Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google to pause Gemini image generation of people after issues

theverge.com

Post reply on HN
Print view View original

901–910 of 1001 posts

Re: Google to pause Gemini image generation of people after issues

#901

Post by dragonwriter »

layer8 wrote: post #733
khokhol wrote: post #688

Earlier quoted context omitted.

Indeed. What's striking to me about this fiasco is (aside from the obvious haste with which this thing was shoved into production) that apparently the only way these geniuses can think of to de-bias these systems - is to throw more bias at them. For such a supposedly revolutionary advancement.

That struck me as well. While the training data is biased in various ways (like media in general are), it should however also contain enough information for the AI to be able to judge reasonably well what a less biased reality-reflecting balance would be. For example, it should know that there are male nurses, black politicians, etc., and represent that appropriately. Black Nazi soldiers are so far out that it sheds…

You are literally saying that the training data, despite its bias, should somehow enable the AI to correct to acheive a different understanding than that bias, which is self-contradictory. You are literally suggesting that the data both omits and contains the same information.

Re: Google to pause Gemini image generation of people after issues

#902

Post by FooBarBizBazz »

sotasota wrote: post #759
duxup wrote: post #694

Earlier quoted context omitted.

It's very strange that this would leak into a product limitation to me. I played with Gemini for maybe 10 minutes and I could tell there was clearly some very strange ideas about DEI forced into the tool. It seemed there was a clear "hard coded" ratio of various racial / background required as far as the output it showed me. Or maybe more accurately it had to include specific backgrounds based on how people looked, a…

If you ever wondered what it was like to live during the beginning of the Cultural Revolution, well, we are living in the Western version of that right now. You don't speak out during the revolution for fear of being ostracized, fired, and forced into a struggle session where your character and reputation is publicly destroyed to send a clear message to everyone else. Shut Up Or Else. https://en.wikipedia.org/wiki/Go…

Nah, America is past "peak woke".

If it gets Trump 2.0 there might be a hyper-woke backlash though (or double backlash?).

But if there's another Biden term, things will be chill, culturally.

Also, Twitter is dead, and that's where the spirals got out of hand.

Re: Google to pause Gemini image generation of people after issues

#903

Post by FooBarBizBazz »

duxup wrote: post #783
sotasota wrote: post #759

Earlier quoted context omitted.

If you ever wondered what it was like to live during the beginning of the Cultural Revolution, well, we are living in the Western version of that right now. You don't speak out during the revolution for fear of being ostracized, fired, and forced into a struggle session where your character and reputation is publicly destroyed to send a clear message to everyone else. Shut Up Or Else. https://en.wikipedia.org/wiki/Go…

I think we're a ways from the severity of the Cultural Revolution.

At least they did something about the landlords.

Re: Google to pause Gemini image generation of people after issues

#904

Post by Sohcahtoa82 »

epivosism wrote: post #883

Dall-E 3 at least exposes the adjusted prompt. Here's an example of it; you can get it if you hit the API directly and look at revised_prompt. https://twitter.com/eb_french/status/1760763534127010074 At least they show it to us; and you can prepare or attempt to convince the GPT which interprets your prompt into not doing it quite as much (although the example above is where I failed; it seems like it's on to me, bec…

Yup, I run an IRC bot with a !dall-e trigger (with protections so people don't run up my OpenAI bill!), and when I get the response back, my bot gives the revised prompt in addition to the image result URL.

Lots of added diversity in the prompts.

Note that I call it added diversity, not forced diversity, because if I ask for a specific race, it will give it to me, and does not override or refuse the requests like Gemini does. If I ask for a crowd of people, I don't mind it changing it to be a racially diverse crowd.

Semi-related note, those revised prompts are also nice because if you create a very non-specific prompt and get something you didn't expect, it gives you insight as to why you got what you got. It added details to your prompt.

Re: Google to pause Gemini image generation of people after issues

#905

Post by mardifoufs »

magnoliakobus wrote: post #818
gedy wrote: post #753

Earlier quoted context omitted.

> How is it that this was somehow approved? If the tweets can be believed, Gemini's product lead (Jack Krawzczyk) is very, shall we say, "passionate" about this type of social justice belief. So would not be a surprise if he's in charge of this.

"very, shall we say, 'passionate'" meaning a relatively small amount of tweets include pretty mild admissions of reality and satirical criticism of a person who is objectively prejudiced. Examples: 1. Saying he hasn't experienced systemic racism as a white man and that it exists within the country. 2. Saying that discussion about systemic racism during Bidens inauguration was good. 3. Suggesting that some level of wh…

I love it when sarcastic white men on twitter tell me how just how much they know about DEI. Surely if there's one person that is going to not be over zealous or completely miss the point of inclusivity and diversity... it's a white dude tech bro like the guy we are talking about here! Always nice to know we minorities can count on such saviors to be saved from the perils of... generating pictures of white people.

Re: Google to pause Gemini image generation of people after issues

#906

Post by vitorgrs »

sega_sai wrote: post #266

That is certainly embarassing. But in the same time, I think it is a debate worth having. What corrections to the training dataset biases are acceptable. Is it acceptable to correct the answer to the query "Eminent scientist" from 95% men, 5% woment to 50%/50% or to the current ratio of men/women in science ? Should we correct the ratio of black to white people in answering a generic question to average across the gl…

There is AI bias. I think the most common scenario on Dall-e prior to the "fixes", was to ask for "doctors" and only get white people. Never black.

The thing is, you don't fix this by changing the user prompt. You fix this by "removing" the bias on your dataset!

Removing under quotes because of course you are just changing to another accepted bias.

Re: Google to pause Gemini image generation of people after issues

#907

Post by vitorgrs »

iteratethis wrote: post #885
sega_sai wrote: post #266

That is certainly embarassing. But in the same time, I think it is a debate worth having. What corrections to the training dataset biases are acceptable. Is it acceptable to correct the answer to the query "Eminent scientist" from 95% men, 5% woment to 50%/50% or to the current ratio of men/women in science ? Should we correct the ratio of black to white people in answering a generic question to average across the gl…

Why is bias a problem? When you prompt "business man" and it outputs a white man, this is quite probably reflective of representation in reality. Whether this overrepresentation is even a problem at all is debatable as the idea that every job, role or subgroup of people is perfectly diverse or that this even should be the goal isn't just ridiculous, it's demographically impossible. If you do have a problem with a spe…

Which reality? For sure "business man" being 100% white is not a reality for some people in Africa and Asia, don't you think?

Re: Google to pause Gemini image generation of people after issues

#908

Post by chadlifter69 »

randounho wrote: post #872
logicalmonster wrote: post #596

Personally speaking, this is a blaring neon warning sign of institutional rot within Google where shrieking concerns about DEI have surpassed a focus on quality results. Investors in Google (of which I am NOT one) should consider if this is the mark of a company on the upswing or downslide. If the focus of Google's technology is identity rather than reality, it is inevitable that they will be surpassed.

[flagged]

This argument could be used for anything.

"I love it when black people cope and seethe about having to use separate water fountains. Imagine what holocaust victims who died of thirst in auschwitz would say about having to use a separate water fountain."

Apologies to HN community for using a "swipe" here but idk how else to characterize how bad this argument is.

Re: Google to pause Gemini image generation of people after issues

#909

Post by seanmcdirmid »

sotasota wrote: post #759
duxup wrote: post #694

Earlier quoted context omitted.

It's very strange that this would leak into a product limitation to me. I played with Gemini for maybe 10 minutes and I could tell there was clearly some very strange ideas about DEI forced into the tool. It seemed there was a clear "hard coded" ratio of various racial / background required as far as the output it showed me. Or maybe more accurately it had to include specific backgrounds based on how people looked, a…

If you ever wondered what it was like to live during the beginning of the Cultural Revolution, well, we are living in the Western version of that right now. You don't speak out during the revolution for fear of being ostracized, fired, and forced into a struggle session where your character and reputation is publicly destroyed to send a clear message to everyone else. Shut Up Or Else. https://en.wikipedia.org/wiki/Go…

You do know that the same time China was having its Cultural Revolution, America and the west were having one as well? With all those baby boomer kids coming of age, 1969 wasn't a calm year anywhere in the world. In China, it meant communism and down with the old culture/elites. In the USA, it meant free love, drugs, and protesting against the Vietnam war.

But this, I don't see any comparison to Google suppressing what images could be generated with AI to any of what happened 55+ years ago.

Re: Google to pause Gemini image generation of people after issues

#910

Post by seanmcdirmid »

FooBarBizBazz wrote: post #902
sotasota wrote: post #759

Earlier quoted context omitted.

If you ever wondered what it was like to live during the beginning of the Cultural Revolution, well, we are living in the Western version of that right now. You don't speak out during the revolution for fear of being ostracized, fired, and forced into a struggle session where your character and reputation is publicly destroyed to send a clear message to everyone else. Shut Up Or Else. https://en.wikipedia.org/wiki/Go…

Nah, America is past "peak woke". If it gets Trump 2.0 there might be a hyper-woke backlash though (or double backlash?). But if there's another Biden term, things will be chill, culturally. Also, Twitter is dead, and that's where the spirals got out of hand.

Trump 1.0 triggered the initial woke wave in the first place (he was a catalyst, not a proponent). Trump 2.0 would rather trigger double woke, which will trigger its backlash like woke 1.0 triggered its own backlash.

Biden as president is boring, which is how I like it. But if you want to rile liberals up, nominate or elect Trump president again, it will definitely drive voter turnout if anything else.

Post reply on HN