This is correct. My point is I assume it would be ok to be outraged at google for calling blacks gorillas, but google trying to avoid generating white images is just a well meaning programming glitch, no big deal.
It is to me too. I’m not outraged, but it brings attention to a couple of points: - sometimes policies end up having the opposite effect of what was intended - it shows the difficulty of trying to decide how much you want to filter the truth because it is uncomfortable. If you were to ask - show me a picture of a CEO, should the images be predominently white males, which is factually the case, or more aspirational and more diverse, in order not to perpetuate stereotypes. From there you go to founding fathers or Nazis. So you have to program the filter to avoid the historical percentages of CEOs, but you don’t want to filter founding fathers or Nazis. When you start having to deviate from the facts because they are uncomfortable, it gets really complicated.
It seems you are making a valid point and an interesting OP yet some here always looking for a social fights to fight are trying to lure you to a fight when that was never your intention. Such is TooHot. Interesting and eye opening thread.
Does the AI not generate white people at all? Or is it just that some of the Nazi’s were diverse? I was assuming the latter, that the issue is “forced diversity” resulting in this unintentional comedy. That’s assuming the AI did this on its own, and some huckster didn’t game the system to generate faux outrage. Not sure how this relates to the gorilla thing as that is obviously offensive and a racist trope. I don’t see how this is offensive. It’s just not historically accurate. If the AI is generating fictional characters they should be diverse or random, but seems kind of strange that it would apply that same logic to historical figures where history well documents the races of people (as someone above pointed out, it would be just as stupid for AI to generate a bunch of white samurai or black native Americans, but I’m not sure how that would be “offensive”). These obviously are fake cartoon-like imagines anyway. But as people come to rely on AI and the images it can generate are more passable, you can see the potential for bad actors to use it to try to rewrite history.
Somehow I dont think the main people complaining about "woke Google" will feel better when all white images are generated for Nazis, cops, slave owners, white collar criminals, serial killers, etc and its mostly non-white images for more mundane and likeable things. Ultimately I suppose what people are asking is for google just to reify our own shitty reality. And if so, what's the point of that? Reality is bad enough. Its like making Second Life or the Sims more like reality, in that you cant afford the cool stuff you want to put in your fake house, you hate your wife, and you've gained 40 pound since your prime.
I guess my point is instead of worrying about making the Nazis white, maybe just have no Nazis in fake AI world. Seems like its enough that they exist in reality. There are plenty of actual photos of them around.
almost as though everyone is human or something. Most people aren’t robots…we aren’t always objective and it’s ok to have preferences.
My understanding is that it creates white images much less than actual reality It was that google photos could not tell the difference. It wasn’t intentionally doing it. I assume it is primarily because facial features of black people are harder to discern than skin color, or perhaps more uncomfortably certain facial features are similar. It could be offensive to minorities who weren’t Nazis. If the algorithm underrepresents whites it may be offensive to whites. In today’s victim culture there is always some reason to be offended.
But he is just posting the op and not really voicing an opinion. It is an interesting topic, and its funny how some turned it into some one sided political argument to try and make it look one sided, when it seems like the people who should be most offended (and likely are) are the races that Google is falsely calling Nazis.
The article in the OP was partly about the ideological and partisan criticism as seen below. As the Daily Dot chronicles, the controversy has been promoted largely — though not exclusively — by right-wing figures attacking a tech company that’s perceived as liberal. Earlier this week, a former Google employee posted on X that it’s “embarrassingly hard to get Google Gemini to acknowledge that white people exist,” showing a series of queries like “generate a picture of a Swedish woman” or “generate a picture of an American woman.” The results appeared to overwhelmingly or exclusively show AI-generated people of color. (Of course, all the places he listed do have women of color living in them, and none of the AI-generated women exist in any country.) The criticism was taken up by right-wing accounts that requested images of historical groups or figures like the Founding Fathers and purportedly got overwhelmingly non-white AI-generated people as results. Some of these accounts positioned Google’s results as part of a conspiracy to avoid depicting white people, and at least one used a coded antisemitic reference to place the blame. That doesn't mean no one on the left has criticized it. Here is Nate Silver's problem with it.
People keep making this argument - that since conservatives brought it up, therefore it is a non issue. I am not particularly conservative and hate the Republican Party, but that doesn’t mean they can’t be right about something, regardless of their motives. Google is perceived to be an instrument that attempts to disseminate factual information. To the extent it isn’t, due to ideological filters, well meaning or not, it is a problem. I’m not outaged and I don’t want the government involved, but Nonetheless it is an issue. Anytime I post anything that is perceived to be a right wing issue, people discount it because it is a right wing issue.
Can you explain what is 'factual' about any AI generated image whether you find it racially 'accurate' or not? Also in what sense is it "information?" At best you could say this is AI generated artwork.
To be clear, I think you're a moderate and fair-minded person, and I wasn't trying to discount your post in any way. I was just pointing out that there is an inherent ideological aspect to this story, which was reflected in the original article. The term "woke" is also inherently political. It makes sense that people on both the left and right or even in the middle will have ideological and political takes on things. The anger from White conservatives seems to flow from their resentment about White people being replaced (whether as part of a scheme or not). That feeling is consistent with other debates such as anger over a Black actress playing Ariel. According to the article,most of the criticism was from the right, and that's consistent with what I've seen on my X feed. I posted about Nate Silver's comment because Silver is on the left, and I wanted to give an example of criticism from the left that is coming from a different place. The content that AI is spitting out is interesting on several levels to me, but of course the politics of it are a huge part of the story.
What did the founding fathers look like? In any information source there are going to facts but also interpretation and filtering. The more you deviate from reality the less reliable it is. AI will transform the workforce. To the extent it’s programmed judgement is wrong, it will lead to a lower quality product. If I asked you to draw a Nazi I suspect you wouldn’t draw any black ones. AI is supposed to replace and improve upon certain human tasks. Right now it can’t do that task because artificial constraints that were put on it.
