Nvidia’s Director of AI research is publicly sharing names of her enemies | Hacker News Reader

Nvidia’s Director of AI research is publicly sharing names of her enemies

twitter.com

68 pointsjdkee35 comments

This should not be flagged. This is real and serious news in tech.

The assymetry of the corporate response to events like this is astounding. It is hard for me to believe that if she did not have a different (skin color / gender / political alignment) that this would go unpunished from Nvidia.

The fact that they did not immediately move to distance themselves from her is a scandal in itself.

It makes you wonder what AI's purposes are worked on to be. Nothing good if this is the standard.

However, I expect professional companies to allow employees to take down inappropriate content. It's tough to navigate stupid though, because they tend to drag everyone down. Social media giving voice to evil, is cause for concern.

On the contrary, I think she's positioned to be asking exactly the right questions about AI's purpose. Specifically, there's a lot of potential societal harm if we let people base too much action on the relatively simple correlative rules that modern ML sports.

Is that why she's doxxing people on her twitter block list, for the greater good?

Is she doxxing them? All I've seen is screenshots of her block list, which is data provided directly and immediately by Twitter. That's as much "doxxing" as calling someone by their common name in public.

I'll call out such antisocial behaviour from Trump as for this Nvidia director. You could say it's childish, but these people get off abusing whatever power they can get. It's disconcerting.

Having never met her, I won't pretend to know what she gets off on.

But I have seen Twitter brigading, and I don't have a good suggestion on how to deal with it. She's taking a novel approach of asking friends of followers to intercede. If it works, it could serve as a model for diffusing brigades, which would be nice to have.

(I wouldn't consider anything regarding Trump to be a useful model here. The power difference between any corporate director and a President is too significant).

I'll let my soul mate dispell such limiting thoughts:

https://www.twitter.com/desiraethinking

What an I supposed to see? I'm not going to read someone's entire Twitter timeline, especially someone I've never heard of.

Not really sure. It's confounding :)

Heh, It doesn't seem to be the case here, does it? Putting people on lists is not preventing societal harm, especially since nvidia is known to put pressure on reviewers.

Don't know her and don't know the people on the list, but that certainly isn't anywhere near the problems that AI might pose to society. The tweet is deleted though and it is obvious why that is the case.

This is shameful and unprofessional. And yet we are lectured about "microagressions" in the workplace and causes many to walk on eggshells - but this bullying is acceptable?

I reported her tweet.

Doing what she did may or many not be allowed, but it's not civilized behaviour.

It's highly unlikely that anyone vets their Twitter block list so well that they can publish it and stand behind it item by item.

I always do research on future coworkers whenever possible after an interview, and if I ran across a tweet like this I definitely would not accept a job offer. Maybe I've been lucky but I shudder to think of having that kind of toxicity in my life.

I don’t know who Pedro is, but Prof Anandkumar was known since her early academic career for turning data science problems into tensor decompositions. At the time I thought it was the most interesting research direction in the field. I met her at a conference and introduced her to Ben Hammer and Richard Sorcher. I didn’t know them either, but it worked out. Point is, it’s easy to feel attacked and lash out on the internet. Most people are pretty reasonable on the outernet though. It’s all about what the norms (algorithm) rewards. I wish tensor decomposition captured the interest of the tech world as much as social justice wars.

Why are people like this still employed

Our society usually rewards these types of people, especially in business. The issue is these types of people usually aren't quite so flagrant about how sociopathic they are... but that seems to be changing more and more.

The fact that people are flagrantly supporting her as a "victim" is truly troubling, it's clear she is both the aggressor and abuser. In what world is a PHD / professor publicly shaming and lamenting graduate students over seeming ideological differences not a fucked up situation? The replies to this tweet are golden, most "asking to be added to the list". Our boy @paulg even got in on the action.

The US laws on blacklisting are complex, fascinating, and they exist at the state level: https://www.nolo.com/legal-encyclopedia/free-books/employee-...

Pedro isn't exactly doing himself any favors in this situation.

https://twitter.com/pmddomingos/status/1338647370364452864

It occurred to me last night that it's probably only a matter of time before I end up on the list. Which made me feel sad because I only disagree with some of her tactics, but don't dislike her.

"Ars, to its credit, responded by pointing out that it is not good practice to boycott an entire site over political disagreement w/one person oh wait nm jk."

Probably because that would have been a terrible idea. People boycott an entire site over a political disagreement with one person all the time. You can disagree with that as a tactic, but it's an obvious tactic for putting economic pressure on a company.

"Listen people, he doesn't even work here" is a much better answer.

She's really pushing the limits of what she can get away with

Pushing the limits of what she can get away with on Twitter? Or get away with at Nvidia? (Or both? Both sounds good...)

Who's Pedro? Supposedly this list consists of his "fanboys".

Pedro Domingas, maybe? UW Roboticist

Apparently Pedro Domingos (UW CS prof) https://twitter.com/pmddomingos

See perhaps: https://news.ycombinator.com/item?id=25418707

which links to some Medium article maybe written by AnimaAnandkuma

and my response here: https://news.ycombinator.com/item?id=25419729

which would land me on her enemies list (as though I would be bothered to be in such esteemed company)

Twitter's model of having everyone equally publicly visible makes it difficult to distance oneself from abusers. In a sense, everyone is standing in the same room.

I don't miss my Twitter account much.

Stalin shows what people will do if people exhibiting similar behaviors get absolute power https://en.wikipedia.org/wiki/Great_Purge

Why should I care?

Perhaps you might like a brief refresher of cancel culture in technology. Exhibit one: Donglegate. https://www.wired.com/2013/03/richards-affair-and-misogyny-i...

"Cancel culture" is one of the dumbest fads this year. It's a new, vague name for things that already have a name: boycott, mobbing, witchhunting, censorship, etc., depending on who or how you are "canceling".

Judging by how I've seen it used in social media, it's a loud way to band together and harass someone before bothering to research facts or context, feel morally superior, and at the same time achieve nothing of consequence.

Better than keeping it secret.

Wow, this is horrible, it's like future-indicting people without presenting any evidence. The block list and the implications of "alt-right" influencer could possibly be actionable. Especially if someone fails to obtain employment based on this list.

Reply on news.ycombinator.com