• 14 Posts
  • 64 Comments
Joined 1 year ago
cake
Cake day: July 9th, 2025

help-circle

  • Most users don’t care about any of the spying. I’d expect a competitor of similar size to do the same. If there’s any improvement competition could bring, it would be for the creators, giving them a larger slice of the ad revenue that otherwise goes to the owners of the platform. I guess that would mean better quality videos for users. Though, I don’t think allowing anyone to make boatloads of money by getting millions of people to sit around for hours and hours watching videos has been a positive for humanity. Incentivizing creators with more money will just make this worse.


  • At some point (maybe we’ve reached it), I fully expect that many of the top-ranked websites by Google will be AI-generated themselves. There’s a cottage industry (and has been a very long time) of websites engineered in such a way to game the algorithm. They target specific queries and load up the pages with ads and links to product pages on amazon which generate a kickback for the website. Google used to turn up pages written by passionate people, using the web of hyperlinks people created out of good will as a proxy for relevancy to a query. But when a measure becomes a target, it ceases to be a good measure. Google Search was so good that everyone started to use it for everything and then we got the cottage industry wanting to generate money from people just seeking information. Vague queries like “best dog food” or “how often to water plants” or “tortilla recipe” are not likely to turn up anything except crap and that’s pretty much a self-fulfilling prophesy. A top-ranked website for a popular query will bring so much traffic so as to overwhelm any site except the ones generating some money per page view. A site which doesn’t do that goes offline and then is deranked by google. And the sorts of people who stoop to the level of putting ads and trackers on their site and aggressively SEO’ing their way to the top are not likely to care about providing you with the highest quality information. It goes the other way too; a passionate hobbyist does not want their website to be top-ranked as the traffic would bring the server or its costs to its knees. Neither do those old-school web forums who don’t want hundreds of noobs coming in and ruining their fun time. I’d consider the possibility that google is ranking pages partly based on whether the site can tolerate and would welcome the load. A metric like that is almost certain to rank up the bad, ad-riddled sites. Google may have no choice in the matter. When a random person wants to know why their refrigerator smells bad, the only other person willing to spend their own resources answering their dumb question will be evil as fuck.

    It was poor English literature grads writing those pages only because those were the only people who could write well. Now a computer can do that job. Take a guess where it’s going to go. The people running these sites and prompting the AI will be the poorest people in the world with internet access, where the money from the kickbacks goes a lot further than it does in the West. The internet has no borders. You’ll be getting information from an AI prompted by Raj in India, claiming to be Dave from NYC.



  • That’s probably the way things are going. All of the decoration on everything you get in a store will be AI generated. Personally, I would like it if the packaging looked little more than like what I imagine 1940 military food rations did: plain tin cans and brown paper packaging with nothing but a description of the contents in plain black lettering. You could have a photo of the contents but that’s it. That might sound depressing, but is it more depressing than decoration meant to get you to buy, buy, buy? Alright, I guess the decoration is also for branding, but you can do that with black lettering too.




















  • Folder is better. On disk, a folder is a list of pairs of a name and number, mapping the items in the folder to their location on the disk i.e. it’s a directory. The days before computerization are before my time, but, as I understand it, library index cards worked this way. You might have a card for each author which listed the books by that author with the location of the index card on that book, and you might have directory cards on subjects or keywords too, and the card on a book might point to the location of the book on the shelf and the card for the author, etc., or something like that. It would be most confusing to call these directory cards “folders”. The computer does the same thing internally, but the user interface has hidden away any notion of directory. They’re logically folders. It’s only a directory if you’re writing a file system implementation.