• 2 Posts
  • 61 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle

  • That argument is a very short (not very detailed) way of surmising the current issue with our world as a whole.

    Don’t like how cars have taken over the world, are the reason cities are hard to live in for low income families, and cause massive amount of climate damage? You can thank the 1% for that.

    Frustrated with how you don’t really own anything, your digital “property” can be taken away from you at a moments notice, and that everything you enjoy gets stuffed with schemes to make more money off of vulnerable people? You can thank the 1% for that.

    Angered that health care costs truly absurd amounts, that medicine is sold to the consumer with a 10,000% mark up, or that a single accident that was not their fault could land someone in debt for life? You can thank the 1% for that.

    A disturbing amount of things that are not good for our planet, keep the poor people poor, and generate inferior products/experiences is directly because of the insane power that the rich hold over our worlds systems.

    “It helps someone I don’t like because they are richer than me” is actually a wonderful definition of harm.

    Reddit used to be an amazing place of community and content that you couldn’t find anywhere on the internet. Then in the pursuit of money and the power that the 1% have Reddit (the company) started implementing practices that actively made the experience worse for the user, violated a person’s ownership of their content, and removed choice just like authoritative/dictatorship governments do.

    It feels to most people that there is nearly nothing that can be done about it. So when a person has the opportunity to directly go against the rich caste in our world they will take that opportunity immediately.

    I recommend taking a hard look at the things that concern you with our world, or cause you pain/annoyance/discomfort and try and learn WHY the issue is the way it is. The majority of the time is because some rich person/group of people (I’m looking at you lobby groups) has an obscene amount of power compared to all of the people affected.

    Lastly there is a reason that “Tax the rich/Eat the rich” is the rally cry of generations.

    It’s because the rich cause us harm.









  • Preservation only but not likely any better than a linguistic historian.

    But it gets tricky because LLMs only function on HUGE sets of data. LLMs are nothing more than complicated probability engines. Give it the question “What color is the sky?” and the math extracted from the massive databases that it has says the highest probability answer is “Blue”. It doesn’t actually KNOW the answer it just knows the probabilities of different words.

    Without large amounts of data on the dying language current gen LLM’s won’t be accurate or able to generate reliable answers. Shoot… LLMs can barely generate reliable answers with the massive datasets they currently have.

    I strongly recommend anyone even remotely interested in LLMs to read this interactive article:

    https://ig.ft.com/generative-ai/


  • If I could figure out how to move all my favorites and playlists AND continue discovering obscure music from around the world with ease, I would replace Spotify.

    I can’t think of another way for me to discover something like Indian Metal, all female Cuban acapella group, or power ambient deep in the middle of nowhere farmland NY. It’s not like those are going to be played on the radio. But I can type a random combo of letters and numbers into Spotify and start a radio based on the first band I don’t recognize. Let the discovery commence.



  • This is a pet peeve of mine right up there with the never ending stream of people calling machine learning AI. We do not have any real kind of AI at all at the moment but I digress.

    LLM is literally just a probability engine. LLM’s are trained on huge libraries of content. What they do is assign a token(id) to each word (or part of word) and then note down the frequency of the words before and after the word as well as looking specifically for words that NEVER come before or after the word in question.

    This creates a data set that can be compared to other tokenized words. Words with vary similar data sets can often be replaced with each other with no detriment to the sentence being created.

    There is something called a transformer that has changed how efficiently LLM’S work and has allowed parsing of larger volumes by looking at the relation of each tokenized word to every word in the sentence simultaneously instead of one at a time which generates better more accurate data.

    But the real bread and butter comes when it starts generating new text it starts with a word and literally chooses the most probable word to come next based off of its extensive training data. It does this over and over again and looks at the ending probability of the generated text. If it’s over a certain threshold it says GOOD ENOUGH and there is your text.

    You as a human (I assume)do this kind of thing all ready. If someone walked up too you and said “Hi! How are you…” by the time they got there you have probably already guessed that the next words are going to be “doing today?” Or some slight variation thereof. Why were you able to do this? Because of your past experiences, aka, trained data. Because of the volume of LLM’S data set it can guess with surprisingly good accuracy what comes next. This however is why the data it is trained on is important. If there were more people writing more articles,more papers,more comments about how the earth was flat vs people writing about it being round then the PROBABLE outcome is that the LLM would output that the earth is flat because that’s what the data says is probable.

    There are variations called the Greedy Search and the Beam Search but they are difficult for me to explain but still just variations of a probability generator.



  • Let me tell you about my 2007 Toyota Yaris Hatchback manual drive.

    In my opinion this is nearly as perfect of a vehicle as it is possible to get.

    The 1.5 liter engine is small and efficient which means your gas bill is nice and tiny compared to the average vehicle I see on the road. It also has this weird quirk of FEELING really fast and exciting while driving while actually being rather pedestrian. A year after this vehicle was released motortrend came out with an article about the slowest cars they have ever tested: The Toyota Yaris was the 5th slowest. Probably due to the manual gear box, the sharp and agile steering, and the noise it makes, it simply feels a LOT faster than it really is. THIS IS A POSITIVE. It means you can have a good time and enjoy driving it but unless you are trying REALLY hard you won’t be speeding all that often and even with your foot to the floor you won’t be ripping away from traffic and drawing attention to yourself. That doesn’t mean you can’t red-line the engine, drop the clutch , and rip a nicely satisfying burnout, because you can… And I have.

    Oh BTW… if you want to have some fun you can buy a ready to install everything included SUPERCHARGER kit for the Yaris. It’s on my bucketlist.

    The cargo space is MASSIVE! because the rear seats fold flat and it’s a hatchback with a wide trunk opening and a flat-ish roof (instead of aggressively raked back) the amount of stuff you can fit in it is kind of insane. Several hundred pounds of firewood? Check. Two fully assembled kitchen cabinets to be turned into a kitchen island? Check. 55" TV in box? Check. 6.5’ Christmas tree? Check. Just look at that cavernous space!

    Shoot, my wife and I regularly go car camping out of the Yaris. If you push the front seats as far forward as they go you can fit an inflatable mattress in the back with only a slight bend in it. It’s remarkably comfortable and unless you need to sleep perfectly straight as a log it serves very well as a mini RV. Back when I commuted 26 miles to my full time job and then another further 55 miles to my full time schooling I would often sleep in the back of the Yaris between the two and have very restful and replenishing sleep.

    Here we are on Rollins Pass in Colorado at 11,600 ish feet:

    At first having the gauges in the center of the dash was a bit weird but it comes with two bonuses. The first one took me a while to notice: You feel more connected to the road and your journey. Putting the dashboard gauges directly in front of the driver actually puts a barrier between the driver and the road ahead of them. It’s a wall of information density that permanently exists between you and the world ahead and you have to go through it before you can experience what’s before you. It might be a borderline subconscious thing but not having something that constantly wants your attention in front of you really lets your mind focus on the road ahead of you and the journey you are on. If you NEED the information, it’s still there, just politely sitting off to the side waiting to tell you whatever you need to know.

    The second bonus to the center gauges? MOTHER FUCKING GLOVEBOXES BABY! THIS CAR HAS THREE! There is the standard glove box around the knees of the passenger but there is also one above that and a THIRD one above the steering column on the driver’s side. I never would have guessed how excited a grown man could be (me) about the discovery of multiple GLOVEBOXES in a car.

    Almost nearly as much as I enjoy the gloveboxes I really am impressed by the setup of the cup holders. You have your standard 2 cup holders down by the hand brake in the center console but the really awesome ones are seamlessly folded into the dashboard near the doors. These aren’t your tiny popout cupholders you find in most cars that break the second time you put a big gulp in one. No… these are chunky, heavy duty cup holders that make an audible ca-thunk when deployed. The amount of times I’ve deployed the cup holder at a driver through and had the teller make a visible reaction or even stop to say something is significant. My words probably do not do them justice so look at these pictures of their location and diagram from the manual and tell me that they don’t inspire confidence.

    There are only TWO things I would like changes about this car. Give me a Bluetooth enabled head unit with better speakers and a good place to put a chi charger for my phone. That is all.

    I could go on for hours about this car but my last point about how epic this vehicle is and how we don’t deserve it is this: It’s a Toyota. A proper old fashioned bulletproof, reliable, affordable Toyota. Parts are dirt cheap and easy to replace.I’ve got 266,000 miles on mine and let me tell you, they have not been kind miles. We regularly take this on off road trails bouncing off of rocks and occasionally trees. I’ve torn the O2 sensor clean off of the car a couple of times and got it stuck up to the bottom of the door in deep snow while driving a dirt road pass in the Rockies. I have treated it like dirt and only done basic maintenance far less than it deserves. I’ve only had to replace the clutch once and this next summer will be the first time ever that I need to do anything even approaching major service. It’s got a water pump leak and a front timing cover leak. Neither of these stop the car from functioning at all but as long as I keep an eye on the fluid levels we are good to go.

    All this and it takes it like an absolute champ. It trucks along being the best little car it can be. The snow, dirt, and mud, and neverending miles of cross country journeys this car has never failed me. I will not part with my beautiful little car for anything less then total destruction. The day that happens I will remove the logo from it’s sad lifeless carcass, frame it and hang it on my wall for all to know what an amazing being was part of our lives for so long and yet not nearly long enough.

    I love my car.



  • oh man this actually pisses me of so much. I’ve been using the swiftkey keyboard long before it was bought by Microsoft, back when it was a paid app, and I swear it has gotten significantly worse at word prediction , key detection, and swipe accuracy.

    I’m finding that I have to type letter by letter if I want any kind of garuntee that it will choose the right words. Swipe has become a waste of time because I am constantly going back to fix words it got wrong. And B instead of space is a constant!


  • I got 17 out of 20. I pegged the bezerk drawing as generated because the bottom part of the armor lacked symmetry and didn’t make any sense. I got the other three line drawings incorrect.

    I have spent WAAAAY to much of my freetime generating images and apparently have picked up an eye for the weird types of artifacts that these generators produce. The hardest one to articulate is that generated images have a very specific type of noise. Images create a very nice grainy type noise while digital images get more of the blocky jpeg artifacts and banding. Generated images get this weird hybrid of the two that isn’t consistent across the whole image.