chloé bakalar

@benzevenbergen @vineshgkannan @alfredtwo @ed_saber @EdFelten Thanks, @benzevenbergen! This is definitely Phase 2 for the project. We don't just want to provide access to the cases, but we also want to give people the tools to teach them. That's the only way to scale them out. We're playing around with ideas, so stay tuned!

A3: I worry about bias in algorythms in AI. I also worry about the motives of people using it. Marketing? Government? Police? Bad algorythms could result in a lot of hurt for people. #ethicalcs

@sophiebakalar Thanks for getting the word out, Soph!

Amazing and incredibly thought-provoking case studies on AI and ethics from Princeton. My sister Chloé co-authored these and I am very proud to be related to her! 😬 she’s not on Twitter but I’ll pass along any comments. twitter.com/princetoncitp/…

#EthicalCS A6: As for resources, the case studies @benzevenbergen, @EdFelten and I developed this year are a good place to start. aiethics.princeton.edu/case-studies/ The idea isn't to "solve" AI Ethics, but to help students begin learn to ask and identify "morally awake" questions. twitter.com/ed_saber/statu…

@ed_saber Even just deciding which news gets prioritized in our individual newsfeeds! It's a necessary function, but each design choice comes with tradeoffs.

#EthicalCS A5: As a political theorist, I especially worry about the way automated content moderation algorithms on social media and search engines can contribute to degraded democracy. Even when that's not the intent. twitter.com/ed_saber/statu…

@ShlomoArgamon Absolutely! Calls for algorithmic transparency are misguided, but given the role AI systems are increasingly playing in day-to-day life, there needs to be some method of holding proprietors accountable for their decision-making.

#ethicalCS Yeah, and the thing is that the operation of these systems is extremely opaque to users, and even often to their creators. Transparency of operation is a must-have. twitter.com/cbakalar/statu…

#ethicalCS AI: neither inherently good nor inherently bad twitter.com/c_valenzuelab/…

@vineshgkannan @benzevenbergen @alfredtwo @ed_saber @EdFelten Thanks! @benzevenbergen and I starting work on that. For now, the cases are freely accessible to all.

@benzevenbergen @alfredtwo @ed_saber @cbakalar @EdFelten These are gripping! Are there resources for people who want to facilitate discussions around these case studies?

#ethicalCS A4: AI is not value neutral--in design or use--nor are the tools perfect. While AI may be the best was to address certain problems, it's important to maintain healthy skepticism towards claims of their omnipotence. Imperfect, like all tools. twitter.com/ed_saber/statu…

@Aidan_Power_ Even where the effects are equally distributed, they may have unforeseen ethical and political consequences.

@Aidan_Power_ Exactly! Treating people as data points (or clusters thereof) influences the way that we think of what it means to be human, especially for those particular humans.

#ethicalCS Broadly, the wide-spread use of data-driven processes to analyze the world, identify problems and produce solutions represents may often be for the best, but we don't know how it'll impact the way we view society and one another. twitter.com/ed_saber/statu…

#ethicalCS A3: I think out largest concern should be uncertainty. Uncertainty about how data that’s already been collected may be stored and used in the future. And not even nefarious uses, necessarily. twitter.com/ed_saber/statu…

#ethicalCS A2: Various stakeholders participate (rightly) in different ways - from identifying questions AI ought to be used to address to regulation and oversight. The point is everyone has a role to play in ensuring the ethical development of AI twitter.com/ed_saber/statu…

@ed_saber #ethicalCS Q1: Intro yourself & describe your relationship to tech, computing, and CS education. I'm a senior researcher at @PrincetonCITP with a background in political theory and philosophy. Particular interests in the relationship between democracy and internet-mediated speech

We’re excited to announce the publication of four original case studies from the “Princeton Dialogues on AI and Ethics” project, in collaboration with the University Center for Human Values: aiethics.princeton.edu and freedom-to-tinker.com/2018/05/21/pri… #AIEthics

social speech (in the form of the beatles) overwhelms hate politics: youtube.com/watch?v=mxoVis… #thepersonalispolitical #dissertationinaction

@bittims break both legs (and maybe an arm)! #thinkofthefishes

@theBossOfFloss ok, you win. i finally stopped feeling sorry for romney. (in the time it took me to post this, he made $54.)

supporting the downtown economy by eating and drinking with @NDFerguson. much love to @AppellationNYC @barbutonyc