• 0 Posts
  • 90 Comments
Joined 2 years ago
cake
Cake day: July 26th, 2023

help-circle
  • People don’t predict values to determine their answers to questions…

    Also, it’s called neural network, not because it works exactly like neurons but because it’s somewhat similar. They don’t “run on neural networks”, they’re called like that because it’s more than one regression model where information is being passed on from one to another, sort of like a chain of neurons, but not exactly. It’s just a different name for a transformer model.

    I don’t know enough to properly compare it to actual neurons, but at the very least, they seem to be significantly more deterministic and way way more complex.

    Literally, go to chatgpt and try to test its common reasoning. Then try to argue with it. Open a new chat and do the exact same questions and points. You’ll see exactly what I’m talking about.

    Alzheimer’s is an entirely different story, and no, it’s not stochastic. Seizures are stochastic, at least they look like that, which they may actually not be.


  • It’s not. It’s a math formula that predicts an output based on its parameters that it deduced from training data.

    Say you have following sets of data.

    1. Y = 3, X = 1
    2. Y = 4, X = 2
    3. Y = 5, X = 3

    We can calculate a regression model using those numbers to predict what Y would equal to if X was 4.

    I won’t go into much detail, but

    Y = 2 + 1x + e

    e in an ideal world = 0 (which it is, in this case), that’s our model’s error, which is typically set to be within 5% or 1% (at least in econometrics). b0 = 2, this is our model’s bias. And b1 = 1, this is our parameter that determines how much of an input X does when predicting Y.

    If x = 4, then

    Y = 2 + 1×4 + 0 = 6

    Our model just predicted that if X is 4, then Y is 6.

    In a nutshell, that’s what AI does, but instead of numbers, it’s tokens (think symbols, words, pixels), and the formula is much much more complex.

    This isn’t intelligence and not deduction. It’s only prediction. This is the reason why AI often fails at common sense. The error builds up, and you end up with nonsense, and since it’s not thinking, it will be just as confidently incorrect as it would be if it was correct.

    Companies calling it “AI” is pure marketing.


  • We teach the fundamentals

    Sure. They are, however, not the focus. At least that’s not how I’ve been taught in school. You’re not teaching kids how to prove the quadratic formula, do you? No, you teach them how to use it instead. The goal here is different.

    They only teach order of operations.

    Again, with the order of operations. It’s not a thing. I’ve given you two examples that don’t follow any.

    The constructivist learners…

    That’s kinda random, but sure?

    And many proofs of other rules…

    They all derive from each other. Even those fundamental properties are. For example, commutation is used to prove identity.

    But the order you apply operators does matter

    2+2-2 = 4-2 = 2+0 = 0

    2 operators, no order followed.

    If we take your example

    2+3×4 then it’s not an order of operation that plays the role here. You have no property that would allow for (2+3)×4 to be equal 2+3×4

    Look, 2+3×4 = 1+3×(2+2)+1 = 1+(6+6)+1 = 7+7 = 14

    Is that not correct?

    Notably you picked…

    It literally has subtraction and distribution. I thought you taught math, no?

    2-2 is 2 being, hear me out, subtracted from 2

    Same with 2×(2-2), I can distribute the value so it becomes 4-4

    No addition? Who cares, subtraction literally works the same, but in opposite direction. Same properties apply. Would you feel better if I wrote (2-2) as (1+1-2)? I think not.

    Also, can you explain how is that cherry-picking? You only need one equation that is solvable out of order to prove order of operation not existing. One is conclusive enough. If I give you two or more, it doesn’t add anything meaningful.


  • Yes we are

    Yes and no. You teach how to solve equations, but not the fundamentals (and if you do then kudos to you, as it’s not a trivial accomplishment). Fundamentals, most of the time, are taught in universities. It’s so much easier that way, but doesn’t mean it’s right. People call it math, which is fair enough, but it’s not really math in a sense that you don’t understand the underlying principles.

    Yes there is!

    Nope.

    There’s only commutation, association, distribution, and identity. It doesn’t matter in which order you apply any of those properties, the result will stay correct.

    2×2×(2-1)/2 = 2×(4-2)/2 = 1×(4-2) = 4-2 = 2

    As you can see, I didn’t follow any particular order and still got the correct result. Because no basic principle was broken.

    Or I could also go

    2×2×(2-1)/2 = 4×(2-1)/2 = 4×(1-0.5) = 4×0.5 = 2

    Same result. Completely different order, yet still correct.

    My response to the rest goes back to the aforementioned.


  • Wasn’t it a kickstarter product? I wouldn’t consider venture a pre-order, tbf.

    Pre-orders are reservations with pre-payment.

    Crowdfunding is, well, funding. You aren’t buying a product. You’re funding it, which comes with additional risks and benefits.

    Of course, there’s always a possibility that a product is being funded using pre-orders, which is financially irresponsible (norm varies from industry to industry). But you must be a moron to pre-order a product from a startup you know nothing about and expect not to get scammed. Outright buying their product would be risky enough.

    Take housing market. You’re pretty much always either pre-ordering or buying second-hand.




  • The “why” goes a little further than that.

    In actuality, it’s because of fundamental properties of operations

    • Commutation

    a + b = b + a

    a×b = b×a

    • Association

    (a + b) + c = a + (b + c)

    (a×b)×c = a×(b×c)

    • Identity

    a + 0 = a

    a×1 = a

    If you know that, then PEMDAS and such are useless because they’re derived from those properties but do not fully encompass them.

    Eg.

    3×2×(2+2) = 3×(4+4) = 12+12 = 24

    This is a correct solution that is improper if you’re strictly adhering to PEMDAS rule as I’ve done multiplication before parenthesis from right to left.

    I could even go completely out of order by doing 3×2×(2+2) = 2×(6+6) and it will still be correct





  • Mistic@lemmy.worldtomemes@lemmy.worldI mean... I don't see the problem?
    link
    fedilink
    arrow-up
    10
    arrow-down
    1
    ·
    edit-2
    2 months ago

    Equal sides in a triangle are only possible if the corners are equal. So, 60⁰ each.

    But its height cannot be half of base because of the same Pythagorean theorem

    (1,5)²+(1,5/2)²=2,8125

    sqrt(2,8125) ≈ 1,677, which is half of a diagonal

    So, we get 4 sides that are 1,5 in a parallelogram, but diagonals are 1,5 and 3,354, as opposed to both being 1,5 as shown on the picture

    TL;DR: Won’t work because Pythagorean theorem





  • Mistic@lemmy.worldtoTechnology@lemmy.world*Permanently Deleted*
    link
    fedilink
    English
    arrow-up
    10
    arrow-down
    1
    ·
    edit-2
    4 months ago

    Ads, complete lack of privacy, kremlinbots.

    It used to be the platform that did one thing good. Build communities. Not it does a lot of things, but all of them badly.

    Not to mention that their owners are the main reason why “YouTube works bad, because Google’s servers are degrading”. Basically, Roscomnadzor (an official government censorship service) has been making YouTube unusable for over a year now to make people switch to VK Video or Rutube (both owned by same entity) whilst pretending that it doesn’t. Did it work? According to them, yes. According to reality, however, the vast majority just switched to using VPN.


  • Ah, I see. Yes, you’re right, they do manufacture reference design cards. First time dealing with reference models, since those aren’t sold where I live, hence the confusion, haha

    Those will all be the same in terms of temperature and clock speeds. Build quality should only vary insignificantly, although I do not trust Gigabyte, due to 3000 series PCB issues and how they handled it, and ASUS due to their borderline scam customer support.

    You won’t be getting any warranty buying used, I don’t think. So, imo, just get the cheapest one. You should concern yourself more with the seller, and do make sure to thoroughly check everything after buying. Both physically and performance-wise.



  • Those aren’t reference models, the ones you named are called something like AIB or aftermarket cards. (Reference models are those made by GPU manufacturers)

    Main differences between all are temperatures, clock speeds, build quality, and price.

    They also have different “tiers” of cards using same GPU. Those of higher tier cost more, but have better coolers and higher clock speeds. The premium you’d be paying isn’t usually worth it, unless it’s a small amount.

    Personally, I prefer Sapphire, PowerColor, or XFX for AMD cards.