• 0 posts
  • 12 comments
Joined 3 years ago
Cake day: June 30th, 2023
  • Most games, image processing libraries and other compute demanding libraries already use WASM in the browser. Have been for quite some time. It’s already widely used in every case where it provides a substantial benefit.

    But writing for GUI, which is what 99% of JS is used for, WASM provides little benefit. The speed bottleneck is mostly in DOM manipulation. And every web GUI framework uses 200 npm packages with something like webpack. Getting that to somehow work with your WASM code would be a nightmare if it’s even feasible.

  • It’s not a dark personality to misinterpret things. It’s just a poor grasp of the finer aspects of communication.

    For instance, Kathy Griffin made a controversial video holding a bloodied severed head of Trump, who was the US president at the time. That rightfully received a lot of backlash for being in poor taste. Making death threats against the president of the United States in a serious crime, however she was not arrested because it was correctly not seen as a threat by the law enforcement.

  • They can’t. Not without removing the part that makes reddit worthwhile, which are small dedicated communities.

    Sure, an AI can moderate shit like /pics /politics /funny /gaming. Those are basically just garbage article/repost feeds anyway.

    Can an AI moderate say… buildapcsales? Where you need a human moderator to verify if the deal is valid, add additional info to the flair and mark deals as expired? Technically yes, but it would require a specially trained AI, that knows how to scrape particular websites and avoid detection, which is not a great look for reddit and is also a ton of work.

    What about a TV show subreddit? There has been instances where episodes were leaked, and the moderators had to protect the community from spoilers. Can an AI do it? Again, if it’s manually trained on this new data it could. But are reddit employees going to illegally download leaked episodes to feed the to the algorithm?

    At that point, you’d be spending more time and money on babysitting each AI than if you just moderated manually.

  • The percentage of users that solely used the 3rd party apps to view and comment was relatively small.

    Reddit doesnt produce any content itself, so viewing and commenting in general isn’t particularly important. What matters more are valuable contributions. I would posit that 3rd party app users provided disproportionately more valuable content than the official app users.

    There is already an army of repost bots which aren’t going away. The bots don’t care about the health of the platform, so we can assume they are at maximum repost saturation.

    And reposts still require new content generation to make reposts. You can’t repost the same stale content perpetually.

    I don’t think reddit is going to just die. But it’s popularity and userbase can dwindle over time. Tumblr still exists, but it’s a shell of its former self.

  • What in the fresh fuck are you talking about?

    The overwhelmingly left leaning administration of Reddit want pro-Nazi spam accounts to convert the vehemently left leaning userbase of reddit into Republicans? This is some Alex Jones level shit.

    They just want to force people into using the site and their app so they can serve ads and harvest user data. They are trying to IPO so they are doing everything to make it seem profitable in the short term, they don’t care about the future of the site.