…
The second act is the story of Lenny Pozner, the father of a Sandy Hook victim, who was harassed, threatened, and stalked by Alex Jones fueled conspiracy theorists accusing him of being a "crisis actor." And one tactic was making cruel memes out of photographs of his son.
And after trying to report content and get things taken down for lies and harassment, he finally realized that his best course of action was reporting copyright violations since he owned the photographs which were e.g. used in a YouTube video.
Not that I was *surprised* to see this study about predicting "political orientation," but since I've been talking about the "gaydar" (sigh) algorithm from the same researcher for a while now, here's some reflection. nature.com/articles/s4159…
Given criticism of the previous paper (which if you're not familiar is here: psyarxiv.com/hv28a/ ) I was genuinely expecting to see an ethical considerations section by the end of this paper (since that criticism pretty much constructed it exactly!). There is not one.
There is a lengthy "author notes" document linked to from the article that includes FAQs (like "physiognomy????") and twice warns to not "shoot the messenger" so I guess that's the ethics statement.