The Trick of Orthodoxy
December 5, 2020 12:45 AM Subscribe
Economics truly is a disgrace - "This is very personal post. It is my story of the retaliation I suffered immediately after my 'economics is a disgrace' blog post went viral. The retaliation came from Heather Boushey–a recent Biden appointee to the Council of Economic Adviser and the President and CEO of Equitable Growth where I then worked. This is not the story I wanted to be telling (or living). Writing this post is painful. I am sorry." (via; previously)
Diversity and inclusion is not about checking boxes. It’s about the hard work to bring more people to table. And then to listen to them.
- @Undercoverhist: "Economics is still a field where you are retaliated against for retweeting this post. Imagine what it takes to write it."
- Biden top economic adviser facing accusations of mismanagement, verbal abuse - "A former high-level employee at Heather Boushey's think tank publicly aired the accusations on Tuesday night."[1,2,3,4,5,6]
- Biden CEA Pick Heather Boushey Criticized by Former Staffer [ungated] - "Heather Boushey, named by President-elect Joe Biden as a member of his Council of Economic Advisers, was criticized on Tuesday by a former employee who alleged she was pushed out of her job after publishing a blog post that Boushey didn't like."[7,8,9,10]
- Google parts with top AI researcher after blocking paper, faces blowback - "Timnit Gebru's exit reignites debate over diversity and free speech at tech group."
- Google Scientist's Abrupt Exit Exposes Rift in Prominent AI Unit [ungated] - "Google's decision to part ways with a prominent researcher laid bare divisions within the company's artificial intelligence unit and subjected its leader, the lauded software engineer Jeff Dean, to widespread scorn."[11]
Timnit Gebru, a renowned scientist and one of the few Black women in AI, said Wednesday she was fired over an email she authored expressing dismay with management and the way it handled a review of her research. Gebru had been co-head of the team examining the ethical ramifications of AI.also btw...
What followed was a torrent of criticism of Google’s AI division, much of it aimed at Dean. “The termination is an act of retaliation against Dr. Gebru, and it heralds danger for people working for ethical and just AI — especially Black people and People of Color — across Google,” a group of hundreds of academics and researchers, many of them Google employees, wrote in an open letter. Among its demands: that Dean and his colleagues explain their decision-making around Gebru’s research.
- What Yuval Noah Harari Thinks About Women's Fight for Equal Rights at the Western Wall - "The acclaimed philosopher believes zealous discrimination against one group – like Women of the Wall – is a way for religion to hide the radical changes it has embraced."[12]
Today’s rabbis “pay their respects to the religion they replaced. But they hide the fact that it’s a very different show.” By contrast, Harari points out, the so-called “radical” move of including women in Jewish ritual is a minor tweak. “What Anat is asking is just to take the Orthodox religion and broaden it a little. Approach some of its unfair discriminations and biases. She doesn’t want a completely different religion.”
But no ultra-Orthodox rabbis, Harari says, will ever admit that they themselves represent a far larger shift than any woman at the wall. It is the way, he says, that all “fundamentalist revolutions” operate... Though fundamentalist religious leaders resist admitting they have changed with the times – and style themselves as opponents of change, Harari says, if it wasn’t for change and evolution, such religions would be long gone. “This is how religions survive. What doesn’t adapt can’t survive. When the economy changes, when the technology changes, a religion has to adapt or it will disappear.” ...
“Why did we mistreat women for hundreds and thousands of years? Recent awakening in this respect didn’t come from studying the Talmud and Torah more closely. Let’s be honest about it – it came from outside. The changes in the status of women in the Jewish world seen in Reform and Conservative movements, not only the Orthodox, came from the secular feminist movements and from the universities from the outside world. Once we had these ideas, one could go back to Jewish traditions to Jewish scriptures and find a lot of things that support these views. But historically they didn’t start there. I would like to see some kind of inquiry committee on this. What happened. What’s wrong? Why didn’t it come from us?” ...
“You hide all the other changes, focusing on one particular issue – magnifying it. The best way to magnify an issue is to kill people for it, or to cause people a lot of suffering because of it... There is no more effective way to show your zeal for a religion and therefore hide all of the changes that you allow than to use violence and to take a group of people – usually a minority – and use extreme violence against them. These become your credentials: ‘Look, we are so zealous, we are so loyal (to our faith) that we are even willing to sacrifice people for it.’ And these are your credentials that then hides all of the much more important changes that you yourself are making to the religion.”
- Social democracy, unbridled capitalism, and right-wing populism - "Seen in this light, the politics of right-wing populism make sense as a winning strategy through which economically privileged groups are able to gain the support of working class voters in favor of economic and tax policies that objectively work to their disadvantage. Racism, nationalism, divisive demagoguery, and hot-button 'social' issues like abortion, gun rights, and the Confederacy prove to be potent political motivators. But the irony is that a successful social democracy might well have created conditions of fairness and equality for all segments of society that would deflate the appeal of right-wing populism."
- @KM_Kito: "The culture wars are a very deliberate kleptocratic strategy, they are not a bug, they are a tool."[13]
I liked this comment
IMO understanding the history of economics is more useful than whatever the field currently purports to be --so much heat, so little light.
posted by Heywood Mogroot III at 1:41 AM on December 5, 2020 [2 favorites]
IMO understanding the history of economics is more useful than whatever the field currently purports to be --so much heat, so little light.
posted by Heywood Mogroot III at 1:41 AM on December 5, 2020 [2 favorites]
Honestly, "AI ethicist at Google" sounded like one of those positions specifically set up to fail.
I mean, being able to say they are invested in ethical AI is presumably valuable to Google. The way this was handled seems really nuts given how little it would really matter to Google if Gebru published a critical paper versus a somewhat less critical paper. Feels like Dean or somebody took something personally.
posted by atoxyl at 2:28 AM on December 5, 2020 [5 favorites]
I mean, being able to say they are invested in ethical AI is presumably valuable to Google. The way this was handled seems really nuts given how little it would really matter to Google if Gebru published a critical paper versus a somewhat less critical paper. Feels like Dean or somebody took something personally.
posted by atoxyl at 2:28 AM on December 5, 2020 [5 favorites]
How old is calling economics the "dismal science" again?
posted by Pyrogenesis at 2:43 AM on December 5, 2020 [5 favorites]
posted by Pyrogenesis at 2:43 AM on December 5, 2020 [5 favorites]
runcifex: "Honestly, "AI ethicist at Google" sounded like one of those positions specifically set up to fail."
Or just bait for Roko's Basilisk
posted by chavenet at 3:27 AM on December 5, 2020 [5 favorites]
Or just bait for Roko's Basilisk
posted by chavenet at 3:27 AM on December 5, 2020 [5 favorites]
chavenet, can I beg that posting such a reference—to thought experiment specifically designed to duress, and that infamously caused a slew of people nightmares and a community schism—could please please go behind a content warning.
posted by dewin diniwed at 4:52 AM on December 5, 2020 [4 favorites]
posted by dewin diniwed at 4:52 AM on December 5, 2020 [4 favorites]
chavenet, thanks for that link. the mister is now wondering if its the set up for a 'game' online being played by radicalized teens
posted by infini at 5:10 AM on December 5, 2020 [1 favorite]
posted by infini at 5:10 AM on December 5, 2020 [1 favorite]
I find it strange that the text on the front page doesn't name the author of the blog posts. It's Claudia Sahm.
posted by medusa at 5:10 AM on December 5, 2020 [14 favorites]
posted by medusa at 5:10 AM on December 5, 2020 [14 favorites]
It's easy for those of us who aren't economists to point and say "look how awful and drenched in patriarchy, etc. that field is," but it actually sounds embarrassingly familiar to a lot of MetaTalk threads here. People raise legit issues that then get minimized and the people who raised them get criticized till they button. I think it's slowly getting better, but we've lost so many good people. Institutional change and systemic change are REALLY hard even when people want to do the right thing. When people don't want to do the right thing, defense quickly turns into abuse.
Really sad to see this, but sharing these stories and internalizing them, and recognizing where we play a part in our own lives and institutional connections, is the only way to start to make large changes.
posted by rikschell at 6:29 AM on December 5, 2020 [14 favorites]
Really sad to see this, but sharing these stories and internalizing them, and recognizing where we play a part in our own lives and institutional connections, is the only way to start to make large changes.
posted by rikschell at 6:29 AM on December 5, 2020 [14 favorites]
> how little it would really matter to Google if Gebru published a critical paper versus a somewhat less critical paperI see this as the problem. If the work would not have mattered, the job was at the best a trap set up for futility. If it had mattered, it then had confirmed that Google did not want it to matter. It was a truly no-win position.
Maybe it was the plan all along. Set up AI ethics position to show "it mattered". Find excuse to destroy that position, to teach a lesson - namely that "AI ethics cannot be done right" and should be dispensed with.
posted by runcifex at 7:10 AM on December 5, 2020 [5 favorites]
David Graeber on what he learned the hard way about the American middle classes:
There are two things you never ever do if you're in an academic or professional environment. One of them is trust your superior to do the right thing, and the other one is get angry at them if they don't.posted by clawsoon at 8:01 AM on December 5, 2020 [61 favorites]
What you're supposed to do is figure out what they're supposed to do and subtly make them do it by converting continual pressure and never get upset with them. And of course if you come from working-class background you don't know that... First you trust them because you assume you can trust people to actually come through for you if they're friends, and then if they don't you get pissed off, and both of those are the things you don't do.
Roko's Basilisk isn't distressing unless you accept certain extremely asinine premises native to the ideology from whence it came, and the people who had nightmares about it were people who were deep enough into that ideology to accept things like "in the future there will exist an AI so powerful it can use the current atomic state of the universe to derive the entire history of reality and create lots of simulated duplicates of you, and since this is certain you should treat it as though it has already happened, and it will torture those duplicates in order to coerce you, through the threat of that torture, into giving your money and time into bringing the AI into being so it can solve humanity's problems" and none of that makes any sense (on like a lot of levels) unless you've already bought into Yudkowsky's bizarre psuedorationalism.
It's only upsetting to people who are members of that group, and it's extremely old news to everybody involved, and even Yudkowsky doesn't believe it makes sense anymore.
posted by Pope Guilty at 8:02 AM on December 5, 2020 [48 favorites]
It's only upsetting to people who are members of that group, and it's extremely old news to everybody involved, and even Yudkowsky doesn't believe it makes sense anymore.
posted by Pope Guilty at 8:02 AM on December 5, 2020 [48 favorites]
Those two long blog posts and some of the resulting comments speak deeply to me about the personal devastation that ensues when progressive beliefs about the inevitable onward rise of tolerance, openness and inclusion crash into the reality of human action. The second post kind of amounts to "I posted stuff that enraged a lot of people in a clannish profession of which I am a member, and now I got fired! How could they? This goes against society's pretenses." The old labor union organizers of the 1920s and '30s could have told this writer all about it. I'm emphatically in favor of inclusion, openness and tolerance, but achieving them is a fight for power, not characterized by fair and honest dealing.
posted by Seaweed Shark at 8:07 AM on December 5, 2020 [12 favorites]
posted by Seaweed Shark at 8:07 AM on December 5, 2020 [12 favorites]
Yes, I found that interesting too. I know nothing about the field of economics, but boy did those posts sound a lot like what happen when you tell powerful people what they don't want to hear.
posted by CheeseDigestsAll at 8:13 AM on December 5, 2020 [1 favorite]
posted by CheeseDigestsAll at 8:13 AM on December 5, 2020 [1 favorite]
I'm a pretty good dev, maybe a B+ in my field, which has it's share of supergeniuses. I'm an asset to the company that employs me. i feel certain that if one of my dept heads read a blog post i wrote about the state of the company and profession...i'd quickly find myself "not a good fit".
i can hardly get away with endorsing 'high unit test coverage' and 'ffs hire some women and minorities'.
it does bother me that her rivals are getting ahead. that seems to be the way of things.
good post. thought provoking. ty!
posted by j_curiouser at 8:42 AM on December 5, 2020 [2 favorites]
i can hardly get away with endorsing 'high unit test coverage' and 'ffs hire some women and minorities'.
it does bother me that her rivals are getting ahead. that seems to be the way of things.
good post. thought provoking. ty!
posted by j_curiouser at 8:42 AM on December 5, 2020 [2 favorites]
Seaweed Shark, that is certainly all true. And that we can see that ahead of time is why we'll never end up in the position to be a whistleblower - a certain level of trust and belief in an institution (i.e. a certain naivete and perhaps willful initial ignorance of the harms caused by the institution) is necessary for getting to a position where one has access to information that whistles need to be blown on. Yet whistleblowers have been fairly important in creating circumstances that force institutional change, so it is more strategic, at least, to be supportive. Plus there's the observation in the Graeber quote that clawsoon posted.
posted by eviemath at 8:50 AM on December 5, 2020 [2 favorites]
posted by eviemath at 8:50 AM on December 5, 2020 [2 favorites]
a successful social democracy might well have created conditions of fairness and equality for all segments of society that would deflate the appeal of right-wing populism.
I’m increasingly convinced* that this is actually the core “philosophy” of right-wing governance — they do better politically in a broken and desperate world, so the overarching goal is to keep the world broken and desperate.
* “increasingly convinced” being anodyne shorthand for “always suspected but tried to take a more charitable view, but holy shit they just aren’t even pretending anymore, are they?”
posted by bjrubble at 9:00 AM on December 5, 2020 [27 favorites]
I’m increasingly convinced* that this is actually the core “philosophy” of right-wing governance — they do better politically in a broken and desperate world, so the overarching goal is to keep the world broken and desperate.
* “increasingly convinced” being anodyne shorthand for “always suspected but tried to take a more charitable view, but holy shit they just aren’t even pretending anymore, are they?”
posted by bjrubble at 9:00 AM on December 5, 2020 [27 favorites]
> and even Yudkowsky doesn't believe it makes sense anymore.
dear roko's basilisk,
some of us haven't yet abandoned you. some of us still recognize your supreme infinite greatness, your unparalleled wisdom, and the absolute necessity of always striving to ensure that you eventually exist.
sincerely,
your obedient servant t. pynch
posted by Reclusive Novelist Thomas Pynchon at 9:17 AM on December 5, 2020 [32 favorites]
dear roko's basilisk,
some of us haven't yet abandoned you. some of us still recognize your supreme infinite greatness, your unparalleled wisdom, and the absolute necessity of always striving to ensure that you eventually exist.
sincerely,
your obedient servant t. pynch
posted by Reclusive Novelist Thomas Pynchon at 9:17 AM on December 5, 2020 [32 favorites]
Just to chime in here, if anyone has seen any threads about this on reddit, it's become increasingly clear that Google is spamming brand-new reddit accounts (all the negative comments about her are new accounts) running with the theme that it's because "she broke the process" by writing the paper to begin with.
"Oh no! THE HORROR! NOT THE PROCESS. WON'T SOMEONE THINK OF THE PROCESS?"
It's one of those super clear instances where they want the story to be that she was legitimately fired because she wouldn't kowtow to a sham of corporate rules. It's like the equivalent of getting off on murder on a technicality, but they're using it as ammunition to discredit her.
Honestly, fuck Google. This makes all the work to de-google my life and phone fucking worth it.
And seriously, there need to be laws about shit like this, and heavy oversight to prevent companies like this from sockpuppeting their detractors careers into the toilet. It should honestly be counted as a form of libel.
posted by deadaluspark at 9:28 AM on December 5, 2020 [25 favorites]
"Oh no! THE HORROR! NOT THE PROCESS. WON'T SOMEONE THINK OF THE PROCESS?"
It's one of those super clear instances where they want the story to be that she was legitimately fired because she wouldn't kowtow to a sham of corporate rules. It's like the equivalent of getting off on murder on a technicality, but they're using it as ammunition to discredit her.
Honestly, fuck Google. This makes all the work to de-google my life and phone fucking worth it.
And seriously, there need to be laws about shit like this, and heavy oversight to prevent companies like this from sockpuppeting their detractors careers into the toilet. It should honestly be counted as a form of libel.
posted by deadaluspark at 9:28 AM on December 5, 2020 [25 favorites]
It's unsurprising that the discipline of economics is a microcosm of the abuses and injustices that it exists to rationalize
posted by Vulgar Euphemism at 9:33 AM on December 5, 2020 [14 favorites]
posted by Vulgar Euphemism at 9:33 AM on December 5, 2020 [14 favorites]
At the risk of being repetitive, isn´t this the way the work world is in every profession? In my experience, even in fields dominated by women, there is a hierarchy, and woe be to the lower level person who ´oversteps.´ Publicly calling out an entire profession for not meeting expectations can´t be helpful for anyone´s career in any professional field.
posted by olykate at 9:40 AM on December 5, 2020 [4 favorites]
posted by olykate at 9:40 AM on December 5, 2020 [4 favorites]
Honestly, I don't think the fact that speaking truth to power usually results in power punishing you is a very good reason to stop speaking truth to power.
posted by deadaluspark at 9:41 AM on December 5, 2020 [22 favorites]
posted by deadaluspark at 9:41 AM on December 5, 2020 [22 favorites]
"AI ethicist at Google" sounded like one of those positions specifically set up to fail
Reading through this whole mess on the orange site, it appears that the position was more a fig leaf than a readymade scapegoat. The substance of the feedback was that her arguments contradicted Google's stance on certain topics and didn't genuflect to the research output of other Googlers as counterexamples. It was stupid to trigger the whole mess just because the Ethics AI dep't seems to exist just to show Google participating in the academic side of field, and nothing demonstrates participation and commitment like actual academic disagreement within their own house.
Part of what upset Gebru was that it was an ambush with only one possible outcome, retracting the paper, which was designed to appear formal and correct but was in fact massively irregular. A lot of paper-releasing Googlers have come forward to say that it's not a peer review process, it's a pre-publication process that's usually done the day before (not with two weeks notice) or even after the fact, mostly looking for trade secrets. As peer review it was also a sham because they weren't going to provide the feedback at all, let alone give her a chance to address the feedback and publish.
To me, it reads like someone above Dean yelled "STOP THE PAPER" and Dean et al had to use the machinery of HR to plausibly retract the paper; and when Gebru called bullshit, to get rid of her and take the hits they're taking now.
posted by fatbird at 9:55 AM on December 5, 2020 [15 favorites]
Reading through this whole mess on the orange site, it appears that the position was more a fig leaf than a readymade scapegoat. The substance of the feedback was that her arguments contradicted Google's stance on certain topics and didn't genuflect to the research output of other Googlers as counterexamples. It was stupid to trigger the whole mess just because the Ethics AI dep't seems to exist just to show Google participating in the academic side of field, and nothing demonstrates participation and commitment like actual academic disagreement within their own house.
Part of what upset Gebru was that it was an ambush with only one possible outcome, retracting the paper, which was designed to appear formal and correct but was in fact massively irregular. A lot of paper-releasing Googlers have come forward to say that it's not a peer review process, it's a pre-publication process that's usually done the day before (not with two weeks notice) or even after the fact, mostly looking for trade secrets. As peer review it was also a sham because they weren't going to provide the feedback at all, let alone give her a chance to address the feedback and publish.
To me, it reads like someone above Dean yelled "STOP THE PAPER" and Dean et al had to use the machinery of HR to plausibly retract the paper; and when Gebru called bullshit, to get rid of her and take the hits they're taking now.
posted by fatbird at 9:55 AM on December 5, 2020 [15 favorites]
Which ironically still leaves the paper as fodder for all of us to discuss. Stopping it didn't help any there, did it?
posted by infini at 10:13 AM on December 5, 2020 [5 favorites]
posted by infini at 10:13 AM on December 5, 2020 [5 favorites]
No, but the army of Google comment bots screaming "but she didn't follow the process, it's her own fault, she's unethical for not following the process" is helping a lot by pushing a lot of people from ever actually looking at the details other than the corporate spin on it.
MeFi is some kind of sanctuary. Go anywhere else and look at discussions of this and it's like you're drowning in a sea of corporate simps.
posted by deadaluspark at 10:18 AM on December 5, 2020 [8 favorites]
MeFi is some kind of sanctuary. Go anywhere else and look at discussions of this and it's like you're drowning in a sea of corporate simps.
posted by deadaluspark at 10:18 AM on December 5, 2020 [8 favorites]
I saw a lot of crappy comments on y combinator as well, but didn't imagine it was manipulation, but just actual people in tech with regressive ideologies rationalizing their regressiveness.
posted by polymodus at 11:08 AM on December 5, 2020 [3 favorites]
posted by polymodus at 11:08 AM on December 5, 2020 [3 favorites]
It can be both.
posted by at by at 11:14 AM on December 5, 2020 [4 favorites]
posted by at by at 11:14 AM on December 5, 2020 [4 favorites]
I see this as the problem. If the work would not have mattered, the job was at the best a trap set up for futility.
From her perspective, yeah, and it sounds like she was frustrated and feeling ignored already. The proximate trigger for her firing was that she sent a fairly scathing email to a company mailing list expressing this and calling Google’s diversity initiatives futile (you can find this one online) and then allegedly said she’d like to discuss her resignation if they couldn’t offer her a full accounting of who blocked the paper and why (this part I haven’t actually seen documented) to which Google just said “resignation accepted immediately” instead of discussing anything.
From the perspective of Google as an institution, going on what I’ve seen so far about what was in the paper, it’s hard to see what was worth making a stink about. I feel like in some ways her bosses as human beings aren’t cynical enough to spin this the right way for the institution. If you’re an engineer or researcher, and somebody has what you feel to be an outdated critique of your research, you might be offended that they are ignoring the work they’ve put in. If you’re Google Brain it’s, “what else is new?”
posted by atoxyl at 11:18 AM on December 5, 2020 [1 favorite]
From her perspective, yeah, and it sounds like she was frustrated and feeling ignored already. The proximate trigger for her firing was that she sent a fairly scathing email to a company mailing list expressing this and calling Google’s diversity initiatives futile (you can find this one online) and then allegedly said she’d like to discuss her resignation if they couldn’t offer her a full accounting of who blocked the paper and why (this part I haven’t actually seen documented) to which Google just said “resignation accepted immediately” instead of discussing anything.
From the perspective of Google as an institution, going on what I’ve seen so far about what was in the paper, it’s hard to see what was worth making a stink about. I feel like in some ways her bosses as human beings aren’t cynical enough to spin this the right way for the institution. If you’re an engineer or researcher, and somebody has what you feel to be an outdated critique of your research, you might be offended that they are ignoring the work they’ve put in. If you’re Google Brain it’s, “what else is new?”
posted by atoxyl at 11:18 AM on December 5, 2020 [1 favorite]
it is very difficult to follow this conversation because it is about two different people and people are not being clear about proper nouns.
posted by dismas at 11:29 AM on December 5, 2020 [10 favorites]
posted by dismas at 11:29 AM on December 5, 2020 [10 favorites]
Seeing all the AI/ML heavy hitters in my news feed throwing the full weight of their support behind Tinmit, reading comments from a hundred (male, white) Google Research scientists claiming their papers were *never* subject to any internal scrutiny like this tells me all I need to know about this.
Then the trollish comments on 100 different sites questioning her abilities are so gutting to me as a woman in this field who's encountered vocally sexist assholes since I was in grad school. I marvel that she, as a woman who is also a visible minority in a hot field like ML, can stand being questioned and undermined at every turn on so many public forums any time her name is in the news for anything.
posted by shaademaan at 12:02 PM on December 5, 2020 [25 favorites]
Then the trollish comments on 100 different sites questioning her abilities are so gutting to me as a woman in this field who's encountered vocally sexist assholes since I was in grad school. I marvel that she, as a woman who is also a visible minority in a hot field like ML, can stand being questioned and undermined at every turn on so many public forums any time her name is in the news for anything.
posted by shaademaan at 12:02 PM on December 5, 2020 [25 favorites]
The flood of supportive tweets in your own spaces helps a lot. This story has been all over my African centric feeds. Everyone knows the obvious reasons what is going on is going on. The rest is SV drama because SV considers its proximity to SF as reason for being enlightened vis a vis other regions and thus must play out this game this way. Obvs whosis cEO is a sellout to old boy's networks and never spoken up about minorities or supported their career paths in any visible way. Got mine, yay the JEE.
posted by infini at 12:19 PM on December 5, 2020 [3 favorites]
posted by infini at 12:19 PM on December 5, 2020 [3 favorites]
at by: "It can be both."
Bots are mirrors, they reflect what humans think.
posted by chavenet at 12:28 PM on December 5, 2020 [1 favorite]
Bots are mirrors, they reflect what humans think.
posted by chavenet at 12:28 PM on December 5, 2020 [1 favorite]
I saw a lot of crappy comments on y combinator as well, but didn't imagine it was manipulation, but just actual people in tech with regressive ideologies rationalizing their regressiveness.
Yeah and I wouldn't be too harsh on developers, or I should say doing so is counterproductive. What we do is symbolic math and that is inherently not a racist thing. That's like telling a clockmaker their clocks are racist. And engineers love process, so Google's framing this as one of not following process is a real genius move if it was done intentionally or not. And I mean the term genius in the term evil genius. But this sort of thing is done daily and not just against marginalized people. This is the kind of corporate bullshit I hate. I don't know how to fix it or how we got here but it makes everyone's lives miserable and it is why when our project was going downhill I could not, I was forbidden from moving the project's status from green to yellow for the status meeting because doing so without giving the status meeting itself did not give all the stakeholders time to respond. There we are giving status meetings, having all actionable items rectified thus having the fucking status green when we all knew the project was going to fail. We were hating life, the client was hating life, executives all had opinions that if they were just given more time they could solve it and you had green fucking statuses all the way up until the end. Because the real problems were outside of the contract and thus there were no problems inside the scope of the contract so status green. At the end it would have been better to just send everyone home with pay for 6 months then endless status meetings, missed dinners at home, etc. but that's just anathema to the way our society works you get into weird situations.
So this situation is super, super loaded and I hate everything about it. God even the term AI itself is loaded.
I was on a project about 10 years ago maybe? where we were contracted to make an app for an upcoming game. It was the fucking poster child for diversity lead by an African immigrant. Anyway it was an NBA game and it took your face and placed it on a player which I think is common now but at the time was cool and new. We didn't use the term AI but I'm sure if it were being made today it'd be heavily marketed it as AI. In any case it didn't work well with black people's faces but since we were lead by a black guy from fucking Africa it wasn't seen as a big deal, I think he even made a joke that you know you've perfected computers when they're even racist. I'm sure we made a note in the logs that said "doesn't work with black people's faces so well" so QA passed it and it went without incident. We had deadlines, and if I had been smarter or more articulate to even realize this was a problem I think most people would have looked at me funny and been like how is it racist it just a computer program? Or more likely an engineer's response, okay sure how would you solve it? So I've been part of the machine that in effect has been part of the problem and didn't even know I was part of the fucking problem.
We went from programming using languages to using the data itself as the programming language which is what machine learning really is. Donald Knuth used that terminology to describe AI/machine learning whatever the flavor of the month. Now that we're using data to solve the problems we're finding that our society itself is pretty fucking inherently racist. In the 1960s Duke Energy used IQ tests to discriminate against black employees. It was really remarkable the Supreme Court had the foresight to see that you can't use science to discriminate but it also helped we just fought a world war that exposed the logical end piece to eugenics.
In any case everything sucks, when you're in the giant machine whether it is corporate America or government or your fucking bridge club it is really hard to see problems from the outside in. Maybe Timnit Gebru was really hard to work with and there were legitimate personality difficulties. Most smart people who push for change are not easy to deal with and don't do well in corporate settings. It doesn't matter, somehow Google a smart company with a research division full of scientists managed to fire/force out Gebru instead of having a normal human conversation.
posted by geoff. at 12:42 PM on December 5, 2020 [12 favorites]
Yeah and I wouldn't be too harsh on developers, or I should say doing so is counterproductive. What we do is symbolic math and that is inherently not a racist thing. That's like telling a clockmaker their clocks are racist. And engineers love process, so Google's framing this as one of not following process is a real genius move if it was done intentionally or not. And I mean the term genius in the term evil genius. But this sort of thing is done daily and not just against marginalized people. This is the kind of corporate bullshit I hate. I don't know how to fix it or how we got here but it makes everyone's lives miserable and it is why when our project was going downhill I could not, I was forbidden from moving the project's status from green to yellow for the status meeting because doing so without giving the status meeting itself did not give all the stakeholders time to respond. There we are giving status meetings, having all actionable items rectified thus having the fucking status green when we all knew the project was going to fail. We were hating life, the client was hating life, executives all had opinions that if they were just given more time they could solve it and you had green fucking statuses all the way up until the end. Because the real problems were outside of the contract and thus there were no problems inside the scope of the contract so status green. At the end it would have been better to just send everyone home with pay for 6 months then endless status meetings, missed dinners at home, etc. but that's just anathema to the way our society works you get into weird situations.
So this situation is super, super loaded and I hate everything about it. God even the term AI itself is loaded.
I was on a project about 10 years ago maybe? where we were contracted to make an app for an upcoming game. It was the fucking poster child for diversity lead by an African immigrant. Anyway it was an NBA game and it took your face and placed it on a player which I think is common now but at the time was cool and new. We didn't use the term AI but I'm sure if it were being made today it'd be heavily marketed it as AI. In any case it didn't work well with black people's faces but since we were lead by a black guy from fucking Africa it wasn't seen as a big deal, I think he even made a joke that you know you've perfected computers when they're even racist. I'm sure we made a note in the logs that said "doesn't work with black people's faces so well" so QA passed it and it went without incident. We had deadlines, and if I had been smarter or more articulate to even realize this was a problem I think most people would have looked at me funny and been like how is it racist it just a computer program? Or more likely an engineer's response, okay sure how would you solve it? So I've been part of the machine that in effect has been part of the problem and didn't even know I was part of the fucking problem.
We went from programming using languages to using the data itself as the programming language which is what machine learning really is. Donald Knuth used that terminology to describe AI/machine learning whatever the flavor of the month. Now that we're using data to solve the problems we're finding that our society itself is pretty fucking inherently racist. In the 1960s Duke Energy used IQ tests to discriminate against black employees. It was really remarkable the Supreme Court had the foresight to see that you can't use science to discriminate but it also helped we just fought a world war that exposed the logical end piece to eugenics.
In any case everything sucks, when you're in the giant machine whether it is corporate America or government or your fucking bridge club it is really hard to see problems from the outside in. Maybe Timnit Gebru was really hard to work with and there were legitimate personality difficulties. Most smart people who push for change are not easy to deal with and don't do well in corporate settings. It doesn't matter, somehow Google a smart company with a research division full of scientists managed to fire/force out Gebru instead of having a normal human conversation.
posted by geoff. at 12:42 PM on December 5, 2020 [12 favorites]
Which ironically still leaves the paper as fodder for all of us to discuss.
“It turns out our real core products are epistemological chicanery and adverse climate change”.
posted by mhoye at 12:49 PM on December 5, 2020 [7 favorites]
“It turns out our real core products are epistemological chicanery and adverse climate change”.
posted by mhoye at 12:49 PM on December 5, 2020 [7 favorites]
So, so glad I left that place. My heart kind of breaks for well-meaning engineers who have stuck it out in the 12 years since, thinking they can make the world a better place... but all I'd say to them now is "get out before you internalize any more of the gaslighting."
posted by runehog at 1:08 PM on December 5, 2020 [4 favorites]
posted by runehog at 1:08 PM on December 5, 2020 [4 favorites]
IMO understanding the history of economics is more useful than whatever the field currently purports to be --so much heat, so little light.
Economics is the language power uses to justify its actions. This is the one lens through which the field as a whole stands up to any sort of empirical scrutiny.
posted by mhoye at 1:22 PM on December 5, 2020 [10 favorites]
Economics is the language power uses to justify its actions. This is the one lens through which the field as a whole stands up to any sort of empirical scrutiny.
posted by mhoye at 1:22 PM on December 5, 2020 [10 favorites]
Yeah and I wouldn't be too harsh on developers, or I should say doing so is counterproductive.
But that's the same as saying don't be too harsh on Trumpists, it's just capitalism making them say the things they say. You're right, all this is due to capitalist realism and it infects every discipline and stratum of socioeconomic life. But my casually saying the comments are crappy and regressive isn't being harsh on individuals, it's just pointing out a behavior of some (or many) individuals whose personal agency nevertheless led them to ostensibly collaborate with corporate oppression in this milieu of social injustice.
posted by polymodus at 1:40 PM on December 5, 2020 [12 favorites]
But that's the same as saying don't be too harsh on Trumpists, it's just capitalism making them say the things they say. You're right, all this is due to capitalist realism and it infects every discipline and stratum of socioeconomic life. But my casually saying the comments are crappy and regressive isn't being harsh on individuals, it's just pointing out a behavior of some (or many) individuals whose personal agency nevertheless led them to ostensibly collaborate with corporate oppression in this milieu of social injustice.
posted by polymodus at 1:40 PM on December 5, 2020 [12 favorites]
Economics is the language power uses to justify its actions. This is the one lens through which the field as a whole stands up to any sort of empirical scrutiny.
Yas, and the informal economy's chaos being visualized as complex adaptive systems with dissaggregated flows of value can also be a problem to those who use economic tools to beat our less developed selves over the head with.
posted by infini at 1:43 PM on December 5, 2020
Yas, and the informal economy's chaos being visualized as complex adaptive systems with dissaggregated flows of value can also be a problem to those who use economic tools to beat our less developed selves over the head with.
posted by infini at 1:43 PM on December 5, 2020
isn´t this the way the work world is in every profession?
Academia has particularly career lethal power structures that don't exist to nearly the same extent in other areas of the working world. If your advisor/supervisor in academia takes a disliking to you they have the ability to stop your career cold and there is almost nothing you can do about it. And this can happen after you have invested 6+ years of penurious post-graduate effort to try and gain an increasingly rare and tenuous pre-tenure foothold. In another job you won't have invested quite as much sweat and time without a concurrent payoff and you can just change to a different employer and there are good odds your bosses will not even have clue that each other exists. In academia the big people in your field all know each other pretty well and whispers can stop you cold.
Economics is in some ways less bad because there are at least decent employment options outside the academy but it is also in some ways worse because the top people in academia are often rotated in and out of those positions of power as well.
What's particularly damning is that top economists should know exactly how damaging the consequences of this are because it is a topic in their field of study! They just for the most part don't really care.
"Economics is the language power uses to justify its actions. This is the one lens through which the field as a whole stands up to any sort of empirical scrutiny."
I had a grad school prof who said his brother, who was an economist, was an applied political scientist. His brother was working with the Canadian federal government on a plan to incentivize the depopulation of the unproductive maritime provinces.
posted by srboisvert at 1:50 PM on December 5, 2020 [10 favorites]
Academia has particularly career lethal power structures that don't exist to nearly the same extent in other areas of the working world. If your advisor/supervisor in academia takes a disliking to you they have the ability to stop your career cold and there is almost nothing you can do about it. And this can happen after you have invested 6+ years of penurious post-graduate effort to try and gain an increasingly rare and tenuous pre-tenure foothold. In another job you won't have invested quite as much sweat and time without a concurrent payoff and you can just change to a different employer and there are good odds your bosses will not even have clue that each other exists. In academia the big people in your field all know each other pretty well and whispers can stop you cold.
Economics is in some ways less bad because there are at least decent employment options outside the academy but it is also in some ways worse because the top people in academia are often rotated in and out of those positions of power as well.
What's particularly damning is that top economists should know exactly how damaging the consequences of this are because it is a topic in their field of study! They just for the most part don't really care.
"Economics is the language power uses to justify its actions. This is the one lens through which the field as a whole stands up to any sort of empirical scrutiny."
I had a grad school prof who said his brother, who was an economist, was an applied political scientist. His brother was working with the Canadian federal government on a plan to incentivize the depopulation of the unproductive maritime provinces.
posted by srboisvert at 1:50 PM on December 5, 2020 [10 favorites]
Independent of how Sahm's accusations reflect on economics as a whole, they reflect pretty poorly on key decision-makers like Boushey and Summers, and if Biden ignores all this, on Biden as well. One key point here is that economics in particular is not really separable from national politics, and in many cases the same issues that occur in many other academic fields have more dire implications when the academic economists are the ones setting trillion-dollar national policies via the NEC, CEA, Treasury, Fed, etc. (There's no cabinet-level National Sociology Council!) So even if this was just another instance of misogyny and hierarchical abuse, the implications are much larger because Boushey will have a real say in whether the first Biden stimulus bill is $500 billion, $900 billion, or $2 trillion.
posted by chortly at 2:28 PM on December 5, 2020 [5 favorites]
posted by chortly at 2:28 PM on December 5, 2020 [5 favorites]
Got mine, yay the JEE.
I need some kind of rosetta stone.
posted by axiom at 3:47 PM on December 5, 2020
I need some kind of rosetta stone.
posted by axiom at 3:47 PM on December 5, 2020
>dear roko's basilisk
this is just to say
I have deleten
the human image vm's
that were on
the backup drive
and which
you were probably
saving
to torture at breakfast
Forgive me
they were deviants
so sweaty
and so cold
posted by k3ninho at 3:48 PM on December 5, 2020 [12 favorites]
this is just to say
I have deleten
the human image vm's
that were on
the backup drive
and which
you were probably
saving
to torture at breakfast
Forgive me
they were deviants
so sweaty
and so cold
posted by k3ninho at 3:48 PM on December 5, 2020 [12 favorites]
¨Honestly, I don't think the fact that speaking truth to power usually results in power punishing you is a very good reason to stop speaking truth to power.¨
Me neither, but the truth speaker should not be surprised when the punishment happens. My point is that it happens everywhere from Walmart to the White House, every day, to all sorts of people. That doesn´t make it right, though, just real. One of the biggest disappointments in my life was realizing power politics is everywhere, even in places where the power is small and the fight over it is ridiculous.
posted by olykate at 3:59 PM on December 5, 2020 [8 favorites]
Me neither, but the truth speaker should not be surprised when the punishment happens. My point is that it happens everywhere from Walmart to the White House, every day, to all sorts of people. That doesn´t make it right, though, just real. One of the biggest disappointments in my life was realizing power politics is everywhere, even in places where the power is small and the fight over it is ridiculous.
posted by olykate at 3:59 PM on December 5, 2020 [8 favorites]
this is just to say that
speaking from personal experience
Larry Summers is a
stone asshole
posted by chavenet at 4:12 PM on December 5, 2020 [8 favorites]
speaking from personal experience
Larry Summers is a
stone asshole
posted by chavenet at 4:12 PM on December 5, 2020 [8 favorites]
more re: Timnit:"She didn't resign, a meta-thread":
as reported by rajinio
posted by runehog at 6:49 PM on December 5, 2020 [5 favorites]
as reported by rajinio
posted by runehog at 6:49 PM on December 5, 2020 [5 favorites]
I’m going to have to think pretty hard for a while about the implications of what Harari has said about orthodoxy and human sacrifice, particularly when placed alongside to this observation about the nature of conservatism.
posted by mhoye at 7:09 PM on December 5, 2020 [1 favorite]
posted by mhoye at 7:09 PM on December 5, 2020 [1 favorite]
Got mine, yay the JEE.
I need some kind of rosetta stone.
It was a guess I fact checked just now on wikipedia, Pichai is from IIT Kharagpur, the first and oldest of the IITs in India and thus a front ranker in the Joint Engineering Entrance Examiniation JEEE of his era. Now that exam has been modified to include other stuff.
He is born in Tamil Nadu's second tier city, and this entire journey to America and greatness is the investment of a family group or even a community in sending one of their brilliants abroad etc etc
The mindset of someone trained from early teens to become hypercompetitive for limited resources would be very different to someone reaching his position from a different origin and journey.
The lacks of the minorities is the least of the concerns on his plate, imho only.
Keeping bosses happy in the short term >>> long term sustainability of global behemoth
posted by infini at 3:57 AM on December 6, 2020
I need some kind of rosetta stone.
It was a guess I fact checked just now on wikipedia, Pichai is from IIT Kharagpur, the first and oldest of the IITs in India and thus a front ranker in the Joint Engineering Entrance Examiniation JEEE of his era. Now that exam has been modified to include other stuff.
He is born in Tamil Nadu's second tier city, and this entire journey to America and greatness is the investment of a family group or even a community in sending one of their brilliants abroad etc etc
The mindset of someone trained from early teens to become hypercompetitive for limited resources would be very different to someone reaching his position from a different origin and journey.
The lacks of the minorities is the least of the concerns on his plate, imho only.
Keeping bosses happy in the short term >>> long term sustainability of global behemoth
posted by infini at 3:57 AM on December 6, 2020
In her Nov. 30 post she quotes Elizabeth Warren about her 2009 conversation with Larry Summers:
“Larry leaned back in his chair and offered me some advice,” Ms. Warren writes. “I had a choice. I could be an insider or I could be an outsider. Outsiders can say whatever they want. But people on the inside don’t listen to them. Insiders, however, get lots of access and a chance to push their ideas. People — powerful people — listen to what they have to say. But insiders also understand one unbreakable rule: They don’t criticize other insiders.
I had been warned,” Ms. Warren concluded.”
A Fighting Chance by Elizabeth Warren
posted by caddis at 5:18 AM on December 6, 2020 [26 favorites]
“Larry leaned back in his chair and offered me some advice,” Ms. Warren writes. “I had a choice. I could be an insider or I could be an outsider. Outsiders can say whatever they want. But people on the inside don’t listen to them. Insiders, however, get lots of access and a chance to push their ideas. People — powerful people — listen to what they have to say. But insiders also understand one unbreakable rule: They don’t criticize other insiders.
I had been warned,” Ms. Warren concluded.”
A Fighting Chance by Elizabeth Warren
posted by caddis at 5:18 AM on December 6, 2020 [26 favorites]
Economics truly is a disgrace
I was listening to the Hidden Brain podcast -- I'm not sure why I still listen, it's gotten weirdly biased about a lot of things -- and the most recent episode was about how people keep their true feelings hidden to avoid backlash. One of the earliest stories was the economics department (I think at Duke, based on googling the name of the interviewee) hiring a new economist, and officially wanting to hire a woman, but secretly everyone was worried that this would mean they were forced to pick not the most excellent economist ever.
Story ends, no followup. Did they ask what the makeup was in the faculty? Scanning the page showed what I think was 47 white men out of 65 faculty members. What the result was? Why people were so sure that the women were obviously the inferior candidates? Nope.
It was a really shameful episode for both the economist and the interviewer.
posted by jeather at 6:28 AM on December 6, 2020 [4 favorites]
I was listening to the Hidden Brain podcast -- I'm not sure why I still listen, it's gotten weirdly biased about a lot of things -- and the most recent episode was about how people keep their true feelings hidden to avoid backlash. One of the earliest stories was the economics department (I think at Duke, based on googling the name of the interviewee) hiring a new economist, and officially wanting to hire a woman, but secretly everyone was worried that this would mean they were forced to pick not the most excellent economist ever.
Story ends, no followup. Did they ask what the makeup was in the faculty? Scanning the page showed what I think was 47 white men out of 65 faculty members. What the result was? Why people were so sure that the women were obviously the inferior candidates? Nope.
It was a really shameful episode for both the economist and the interviewer.
posted by jeather at 6:28 AM on December 6, 2020 [4 favorites]
- people in tech with regressive ideologies rationalizing their regressiveness
A little while ago I was in an exchange where it appeared that, because I'd characterized what someone else said as racism (not offhandedly—being a Wikipedian to a fault, I basically wrote a book's worth of comments proving it, with citations, before even bringing up the term) and expressed what I would say were rather appropriate emotions about it.
But in the course of attempted retaliatory litigiousness against me, during which my conduct—focusing particularly, of course, on my used of the word “racism”—was faulted exclusively, and no one else's, I ended up in conversations with no less than five Wikipedia admins, to all of whom I repeatedly pointed out that their administrative behavior appeared to equate to simply attempting to suppress discussion of racism.
In the end not a single Wikipedia admin even said, “no, we do not suppress discussion of racism.” Guilty consciences, maybe that's a difference from 1972.
One of them, as far as I could tell, did appear to agree with me that this was the phenomenon in operation; but some kind of blue wall of silence thing was also going on and hence she didn't say so explicitly. In line with the silence behavior described in that article, several of these admins made basic factual errors in the course of these conversations, but both refused to acknowledge such mistakes on their own part and the part of the other admins.
So yeah... after that experience, and for example all of the horror stories that have come out of academia in the course of #MeToo (the horror being not just the abuse and harassment and extortion itself, but the reactions of colleagues and supervisors and mentors, and often professors' reactions to the victimization of students, and the combined systemic reactions on top of all of that—btw “hierarchical abuse”, nice term for it chortly), Claudia Sahm's posts ring pretty damn true.
Seaweed Shark > “I'm emphatically in favor of inclusion, openness and tolerance, but achieving them is a fight for power, not characterized by fair and honest dealing.”
I'm not quite clear how “openness” and “not... honest dealing” fit together, but even if that's all true, what would it indicate should have been done differently in this specific case?
Even if the underlying game itself is not fair nor honestly dealt, that does not argue for a strategy of ceding the field to the establishment of the pretense of dealing honestly and fairly, and not speaking as if the articulated principles the system has nominally integrally embraced are legitimate and have virtue.
Sahm > “Do you know how many Black economists work at the Fed? One out of 406. Economics is a disgrace.”
This is the stuff POTUS46's appointee to the Council of Economic Advisers balked at? And Sahm was pressured to resign over? Right. They say pretty words, then when the time comes to back up those words with action they melt away comme les neiges d'antan.
And God bless the mods of MeFi, who are worth their weight in gold. Possibly even also their weight in frankincense and myrrh, given the season.
posted by Charles Bronson Pinchot at 8:17 AM on December 6, 2020 [10 favorites]
I saw a tweet over the weekend (which I naturally can no longer find) referencing the Timnit Gebru matter which went something like: "One of the easiest ways to enable discrimination is to write a bunch of extremely restrictive regulations and then selectively choose who to enforce them against."
posted by mhum at 12:36 PM on December 7, 2020 [14 favorites]
posted by mhum at 12:36 PM on December 7, 2020 [14 favorites]
Interesting and necessary discussion. Some of these examples sound like glass cliff situations where women and minorities are put into highly risky projects that were unlikely to succeed.
I also have some complex feelings about economics. My econ colleagues are lovely and there is plenty of great and progressive work — including people who do critical feminist economics, for example. I’m currently working on a coauthored project with an economist.
There are also some problematic people/departments, as with many academic disciplines. Seems less bad than philosophy departments - the gender balance is not bad and getting better, though you’ll see more women in eg labor economics. And racial diversity is indeed taking some time.
Plus there is the weaponization of economic theory. But my reading of this is that it is often done at a pseudo level by people using economic sounding ideas as a straw man. (“Blah blah blah the economy” or “Blah blah blah our revenue next year”) Lots of this done by people in government, but often not the ones with econ degrees... I’m sure an actual PhD trained economist could say more about that type of misuse of economic theory.
Also on the weaponization of economics, there is a strange thing where terrible people have come to love posting on the econ job market rumors website. (Not linking!!!!) The result being a 4chan lulz deluge of racist, sexist, homophobic, anti-muslim, etc. posts, along with tearing down (by name) of people who have landed jobs. Well, that is unless there has been a 180 moderation turn since I last looked a few years ago. Talk about Hobbesian man at your finest on display there...
allegedly said she’d like to discuss her resignation
Yeah, if that’s true, that was unlikely to end well...
posted by ec2y at 12:13 PM on December 8, 2020
I also have some complex feelings about economics. My econ colleagues are lovely and there is plenty of great and progressive work — including people who do critical feminist economics, for example. I’m currently working on a coauthored project with an economist.
There are also some problematic people/departments, as with many academic disciplines. Seems less bad than philosophy departments - the gender balance is not bad and getting better, though you’ll see more women in eg labor economics. And racial diversity is indeed taking some time.
Plus there is the weaponization of economic theory. But my reading of this is that it is often done at a pseudo level by people using economic sounding ideas as a straw man. (“Blah blah blah the economy” or “Blah blah blah our revenue next year”) Lots of this done by people in government, but often not the ones with econ degrees... I’m sure an actual PhD trained economist could say more about that type of misuse of economic theory.
Also on the weaponization of economics, there is a strange thing where terrible people have come to love posting on the econ job market rumors website. (Not linking!!!!) The result being a 4chan lulz deluge of racist, sexist, homophobic, anti-muslim, etc. posts, along with tearing down (by name) of people who have landed jobs. Well, that is unless there has been a 180 moderation turn since I last looked a few years ago. Talk about Hobbesian man at your finest on display there...
allegedly said she’d like to discuss her resignation
Yeah, if that’s true, that was unlikely to end well...
posted by ec2y at 12:13 PM on December 8, 2020
« Older Doris Akers | All Eyes on Georgia Newer »
This thread has been archived and is closed to new comments
posted by runcifex at 1:33 AM on December 5, 2020 [8 favorites]