Profile pic is from Jason Box, depicting a projection of Arctic warming to the year 2100 based on current trends.

  • 0 Posts
  • 845 Comments
Joined 1 year ago
cake
Cake day: March 3rd, 2024

help-circle
  • Reported, or total? That’s the funny thing, people who live off tips already are giving themselves a bit of a tax break by fudging numbers, so this isn’t as big of a deal as it seems. And maybe I found the catch - to get people who have these jobs to report their full income, and then when it’s exposed they get more than they’ve claimed, come after them. Yeah, that’s a bit conspiratorial…but everything else seems to be aimed at the commons and how they are the problem.




  • Absolutely. My post was just about how kids begin with a DEI mindset and that gets removed from them. What you’re talking about is reinforcing the normal, especially for those kids in bad environments. The issue then becomes the schools not having the backing to fight the eventual angry parents and their lawyers because their child is being shown that their parents are wrong. Schools used to be able to push back, but they bend over at anything now for fear of lawsuits and/or funding cut and jobs lost.

    Education is one of the most critical careers we have, and yet it’s underpaid and attacked constantly for doing their job.










  • for what it’s worth, Mbin can see and interact with both Lemmy type communities as well as the Mastodon type of broadcasts. They are still two different parts, but within a single interface. Often I see things on the sidebar from them, usually dropped into “Random” as the algorithm doesn’t know what to put it in, and have the same thoughts as you. That it seems like it’s shouting out into nothingness. But…I could respond to the commentary, and it would bounce back to them. It’s just a different way to communicate, not as “permanent” as a discussion board format.





  • LLMs can be good at openings. Not because it is thinking through the rules or planning strategies, but because opening moves are likely in most general training data from various sources. It’s copying the most probable reaction to your move, based on lots of documentation. This can of course break down when you stray from a typical play style, as it has less to choose from in the options of probability, and only a few moves in there won’t be any more since there’s a huge number of possible moves.

    I.e., there’s no calculations involved. When you play a LLM at chess, you’re playing a list of common moves in history.

    An even simpler example would be to tell the LLM that its last move was illegal. Even knowing the rules you just told it, it will agree and take it back. This comes from being trained to give satisfying replies to a human prompt.