I get the desire for courses on AI ethics, failures, etc, but if the students taking them aren't also the student hired to do stuff with LLMs, is the course effective? Are there courses that integrate ethics etc into a hands-on: "hey, let's (sensibly) build cool (responsible) AI stuff"? - ThreadSky

rohanalexander.bsky.social • 21 days ago

I get the desire for courses on AI ethics, failures, etc, but if the students taking them aren't also the student hired to do stuff with LLMs, is the course effective? Are there courses that integrate ethics etc into a hands-on: "hey, let's (sensibly) build cool (responsible) AI stuff"?

Comments

avivshachak.bsky.social•21 days ago

The people who develop AI are not necessarily the ones who make policies and regulations. And TBH, I wouldn’t want them to. “Technology is too important to leave it in the hands of technologists” (E. Yu). Of course,those who make policy and regulations need to know some things about AI.

avivshachak.bsky.social•21 days ago

A technological literacy framework can help determine who needs what education. Here’s an example of how it can inform health professions education. We left out the AI part bc of word limit but it was in the presentation. https://ebooks.iospress.nl/doi/10.3233/SHTI240699

njr0.bsky.social•21 days ago

@gvwilson.bsky.social talks about this a lot.

I’ve been thinking about it a lot as I write the TDDA book.

I’m not really centring on it. But it’s kind of there as a drumbeat.

It’s tough.

Data scientists really need to take the Hippocratic Oath: First do no harm.

Including the ones at Meta…

mjdonnelly.bsky.social•21 days ago

We offer one to future policy-makers (by the very talented @blairaf.com). In theory, those lessons could get into models via the force of law. In general, though, I think that's right. We need more courses that really integrate both.

rohanalexander.bsky.social•21 days ago

Oh I didn't know about Blair's course - thank you.

mjdonnelly.bsky.social•21 days ago

It's great. They don't do much hands-on modeling, but the students love it. And since they're MPP students, many of them will end up briefing decision makers as soon as July, so it could actually have policy consequences (to the extent Canadian policy can influence these things).

blairaf.com•21 days ago

Agreed that ethics for devs is important! Many of my students aren't devs, but want to learn about capabilities & risks of ML apps in different orgs and sectors so they can help shape corporate/public policy, procurement, or to make more informed decisions about using AI tools in their workplace

zaraworkercouncils.bsky.social•21 days ago

Digital humanities is the language studies version of that which I think could be a great alternative to just letting each professor make different rules about LLM use in their classroom.

Comments

Posting Rules

Reply