• 9 Posts
  • 329 Comments
Joined 3 years ago
cake
Cake day: June 14th, 2023

help-circle








  • Absolutely. The quality drop on discussions was wild. I used to spend hours a week discussing on reddit. Now despite still using it to keep up to date on various topics, I rarely comment.

    The post quality dropped quite a bit as well at that time. But that has always been a steady “eternal September” type of decline so it’s harder to isolate the cause.

    Reddit will become less and less useful to the AI giants for training if it doesn’t contain new and ongoing deep communication like it once did.




  • Stack Overflow has been nearly useless as anything other than an archive of old questions for a long time. It’s a case study in how gamification can backfire.

    For me it was over was before the peak as shown. The gamification that got it started in the first place eventually led to it becoming almost unusable for its intended purpose. Instead it functioned as more of an online-only augmented reality game where people worked to moderate away new questions as efficiently as possible rather than actually answer questions.



  • They were initially based on research into prediction markets. However that research, while interesting, certainly wasn’t able to …predict… what would happen when highly funded commercial enterprises with massive marketing budgets started pushing them right in the middle of an already growing gambling epidemic.

    So, yes, they are now fully gambling apps with little predictive value.

    One of the great ironies of these apps is that people are actively demanding they act more like gambling and less like prediction markets. For example, prediction markets require insiders to participate to function effectively. But the people using these apps, and people who hate these apps alike want to ban insiders from using them to keep them “fair”. This pushes the apps even further from prediction and toward gambling.







  • This article really misses the point. I fundamentally disagree with the conclusion and think it’s actively harmful to neurodivergent people.

    They’re criticizing academic use of the words “neurodivergent” and “neurodiversity” and effectively demanding that all academic work rigorously categorize people into pre-existing diagnoses. They point to disagreement on what terms mean as a reason to not use the terms. But that’s precisely why we need to use the terms, why we need academics studying these things in context.

    Remember, it was only 13 years ago that Aspbergers and PPD-NOS were grouped in with Autism Spectrum Disorder. And ADHD & Autism were finally allowed as co-diagnoses. Yes that’s right, before 2013 you could not be diagnosed with both ADHD & Autism!

    Ultimately, here’s the problem. We weren’t “done” figuring out neurodivergent diagnoses prior to 2013, we weren’t done in 2013, and we aren’t done now. Neurodivergence isn’t as clear and as clean as these authors seem to think. The existing diagnoses can be very helpful to people, but they also fail a lot of people, and can lead to under- or mis-diagnosis. We need sociologists, neuroscientists, psychologists, and other scholars studying this from many angles.

    Maybe one day we’ll get to the point where we don’t need medical diagnoses for neurodivergent people of all ages to have their needs met, but I don’t think we get closer to that day with rigid categorization and unchanging definitions.