Olives @olives@qoto.org

If you think this is the first time I've posted these links, it actually isn't, it's just that it's been a while since I have.

"When thinking about results, it is useful to consider things like the "scientific pile-on effect" (Ferguson, 2013)(Ferguson, 2020) which drive people to go looking for "links" between something like porn and something bad, no matter how weak or tenuous it might be. Factoring this in, you can eliminate a lot of false positives which otherwise don't have value in the real world."

"Ferguson, C. J. (2013). Soda and the scientific Pile-On Effect. TIME.com. https://ideas.time.com/2013/08/28/soda-and-the-scientific-pile-on-effect/

Ferguson, C. J. (2020). “13 Reasons Why” and Teen Suicide. Checkpoints. https://www.psychologytoday.com/intl/blog/checkpoints/202007/13-reasons-why-and-teen-suicide"

I should probably throw that citation in.

"We should also remember that correlation does not imply causation. One classic example of this which is taught in statistics and science is the "ice cream effect" where crime appears to rise with the number of ice cream sales (Peters, 2013). One compelling alternate theory which is pitched is that both ice cream sales, and crime, are associated with warmer weather."

I probably don't need to cite the "ice cream effect" but it might make interesting reading material.

"Dawson et al. (2019) fails to find a link between porn use and sexual aggression among adolescents which suggests that sexual aggression might not necessarily be an outcome of porn use among adolescents. This might further call into question whether disproportionate interventions are warranted here."

"Dawson, K., Tafro, A., & Štulhofer, A. (2019). Adolescent sexual aggressiveness and pornography use: A longitudinal assessment. Aggressive Behavior, 45(6), 587–597. https://doi.org/10.1002/ab.21854"

I'm still working on a new experimental post. As you can see, I'm adding more information as to how each bit of science is relevant to the post.

I don't have a lot of time to look at everything in the U.K. so I'll leave it at that.

Includes things like domestic violence, which I suppose has less of a Q vibe to it.

"The British police claim there are millions of sexual abusers."

Curiously, this claim from London police is unsourced and appears in a document which is all about asking the government for more money. It also makes up a tiny part of it and is an "estimate".

Have you considered the possibility that they're exaggerating to try to get more funding?

Instead of vague and intangible estimates, a more relevant statistic might be how many cases they have to process, how many are likely to lead to a judicial outcome, and whether they feel those are adequately resourced. After all, it's a funding pitch document.

They even take the time to complain about the media criticizing them, presumably this refers to articles calling them racist. Or when controversial "emergency" covid powers were misused by a cop to rape and kill a woman.

By contrast, an AI translator could be useful, because there really is a shortage of decent translators, and a lot of these tend to be fans.

This is like some EA crap where they try to sell the same game to you over and over and it's somehow progressively getting worse in each iteration. Shifting from "hey, let's build a real game" to "hey, let's build something which can appear in a 30 minute stream".

Imagine if for the 2011 remake of Hunter x Hunter, instead of an updated and retouched show, someone just made a bunch of generic looking characters like that.

It's really just a publisher opening up the "back catalogue" and trying to sell you crap based on a brand.

https://en.wikipedia.org/wiki/Hunter_×_Hunter
https://en.wikipedia.org/wiki/Hunter_×_Hunter_(2011_TV_series)

https://nichegamer.com/manga-publisher-is-remaking-classic-series-eiken-with-ai/ It flattens any unique charm it had and makes it look like generic crap.

A couple of years ago, an influencer made the take that "AI images" would end child abuse and the like. But, I really think they over-estimate the capabilities of this technology. A lot of what it generates is crap. Garbage. It might have a slight novelty to it but otherwise it's just not good.

For just about anything, it would be worse than just getting a human to create the same thing by creating it by hand.

"A number of art communities have even banned AI generated content."

I can't say I'm surprised. It can be very spammy and can crowd out original content. A lot of it is just not really interesting, and how many pieces of it are there which look pretty samey (or generic / tacky)?

It's not a problem unique to diffusion models either. When someone used to feed like a million anime pics into StyleGAN to train a model, there was also an issue where it would be "cool geekery" but would otherwise look boring.

https://reason.com/2024/05/30/the-illusion-of-financial-privacy/
"It's strange how quickly we have accepted the current state of financial surveillance as the norm. Just a few decades ago, withdrawing money didn't involve 20 questions about what we plan to use the money for, what we do for a living, and where we are from. Our daily transactions weren't handed over in bulk to countless third parties."
#FourthAmendment #privacy

https://www.psychologytoday.com/us/blog/the-modern-brain/201909/how-negative-news-distorts-our-thinking Another article about negativity in the media.

https://www.theatlantic.com/newsletters/archive/2023/03/negativity-bias-online-news-consumption/673499/ One article about the media's tendency towards negativity.

Some journalists might identify "AI" as a "spooky new tech", so they jump on virtually everything, minor, or not, that someone might / has done with it, and speak of it in apocalyptic tones. That interest in doing so in itself says a lot.

From what I remember, there was even a study showing that the media didn't used to be as negative.

I don't see how someone can simply ignore sensationalism being a thing.

It's one thing when the algorithm has a silly fail but quite a bit of it is like complaining to Ford about how someone uses their cars.

Someone can practically take anything, add the presence of "AI" to it, and pitch it as a unique "shocking story".