I can’t tell if you are being serious or sarcastic. Assuming you are being real, and as someone who has to listen to boring youtube videos to help falling asleep, i have a few questions: how long is the story? What happens after the story ends? Like, does it turn off the voice feature or does it stay on and randomly speaks during the night?
I have it as well. This is how i explain it to regular folks: most people have an whole neural network dedicated to processing the human face, memorizing its feutures and matching them to the people in front of them. But when they look at trees they cant really distinguish a pine tree from another pine tree, unless the differences are extreme: one is brunt down, or has a weird shaped trunk, etc. If they try to recall a specific pine tree all they can recall is a messy green spiky blob. Why? Because their brain has no neural network dedicated to identifying trees. For us, the human face is as distinguishable as a tree
to be fair you need to create a dozen images to get a decent one. Many times i need to get to almost what i want using paint.net and hope chatgpt removes the aliasing/cleans up the pixel boundaries without messing anything up - i'm not complaining though, still faster and cleaner than photoshop -- when it works
I asked gemini 3.8 high to review the site I'm working on for points of high cpu/ram consumption - it failed spectacularly and also halucinated the server i/o limits
I'm seeing it more like: take the concept of "good" put it in a scale of -10 (pure evil) to +10 (pure good). These concepts and the inbetweens have been ingrained into the model, the model can multiply its weights in any combination to express any level and any in between, even -2.16541 etc. So, suppose you give it a short story and ask the model to reason about it and how a character displayed good vs evil behaviour towards the story: Internally the model is making calculation that are very nuanced and extremely precise. This calculations are not the reasoning trace. The reasoning trace itself does not influence the calculations. You may read: "John starts bad and slowly becomes good" when inside the calculations are John goes from -5.245 to -4.24 to -5.221 again, and then 2.1. What matters for nuance is the inner calculations across the many matrix layers. What you see is like an independent program that looks at "-5.245 to -4.24 to -5.221 again, and then 2.1." consults the tokenizer and outputs: "John starts bad and slowly becomes good" or even "John first bad, then good". When in reality, inside, the model as been processing something more akin to "John starts the story as a despicable person, with a redemption arc that builds slowly, he can't yet be considered a good person, certainly not the kind of good person you'd leave your dog with, but he's certainly not as bad as before" the whole time. Now, what if when you continue the conversation, what does the model receive as context? It's original nuanced sentiment, or the brute reasoning trace? That I don't know. It might be that when the reasoning trace is converted from tokens back into numbers it loses all nuance, or it might be that the trace (the words you see) are not the only thing that is being saved and is not the only thing being fed back as context
It's my understanding that the llm is not literally thinking those words, they are just the conversion of the matrix multiplication results (numbers) into the tokens. So the matrix is "multiplying" concepts and directions to come up with the final answer - which produces a somewhat readable reasoning trace. As far as the llm is concerned the reasoning trace could be random (to us) symbols. In fact, the reasoning traces are not necessarily optimized for readability as much as they are an emergent property of the way a reasoning model is trained
My father is blind, the only tech he knows how to use is a smart tv, when i build him custom apps the first thing i do is turn off talkback when the app opens, so i can control what tts says. The regular talkback is so so annoying and repetitive
Maybe i’m recalling the youtube videos I saw on Nuremberg incorrectly, but i recall something on the lines of: the concept of ‘crimes against humanity’ being applied retroactively
Also rare genetic issues like GSDs and FAODs may benefit immensely from dual glucose and ketone metering. Next we need potassium and magnesium meters and that will benefit the people with rare renal wasting syndromes like Gitelman and Bartters
If they had ask Claude it would probably look like this: Introducing the all new Cerebras CS-4, a revolutionary rack-scale solution that delivers up to 30x faster inference compared to GPUs, enhanced economics, and a simple path to load-bearing hyper scale capacity.
There are MANY facebook groups filled with people fully convinced that they have Lyme disease. These groups teach that virtually any and every symptom is caused by the infection and that basically everyone has it. As if this wasn’t insane enough, the people in the group will push each other to go on antibiotics - via so called lyme literate doctors - until they are on the verge of death due to the side effects. Worse: if a lab test is negative they say you should treat by symptom (referring people back to the massive list of virtually every known symptom as proof that they have lyme). It gets worse: even if you get really bad reactions to the treatment, they say you are having a herxheimer reaction and to keep pushing. It gets worse: having followed many of those groups for years, it’s rare that anyone actually gets better from whatever symptom they were having that led them down the self destructive path of lyme self-diagnosis. Well, it actually is a bit worse: even if by some chance you did had a positive lyme test, and if by chance you did got years worth of antibiotics, SOT treatments, etc etc etc… and now you have finally a clean test- well actually it’s bartonella, or parasites, or babesia, or EBV or whatever … so you must carry on. And, it’s not that these people are idiots, it’s just that they fall into a very specific category: 1) have real symptoms, 2) have no real answers from medicine, 3) have internet access, 4) have no notion, not even a tiny spec, of the existence of a concept called falsifiability. So when i read that there’s a new at home test for lyme… on tics(!) i can only imagine the massive amount of suffering this will cause on these poor people… and i imagine many if them will put their blood on the test and come up with all kinds of ways to use this to find yet another proof that they have it
Yeah, it’s hard to believe, specially when you are coding and there’s only one best way to do things, unless it plays with variable naming, or comments
From the little i understand that wouldnt be an issue because the model is ‘just’ using interchangeable words in a mathematical non-random way. Like using the same number of adjectives and the exct same words, but in a order that wouldn’t be mathematically plausible unless it was the watermark
Well, you do get 1 million tokens and the ability to reason over video natively and many of us are forced to pay for 20usd plan anyway due to google drive 5TB, not to mention notebooklm, so it’s not a nothing burguer, it’s just an almost nothing burguer