• 2 Posts
  • 2.46K Comments
Joined 2 years ago
cake
Cake day: July 15th, 2023

help-circle





  • Try Changing the width and Hight of the First Layer to smaller Values, no clue Why, but it Fixed this kinda issue before for me!

    reducing the width of the first layer seems counter intuitive to me. usually adhession is improved by using widths wider than the nozzle since the plastic has to get smooshiefied into the surface. I don’t know there’s much difference on the first layer 125%-200% of the nozzle width, but I find going over definitely helps. (I use 150% for stronger parts in general.)

    or at least that’s been my experience. nozzle height/z offset would be the first thing for me to check, but PLA shouldn’t be warping that bad, IMO, so there might be some thermal issues as well.








  • This is incorrect. I’m a teacher and CONSTANTLY use the analog clock for multiple reasons.

    It’s the clock of record. Doesn’t matter what YOUR particular clock says, the clock on the WALL is the time we all go on.
    

    Irrelevant. This function would be served regardless of what type of clock is on the wall. A digital wall clock would serve just as well, and likely be less expensive to maintain than the electromechanical clocks you presumably use. (something about obsolete technologies becoming increasingly more expensive to maintain.)

    Reaching into your pocket to pull out a phone and look, no matter how much you want to pretend it’s trivial, still takes SIGNIFICANTLY more effort and time than glancing at the wall. Those seconds add up.
    

    See the above reply.

    Momentum. Are you PERSONALLY going to provide the billions of dollars in funding to replace every analog clock in a public space with digital ones?
    

    an irrelevant red herring with a false dichotomy. Clocks in public places are mostly installed by the people who maintain those public places. We all pay taxes to keep those places up. I don’t need to personally fund such a project and you know it.

    This is not like learning to write cursive - reading an analog clock is a trivial skill that should not take longer than a day for anyone to master.

    It may be trivial, and there may be some benefit besides learning to read said clock. But like cursive, it’s an irrelevant skill that generally won’t be used outside of class. That’s how it’s like cursive. Or, if you prefer, using a slide rule. as a teacher, I am, however, sure you understand that you only have so many hours of instruction available. A day spent on this, is a day not spent on something else. You might have arguments for why this is more important than that something else, but its still an obsolete technology that, like the slide rule and fountain pens, is going away. Nostalgia is not a good enough reason to keep it around.

    btw, my middle school had all digital wall clocks back in the mid 90’s, and probably had them for quite some time before i was there.


  • or maybe they should invest in digital clocks rather than continuing to use an archaic, obsolete technology.

    Yes. You read that right.

    Analog clocks are obsolete.

    Same with the fountain pen and dip pens and calligraphy (and the cursive writing styles that relied on them,) carrier pigeons and the telegraph. not to mention all sorts of other technologies that are of only passing interest.

    You probably don’t know how to read a sundial, or to locate yourself on a map using a magnetic compass.

    While there’s some esoteric value in such skills, the skills themselves are obsolete and useless to modern life. We’re not preparing them for the past. we’re preparing them for the future.

    Guaranteed the teachers aren’t using the analog clocks if they don’t have to either.

    (edit, I can’t find the full article, everything points back to that atlantic article. The gyst that they cropped out so crudely is two fold: the point of cursive was to minimize blotting caused by lifting and setting the pen across each letter when using a pen with a nib. It required a relatively light hand when writing, so as to glide over the page and not dig itn. Ball point pens, on the other hand, transfer ink differently- you’re rolling a ball over the surface- and they require significantly more pressure. the fluid motions of cursive writing cramp the hand sooner compared to print, if you’re writing with a ball point.)


  • FuglyDuck@lemmy.worldtopolitics @lemmy.worldWho Controls AI Exactly?
    link
    fedilink
    English
    arrow-up
    2
    arrow-down
    1
    ·
    7 days ago

    This is incorrect. Generative image models don’t contain databases of artwork. If they did, they would be the most amazing fucking compression technology, ever. … snip… The training is mostly a linear process. So the images never really get loaded into an database, they just get read along with their metadata into a GPU where it performs some Machine Learning stuff to generate some arrays of floating point values. Those values ultimately will end up in the model file.

    Where does it get read from? a database, right? yeah. that’s called a database. It may not be a large massive repository of art to rival the Vatican’s secret collection, but it is a database of digital art.

    as for it being complex… yeah. that’s why I kept it simple and glossed over all the complex stuff that’s not really, you know. relevant to the question of who owns it.


  • the people that own it.

    Keep in mind what we’re calling “AI” isn’t artificial general intelligence (C.F. Kryten, Data or R2D2). the most visible AI is a Learned Language Model- basically a predictive algorithm that goes through it’s training material and says “99% of the time, when someone says ‘69’, people respond ‘nice’, therefore, when people say ‘69’ I should respond with ‘nice’.”

    Or, with AI image gen, it knows that when some one asks it for an image of a hand holding a pencil, it looks at all the artwork in it’s training database and says, “this collection of pixels is probably what they want”.

    But the models don’t know why 69 is nice, nor what a hand is. It just spits out the proper response based on statistical probability.

    The thing is that the ‘proper’ response can be weighted by giving priority to certain responses- or rejecting certain responses- based on whatever motives the owner has. Take Grok as an example, and it’s blatant framing of Musk as the Greatest Man who Ever Lived™, but whoever weighted those responses failed to consider what happens when you ask if Musk is the best nazis or whatevers. You’ll notice those responses suddenly changed after people started figuring out how to game the prompts to get them.

    AI chatbots are the mouthpiece of whoever owns it… and it gives a level of sophistication that we’ve never seen before in the billionaire’s attempts to manipulate us.