• 1 Post
  • 22 Comments
Joined 1 year ago
cake
Cake day: June 10th, 2023

help-circle


  • And it’s prioritizing short-term views over long-term stability. Sure, rushing the review gets you views now, but if companies realize that you’re not going to give their product a fair shake, they’ll stop sending you products. Then to review things, you’ll need to buy them yourself, further cutting into your profits. If Billet Labs ever makes another product, they’re not going to send LTT a review sample because of this whole shitshow. Other startups are now going to be hesitant to send LTT review products because 1. They may not get a fair review, and 2. They may not get their review product back.



  • Like, is the Billet Labs issue supposed to be sabotage or something? He’s shitting on it right from the beginning, uses the wrong card, installs it poorly, then refuses to retest because…it’ll cost him…like…$500?

    It’s like if I was reviewing a screwdriver, decided to use nails because I couldn’t find any screws, held the thing upside down, then bitched about how shitty it was. And when it’s pointed out that my review isn’t fair, refuse to retest because a box of screws is $8 at Home Depot and the screwdriver probably sucks anyway. And on top of that, just sell the screwdriver to someone else instead of giving it back.

    Does LMG have investments in a competitor or something? It is so willfully irresponsible that I almost want to claim conspiracy because I can’t believe that a company would make so many poor decisions by mistake. What is going on over there where a $500 reshoot that would ensure a fair and balanced review of the product is such an nonnegotiable prospect?










  • There needs to be a distinction between “I did my science badly” and “I knowingly published false information”. Wakefield’s paper linking vaccines and autism faked its data to imply a causal relationship between the two for the purposes of financial gain. You should absolutely be able to sue that guy if his paper damaged you in any way. Fuck 'em.

    On the other hand, if you publish a study in earnest, but that study is full of mistakes and comes to an incorrect conclusion, you should not be able to be sued. If the study is bad, it would be easy enough to publish a response pointing out flaws with the original study. This is especially true since so many papers are published with the caveat of “this requires future study to confirm”.

    In order to sue, you should be required to show some sort of malicious action behind the bad science, such as faked data.


  • This would be a dangerous precedent. If you disagree with scientific findings, you just conduct your own research to disprove the original study. If companies can sue researchers for publishing claims that damage them, it’ll just result in researchers withholding studies in fear a multibillion dollar corporation coming after them. Scientists need to be able to publish their research without fear of retribution.

    The only exception I would accept is if someone published knowingly false research, a la Andrew Wakefield.






  • Another thought is that they’re not trying to kill Mastodon, they’re trying to kill Twitter.

    Mastodon has a bit of a community already, so by implementing ActivityPub, Meta can make its platform seem bigger than it is by pulling in Mastodon content. Gives it another edge over Twitter.

    Best case scenario is Threads sees ActivityPub as just the cost of doing business. That way, even people who won’t use your platform are still interacting with it. Downside, people on your platform can leave for a federated alternative and not miss out on any content. Not sure if that downside makes up for the potential gains.

    I think the default approach needs to be defederate first unless Meta shows actual interest in developing the fediverse with good intentions. If Threads become the majority provider of content to the fediverse and then we defederate, we lose all that content. It could lead to Mastodon, Lemmy, and Kbin withering and dying as everyone goes where the content is.