• 1 Post
  • 53 Comments
Joined 1 year ago
cake
Cake day: June 5th, 2023

help-circle

  • I mean, bosses input reading my heavy attack to suddenly turn their three move combo into a four move combo 50% of the time feels a bit lame. For instance the dancing lion suddenly going into the spray carousel after it would have exhausted its combo and rested otherwise. My main issue is on the inconsistency.

    Don’t get me wrong, that fight was really fun and I overcame it, but there are many such cases where it feels overtly like the game just threw in the extra attack as a “fuck you” while trying to learn the mechanics. There might be a subtle cue to the boss’s body language I didn’t see but there’s also the issue of the camera in encounters with large enemies.

    On the whole though, as frustrating as it may be at times, often there’s still an underlying pattern. The only fights I think are explicitly unfair are the ones with adds or multiple enemies that add a lot of uncertainty especially if some are off camera. The twin gargoyle fight comes to mind, as does the Godskin duo where you explicitly have to kill both around the same time or the other respawns.



  • >dinner is done
    >announce dinner
    >everyone shows up 10 minutes later

    Some of the food is cold because I didn’t cover it, but why should I? It would have been fine if they came to the table when I announced dinner.

    >next day
    >dinner is almost done
    >remembering yesterday, I decide to announce dinner 10 minutes early so that 10 minutes later is “on time”
    >everyone arrives to the table immediately, remembering that it was cold yesterday not wanting that to happen again
    >mfw




  • The device wouldn’t necessarily have to be constantly streaming the audio to a central server. If it’s capable of hearing wake up words like “Ok Google” it’s capable of listening for other phrases and having onboard processing to relay back the results much more compressed. Whether or not this is common practice is another matter, and yes the algorithms are scary good even without eavesdropping.







  • I had to resolder the switches on my G Pro Wireless. I absolutely love the mouse but they use pretty much the cheapest Omron switches they can, unless that’s changed recently.

    Was surprisingly not terrible to take apart and reassemble but you do have to pull off the adhesive-applied low-friction sliders at the bottom, so I would recommend buying replacement ones if you find yourself in that situation. Doesn’t help that I also had a $150 mousepad that goes with it that’s proprietary to specific Logitech gaming mice.

    For what it’s worth I only did it because I absolutely love the mouse and setup; I love the design and weight plus the convenience of a low-latency wireless mouse that literally never runs out of battery.



  • That’s fair. I think fundamentally a false positive/negative isn’t that much different. Pretty much all tests—especially those dealing with real world conditions—are heuristic, as are all LLMs by necessity of the design. Hallucination is a pretty specific term given to AI as an attempt to assign agency to a system that doesn’t actually have any (by implying it’s crazy and making stuff up instead of a black box with deterministic inputs and outputs spitting out something factually wrong but with a similar format to what is trained on). I feel like the nature of any tool where “you can’t trust this to be entirely accurate” should have an umbrella term that encompasses both types of providing inaccurate info under certain conditions.

    I suppose the difference is that AI is a lot more likely to randomly go off, whereas a blood test is likelier to provide repeated false positives for the same person with their unique biology? There’s also the fact that most medical tests represent a true/false dichotomy or lookup table, whereas an LLM is given the entire bounds of language.

    Would an AI clustering algorithm (say, K-means for instance) giving an inaccurate diagnosis be a false positive/negative or a hallucination? These models can be programmed on a sliding scale and I feel like there’s definitely an area where the line could get pretty blurry.


  • I mean, AI is used in fraud detection pretty often; when it hits a false positive (which happens frequently on a population-level basis), is that not a hallucination of some sort? Obviously LLMs can go off the rails much further because it’s readable text, but any machine learning model will occasionally spit out really bad guesses almost any person could have done better with. (To be fair, humans are highly capable of really bad guesses too).




  • I’m picturing a CEO in a board meeting suddenly turning heel on the fiscal policy and spinning it as “y’know it really is our duty as a fortune 500 American company to be paying our taxes because it pays dividends back into this country…” met by the credulous looks from everyone else in the room and mix of eager to pensive nods from other Executives.

    The C-suite trying to pull fast ones by Accounting hoping they won’t notice but getting their plots subtly foiled by the existence of so many tax loopholes.

    The only thing across the runtime that definitely 100% for sure won’t happen is the blanket lowest-pay raises, but it frequently gets brought up as a joke that’s only funny for the first few episodes and quickly becomes a tired cliche; the last resort suggestion that has to be tried and shot down every. single. time even though it becomes increasingly clear that it would be a net positive for everyone. In that sense, the humor is still there, but it’s in the cynical acknowledgement of the situation rather than the lowest paid employee actually being the butt of the joke.