News for hackers. Stuff that matters.
Re:Imagen, a text-to-image diffusion model by josho on Monday May 23, 2022 @05:39PM (#31485100)
Translation: AI has the potential to transform society. When we release this model to the public it will be used in ways we haven’t anticipated. We know the model has bias and we need more time to consider releasing this to the public out of concerns that this transformative technology further perpetuate mistakes that we’ve made in our recent past.
Re:Imagen, a text-to-image diffusion model by curiousgal on Monday May 23, 2022 @05:44PM (#31485148)
> it will be used in ways we haven’t anticipated

Oh yeah, as a woman who grew up in a Third World country, how an AI model generates images would have deeply affected my daily struggles! /s

It's kinda insulting that they think that this would be insulting. Like "Oh no I asked the model to draw a doctor and it drew a male doctor, I guess there's no point in me pursuing medical studies" ...

Re:Imagen, a text-to-image diffusion model by colinmhayes on Monday May 23, 2022 @06:07PM (#31485403)
Yes actually, subconscious bias due to historical prejudice does have a large effect on society. Obviously there are things with much larger effects, that doesn't mean that this doesn't exist.

> Oh no I asked the model to draw a doctor and it drew a male doctor, I guess there's no point in me pursuing medical studies

If you don't think this is a real thing that happens to children you're not thinking especially hard. It doesn't have to be common to be real.

Re:Imagen, a text-to-image diffusion model by paisawalla on Monday May 23, 2022 @07:01PM (#31485912)
> subconscious bias due to historical prejudice does have a large effect on society.

The quality of the evidence for this, as with almost all social science and much of psychology, is extremely low bordering on just certified opinions. I would love to understand why you think otherwise.

> Obviously there are things with much larger effects, that doesn't mean that this doesn't exist.

What a hedge. How should we estimate the size of this effect, so that we can accurately measure whether/when the self-appointed hall monitors are doing more harm than good?

Re:Imagen, a text-to-image diffusion model by astrange on Monday May 23, 2022 @09:29PM (#31487005)
> Yes actually, subconscious bias due to historical prejudice does have a large effect on society.

The evidence for implicit bias is pretty weak and IIRC is better explained by people having explicit bias but lying about it when asked.

(Note: this is even worse.)

Re:Imagen, a text-to-image diffusion model by curiousgal on Monday May 23, 2022 @06:11PM (#31485458)
> If you don't think this is a real thing that happens to children you're not thinking especially hard

I believe that's where parenting comes in. Maybe I'm too cynical but I think that the parents' job is to undo all of the harm done by society and instill in their children the "correct" values.

Re:Imagen, a text-to-image diffusion model by colinmhayes on Monday May 23, 2022 @06:18PM (#31485537)
I'd say you're right. Unfortunately many people are raised by bad parents. Should these researchers accept that their work may perpetuate stereotypes that harm those that most need help? I can see why they wouldn't want that.
Re:Imagen, a text-to-image diffusion model by holmesworcester on Monday May 23, 2022 @06:49PM (#31485833)
> I think that the parents' job is to undo all of the harm done by society and instill in their children the "correct" values.

Far from being too cynical, this is too optimistic.

The vast majority of parents try to instill the value "do not use heroin." And yet society manages to do that harm on a large scale. There are other examples.

Re:Imagen, a text-to-image diffusion model by cgrealy on Monday May 23, 2022 @07:20PM (#31486069)
Isn't that putting an undue load on parents?

It seems extremely unfair that parents of young black men should have to work extra hard to tell their kids they're not destined to be criminals. Hell, it's not fair on parents of blonde girls to tell their kids they don't have to be just dumb and pretty.

(note: I am deliberately picking bad stereotypes that are pervasive in our culture... I am not in any way suggesting those are true.)

Re:Imagen, a text-to-image diffusion model by boppo1 on Monday May 23, 2022 @05:58PM (#31485310)
I don't think the concern over offense is actually about you. There's a metagame here which is that if it could potentially offend you (third-world-originated-woman), then there's a brand-image liability for the company. I don't think they care about you, I think they care about not being hit on as "the company that algorithmically identifies black people as gorillas".
Re:Imagen, a text-to-image diffusion model by pxmpxm on Monday May 23, 2022 @05:59PM (#31485323)
Postmodernism is what postmodernism does.
Re:Imagen, a text-to-image diffusion model by contingencies on Monday May 23, 2022 @06:16PM (#31485510)
Re:Imagen, a text-to-image diffusion model by pxmpxm on Monday May 23, 2022 @07:08PM (#31485960)
Ha! However different pxmpxm on github, I'm afraid.
Re:Imagen, a text-to-image diffusion model by contingencies on Monday May 23, 2022 @07:13PM (#31486016)
That's almost poetic. Watch them attempt to make sense of the situation.
Re:Imagen, a text-to-image diffusion model by renewiltord on Monday May 23, 2022 @06:46PM (#31485794)
It's not meant to prevent offence to you. It is meant to be a "good product" by the metrics of their creators. And quite simply, everyone here incapable of making the thing is unlikely to have an image of what a "good product" here is. More power to them for having a good vision of what they're building.