• 1 Post
  • 108 Comments
Joined 2 years ago
cake
Cake day: June 1st, 2023

help-circle




  • I very much agree that tyre dust is a problem, and that weight is a large issue.

    However, these kinds of caveats are routinely used to downplay the level of harm reduction that transitioning ICE cars to EVs would bring. Note how right-wing media basically uses this technique - mostly with the emissions associated with making EV batteries - to justify the continued use of ICE cars.

    The antidote is to require numbers for this type of claim.

    Fwiw, I don’t own any kind of car, I bike and take transit everywhere, and I’m broadly against cars on account of their outsized negative impact on society. I still believe EVs represent a necessary amount of harm reduction.








  • The concern is that the model doesn’t actually see the world in terms of distinct hexadecimals, but instead as tokens of variable size - you can see this using the tiktokenizer-webapp: enter some text and it will split it into the series of tokens the model actually will process.

    It’s not impossible for the model to work it out anyway, but it is a reason for this type of task to be a bit harder on LLMs.




  • Still, this does not quite address the issue of tokenization making it difficult for most models to accurately distinguish between the hexadecimals here.

    Having the model write code to solve an issue and then ask it to execute it is an established technique to circumvent this issue, but all of the model interfaces I know of with this capability are very explicit about when they are making use of this tool.