5 ms·

I was disappointed to find out that AI ethicists at Google aren't LessWrong style AI ethicists. I've just read one Google AI ethics paper in it's entirety but i

I was disappointed to find out that AI ethicists at Google aren't LessWrong style AI ethicists. I've just read one Google AI ethics paper in it's entirety but it seemed pretty non-insightful and not actionable. That said, if Google is "investigating" her because she was trying to find evidence of discrimination or bad treatment of Dr. Gebru, that seems borderline criminal behavior. (Assuming her attempts at trying to find evidence did not involve exfiltrating secret information).

actuator 6y ago
I think sending it to external accounts probably triggered this, there is no way this would not have raised a red flag in any company. I mean, Alphabet is the same company where a person took self driving documents with him to start a new company. Hopefully, if it was just normal conversations and after identities had been redacted, their access is restored. Still, analysing this information on their work machine itself and then sending out a summary seems like a better course of action than this.
- fossuser 6y ago
  Somewhat related - Levandowski (the crook that stole the self driving docs) was just pardoned by Trump. The pardon was supported by Thiel, Luckey, and other Founders Fund people. I find this pretty irritating, but I guess when you’re in the company of actual war criminals like Eddie Gallagher then the bar is pretty low.[0] —— Separately, from the drama I’ve seen around the recent AI ethics stuff, I’d bet money Google is on the right side of this and the activists are not. [0]: https://news.ycombinator.com/item?id=25843040 https://news.ycombinator.com/item?id=25843040
  - bergstromm466 6y ago
    > Levandowski (the crook that stole the self driving docs) Why is your username 'foss user' when you're showing you are for intellectual property rights by calling this person a crook?
- ALittleLight 6y ago
  True, but it's only Google/Alphabet saying she sent data to external accounts and I don't think they're especially trustworthy. If she sent secret info to outsiders she deserves to be fired, but if Google is investigating her because they think she's looking for evidence of discrimination then Google deserves whatever civil and legal penalties apply.
cauthon 6y ago
> I was disappointed to find out that AI ethicists at Google aren't LessWrong style AI ethicists. I've just read one Google AI ethics paper in it's entirety but it seemed pretty non-insightful and not actionable. This seems like a contradictory criticism to make. The Less Wrong folk, in my perception, mostly dabble in thought experiments (such as Roko’s basilisk) that aren’t terribly relevant to real world AI work. Timnit’s work on identifying racial and gendered inaccuracies in facial recognition [1] seems much more actionable. [1] https://www.technologyreview.com/2020/06/12/1003482/amazon-stopped-selling-police-face-recognition-fight/ https://www.technologyreview.com/2020/06/12/1003482/amazon-s...
- ALittleLight 6y ago
  Gebru's (and co-author's) work on racial error in facial recognition was done while they were academics at Stanford and MIT respectively and not while a Google AI ethicist to my knowledge. Roko's Basilisk is a banned comment from Less Wrong (though interesting) and doesn't really represent the totality of that site or the associated ideas. https://news.mit.edu/2018/study-finds-gender-skin-type-bias-artificial-intelligence-systems-0212 https://news.mit.edu/2018/study-finds-gender-skin-type-bias-...
  - fossuser 6y ago
    Yeah - the basilisk thing is mostly a joke as far as I can tell and has little to do with the AGI goal alignment friendly AI work or the control problem. Miri is mostly focused on the AGI control problem and thinks that AGI is closer than others believe. If true, all other problems are pretty irrelevant. There is real work in ML bias and fairness to be done too, but it seems overrun by toxic personalities and partisan politics.
  - cauthon 6y ago
    I’m not sure which Google AI paper you’re referencing. The recent paper at the center of the controversy did have actionable suggestions regarding model training costs (financial and environmental), appropriate training sets, and generally incorporating domain expertise to avoid misinformation or mistranslations. [1] That’s fair, I haven’t visited LW in a while. I did thoroughly enjoy HPMOR, but thought that most of Eliezer’s other work fell closer to speculation than practice. As other commenters have mentioned, these AI algorithms are already being implemented in the real world with real consequences. Their associated ethical concerns therefore seem more urgent, and they can be acted upon. [1] https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru/ https://www.technologyreview.com/2020/12/04/1013294/google-a...
unhammer 6y ago
https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru/ https://www.technologyreview.com/2020/12/04/1013294/google-a... is a summary of the paper that started it, full of actionable points – and co-authored with Emily Bender, no less! Also, if you're into the LessWrong mindset and are susceptible to think that that which can be destroyed by the truth should be, you might find this tidbit interesting: > Buried in the recent trillion parameter language model paper is how the dataset to train it was created. Any page that contained one of these words was excluded: https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words/blob/master/en https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and... Two sample banned words: "twink" and "sex" https://twitter.com/willie_agnew/status/1350551463718621184 https://twitter.com/willie_agnew/status/1350551463718621184
- ALittleLight 6y ago
  Could you elaborate on this a little more? I'm not sure I'm connecting all the dots.
- dgellow 6y ago
  I read your links and I have to say that I do not understand how the list of excluded words relates to the rest. And what’s the issue here.

actuator 6y ago

I think sending it to external accounts probably triggered this, there is no way this would not have raised a red flag in any company. I mean, Alphabet is the same company where a person took self driving documents with him to start a new company. Hopefully, if it was just normal conversations and after identities had been redacted, their access is restored. Still, analysing this information on their work machine itself and then sending out a summary seems like a better course of action than this.

fossuser 6y ago
Somewhat related - Levandowski (the crook that stole the self driving docs) was just pardoned by Trump. The pardon was supported by Thiel, Luckey, and other Founders Fund people. I find this pretty irritating, but I guess when you’re in the company of actual war criminals like Eddie Gallagher then the bar is pretty low.[0] —— Separately, from the drama I’ve seen around the recent AI ethics stuff, I’d bet money Google is on the right side of this and the activists are not. [0]: https://news.ycombinator.com/item?id=25843040 https://news.ycombinator.com/item?id=25843040
- bergstromm466 6y ago
  > Levandowski (the crook that stole the self driving docs) Why is your username 'foss user' when you're showing you are for intellectual property rights by calling this person a crook?
ALittleLight 6y ago
True, but it's only Google/Alphabet saying she sent data to external accounts and I don't think they're especially trustworthy. If she sent secret info to outsiders she deserves to be fired, but if Google is investigating her because they think she's looking for evidence of discrimination then Google deserves whatever civil and legal penalties apply.

cauthon 6y ago

> I was disappointed to find out that AI ethicists at Google aren't LessWrong style AI ethicists. I've just read one Google AI ethics paper in it's entirety but it seemed pretty non-insightful and not actionable. This seems like a contradictory criticism to make. The Less Wrong folk, in my perception, mostly dabble in thought experiments (such as Roko’s basilisk) that aren’t terribly relevant to real world AI work. Timnit’s work on identifying racial and gendered inaccuracies in facial recognition [1] seems much more actionable. [1] https://www.technologyreview.com/2020/06/12/1003482/amazon-stopped-selling-police-face-recognition-fight/ https://www.technologyreview.com/2020/06/12/1003482/amazon-s...

ALittleLight 6y ago
Gebru's (and co-author's) work on racial error in facial recognition was done while they were academics at Stanford and MIT respectively and not while a Google AI ethicist to my knowledge. Roko's Basilisk is a banned comment from Less Wrong (though interesting) and doesn't really represent the totality of that site or the associated ideas. https://news.mit.edu/2018/study-finds-gender-skin-type-bias-artificial-intelligence-systems-0212 https://news.mit.edu/2018/study-finds-gender-skin-type-bias-...
- fossuser 6y ago
  Yeah - the basilisk thing is mostly a joke as far as I can tell and has little to do with the AGI goal alignment friendly AI work or the control problem. Miri is mostly focused on the AGI control problem and thinks that AGI is closer than others believe. If true, all other problems are pretty irrelevant. There is real work in ML bias and fairness to be done too, but it seems overrun by toxic personalities and partisan politics.
- cauthon 6y ago
  I’m not sure which Google AI paper you’re referencing. The recent paper at the center of the controversy did have actionable suggestions regarding model training costs (financial and environmental), appropriate training sets, and generally incorporating domain expertise to avoid misinformation or mistranslations. [1] That’s fair, I haven’t visited LW in a while. I did thoroughly enjoy HPMOR, but thought that most of Eliezer’s other work fell closer to speculation than practice. As other commenters have mentioned, these AI algorithms are already being implemented in the real world with real consequences. Their associated ethical concerns therefore seem more urgent, and they can be acted upon. [1] https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru/ https://www.technologyreview.com/2020/12/04/1013294/google-a...

unhammer 6y ago

https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru/ https://www.technologyreview.com/2020/12/04/1013294/google-a... is a summary of the paper that started it, full of actionable points – and co-authored with Emily Bender, no less! Also, if you're into the LessWrong mindset and are susceptible to think that that which can be destroyed by the truth should be, you might find this tidbit interesting: > Buried in the recent trillion parameter language model paper is how the dataset to train it was created. Any page that contained one of these words was excluded: https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words/blob/master/en https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and... Two sample banned words: "twink" and "sex" https://twitter.com/willie_agnew/status/1350551463718621184 https://twitter.com/willie_agnew/status/1350551463718621184

ALittleLight 6y ago
Could you elaborate on this a little more? I'm not sure I'm connecting all the dots.
dgellow 6y ago
I read your links and I have to say that I do not understand how the list of excluded words relates to the rest. And what’s the issue here.