• 0 Posts
  • 348 Comments
Joined 2 years ago
cake
Cake day: July 6th, 2023

help-circle





  • QuaternionsRock@lemmy.worldtoJust Post@lemmy.worldNope. No.
    link
    fedilink
    arrow-up
    0
    arrow-down
    1
    ·
    8 months ago

    Out of curiosity - do you think your opinion will change once on-device (i.e., power efficient) AI becomes the norm?

    The capabilities and utility of contemporary LLMs are wildly overstated by many, but the claim that they are completely useless is dubious imo. Nothing they generate can be treated as fact (and shame on those who suggest you do), but I can say with certainty that it has made my life as an indie programmer much easier, and I know I’m not alone in that.




  • I can appreciate that contemporary neural networks are very different from organic intelligence, but consciousness is most definitely equivalent to a computer program. There are two things preventing us from reproducing it:

    1. We don’t know nearly enough about how the human mind (or any mind, really) actually works, and
    2. Our computers do not have the capacity to approximate consciousness with any meaningful degree of accuracy. Floating point representations of real numbers are not an issue (after all, you can always add more bits), but the sheer scale and complexity of the brain is a big one.

    Also, for what it’s worth, most organic neurons actually do use binary (“one bit”) activation, while artificial “neurons” use a real-valued activation function for a variety of reasons, the biggest two being that (a) training algorithms require differentiable models, and (b) binary activation functions do not yield a lot of information per neuron while requiring effectively the same amount of memory.



  • Fixed lidar sensors are not as reliable as it’s made out to be, unfortunately. Dome lidar systems like those found on Waymo vehicles are pretty good, but way more advanced (and expensive) than anything you’d find in consumer vehicles at the moment. The shadows of trees are enough to render basic lidar sensors useless, as they effectively produce an aperiodic square wave of infrared light (from the sun) that is frequently inseparable from the ToF emission signal. Sunsets are also sometimes enough to completely blind lidar sensors.

    None of this is to say that Tesla’s previous camera-only approach was a good idea, like at all. More data is always a good thing, so long as the system doesn’t rely on the data more than the data’s reliability permits. After all, cameras can be blinded by sunlight too. IMO radar is the best economical complementary sensor to cameras at the moment. Despite the comparatively low accuracy, they are very reliable in adverse conditions.




  • I mean, futures have an effective place in capitalist societies; the onion market is rather volatile because onion futures are banned (for hilarious reasons, I might add). The volatility makes is more difficult to run a sustainable onion farming operation.

    As for these water futures, while the concept does seem repulsive to me, I don’t really understand how they are responsible for anything here. If water futures were illegal, water suppliers would still be charging a huge markup today. Water companies sold undervalued futures years(?) ago, and they’re the only ones really losing out here. No skin off my back.

    (I think that everything I just said is stupid btw. Don’t take this as a defense of the “system”.)