• 0 Posts
  • 186 Comments
Joined 10 months ago
cake
Cake day: November 22nd, 2023

help-circle




  • Again, that’s literally what airlines and some hotels do. Based on how often you frequent the site and how often you search for flights for a specific date and location, they will change their prices for you specifically. The more interest you show and the closer it gets to that date, the higher the prices go. And your local pizza shop does this on a broad scale. They base their prices on demand - the more people willing to come in, the more they can raise their prices until they hit the threshold of what people are willing to pay.

    This is literally just taking targeted ads and applying it to pricing. A cross section of different values can identify you as an individual based on things like browsing habits and web searches, and companies can use that digital fingerprint to tailor online prices for you the same way that the airlines do. Even at a broad scale, they can tailor prices based on your income level, hobbies, and predicted price tolerance. Hell, with this concept they could even run fake sales at an individual level instead of site-wide like Amazon does during their Prime Day “sales.”

    This is one of the more irrational fears/predictions about the dynamic pricing infrastructure grocery stores want to implement - that they’ll start tailoring prices on things that you buy frequently or try to get you to buy extra with prices that look like a good deal. But it’s a lot more practical to do online than in a physical store.


  • Airline companies and hotels have been doing this for years. They track the location, time of year, and how frequently you’re looking to adjust their prices for you. You can sometimes get a different price for the exact same flight or hotel by using a private browser. You know those freezer doors with the display in them instead of a glass panel? Those have a camera in them as well to track which ads you spend the most time looking at so they can roll the most viewed ads more frequently. Some grocery stores are attempting to roll out digital pricing systems in their stores so that they can “dynamically change prices on items due to demand.”

    It’s only a small step from using an algorithm to create a profile on you to serve ads tailored to things that you’re interested in to companies using that same profile to “dynamically adjust prices due to demand.”



  • Because we’re talking pattern recognition levels of learning. At best, they’re the equivalent of parrots mimicking human speech. They take inputs and output data based on the statistical averages from their training sets - collaging pieces of their training into what they think is the right answer. And I use the word think here loosely, as this is the exact same process that the Gaussian blur tool in Photoshop uses.

    This matters in the context of the fact that these companies are trying to profit off of the output of these programs. If somebody with an eidetic memory is trying to sell pieces of works that they’ve consumed as their own - or even somebody copy-pasting bits from Clif Notes - then they should get in trouble; the same as these companies.

    Given A and B, we can understand C. But an LLM will only be able to give you AB, A(b), and B(a). And they’ve even been just spitting out A and B wholesale, proving that they retain their training data and will regurgitate the entirety of copyrighted material.



  • My nebulous understanding is that there’s no specific reason saying that women can’t be space marines, but I assume that it’s probably tradition since all the Primarchs are male. But I do remember hearing that one of the key people in the program that created the Primarchs was upset that they were all men because there was no reason that they couldn’t have been women other than “because the Emperor said so.”



  • The argument that these models learn in a way that’s similar to how humans do is absolutely false, and the idea that they discard their training data and produce new content is demonstrably incorrect. These models can and do regurgitate their training data, including copyrighted characters.

    And these things don’t learn styles, techniques, or concepts. They effectively learn statistical averages and patterns and collage them together. I’ve gotten to the point where I can guess what model of image generator was used based on the same repeated mistakes that they make every time. Take a look at any generated image, and you won’t be able to identify where a light source is because the shadows come from all different directions. These things don’t understand the concept of a shadow or lighting, they just know that statistically lighter pixels are followed by darker pixels of the same hue and that some places have collections of lighter pixels. I recently heard about an ai that scientists had trained to identify pictures of wolves that was working with incredible accuracy. When they went in to figure out how it was identifying wolves from dogs like huskies so well, they found that it wasn’t even looking at the wolves at all. 100% of the images of wolves in its training data had snowy backgrounds, so it was simply searching for concentrations of white pixels (and therefore snow) in the image to determine whether or not a picture was of wolves or not.






  • Except that’s not at all what’s happening here. We’re not talking about somebody we know personally with their permission or anything, we’re talking about an actress who got into pornography after having an emotional video go viral many years ago. Her dead name has nothing to do with that, and if you had even left out the fact that she’s trans, most people probably could’ve figured it out if they even bothered to go check out the original video. Abd if they didn’t? It wouldn’t make a difference in their knowledge of the subject. They’d still know that a woman who had an emotional video go viral years ago later became a porn actress. All her dead name adds to this is a possibly paparazzi style invasion of her privacy.


  • Does it add any useful context, though? I don’t know either name but I do remember the “Leave Britney alone” video being a thing (and the fact that the person in the video turned out to be right all along when the truth about Britney’s situation came out years later), so the added context that she’s trans and what her dead name was is meaningless to me other than to say, “She used to be a man. She’s a woman now, but she was a man before. Did you know that? That she was once a man? Because she was. Here’s what her name was.”

    As a trans woman, whose safety is so dependent on being able to go stealth in society, if I found out people were going around talking about me like this, I’d take a rusty icepick and make sure that they never think in words ever again. Lack of malicious intent doesn’t mean that no harm was caused. Your threat index is not universal.

    This could have very easily been left at “Trans woman X got into porn after her viral video Y” and there would be all the context needed to figure out who they were and what video they were in without using their dead name. Hell, you probably wouldn’t even have to point out that she’s trans for people to figure it out. Cis people treat the privacy of trans people the same way that the paparazzi treats the privacy of celebrities.