Look, ma, a new episode just dropped! This one is full of shit to the brim. Even more so than prof. Ellen Langer who cannot stay within a context and claimed that 1+1 = 10 because in the binary notation it looks like 10 in decimal… anyway, whatever. https://www.youtube.com/watch?v=MlmFj1-mOtg No, the brain ain’t computing any hecking probabilities. It is not a Bayesian machine. It is not a prediction machine. It is not a simulator. It is not a statistical engine. ...
Time waits for no one, the race to the bottom accelerates faster than ever, and the “future” is now. Competition is severe and mostly meaningless, as in some third-world criminal infested ghetto. This is what LLMs turned our world into. So, lets “pee on” so called “prompt engineers”, in the 4chan parlance, of course.. Here is my benchmark prompt to evaluate performance of LLMs. All the “simple” offline models fail miserably, and only Grok and Gemini can produce something adequate. Claude is also good, but it is supposed to be the best, being trained especially for code generation. ...
FUCK THIS SHIT! No, really. It just hallucinated a non-existent package, with such a confidence lmao Yes, yes, I know, it cannot know anything about “existence”, but at least these dorks do not “deserve” the exuberant money they got paid. Uniform Code Block Rendering in Eww and Shr To get syntax highlighting in eww and shr, you need to use a package that intercepts <code> and <pre> tags and applies Emacs’s built-in font-locking. The eww-code-blocks-mode is a good choice for this. ...
Oh, look DeepMind and OpenAI win gold at ICPC (codeforces.com) So, a model memorized a lot more of a convoluted stuff and been able to spit out a coherent code. While this is absolutely amazing considering that it works at the level of syntax, and that its “representation” captures only “possible relations among tokens” and posses no “understanding”, thinking leave alone reasoning capacities whatsoever. It just picks up one of the few next most probable tokens, given the previous “context” (a sequence of tokens “so far”). ...
And while passing by… Yet another “look, look at us, we are soooo smart and clever, give us much more money just because we are so cool” article dropped. https://thinkingmachines.ai/blog/defeating-nondeterminism-in-llm-inference/ The lack of exact precision is not the fundamental issue here. Even if one manages to overcome the “numerical instability issues” and would be able to reproduce always the same structural output from the same linguistic (or otherwise structured) input the “hallucinations” and “subtle bullshitting” won’t go away in principle. ...
To understand what non-deterministic, “syntax level” probabilistic models are actually producing (an illusion), we have to understand how the Mind (of an external observer) works, and how it produces Maya (which has been intuitively understood since the early Upanishads) – an ultimate illusion created by the Mind itself – an inner representations of the “outside” world, which the mind (and body) uses for “decision making”. The “outside” world is inherently complex, non-deterministic and “concurrent” at the level of “compositions” and, at the same “time”, deterministic enough at the level of the “most basic building blocks” (of biology, lets say) – “simple” molecular structure’s (“small” molecules) are exactly the same, exact copies of clones of each other (otherwise everything will break) while larger molecular structures (like whole proteins and their compositions) may have “flaws” or “mutations” or just simple “kinks” – a slightly different shape or a resulting form. ...
I have noticed a resent dramatic change in the behavior of major GPTs online providers – most notably, Gemini is now providing just outline of code, full of stubs and “mocks” of real APIs, and not the full code. This is a significant change from the previous behavior where they would provide a semi-complete (but ridden with errors) code solution. Perhaps, mimicking the behavior of ChatGPT, which has been doing this for a while now – they “optimize” for more what appears to be a “dialogue” (more like a normie-level chat), to create a better illusion of “actually conversing with an artificial intelligence”. ...
We are living through a paradigmatic shift, the one described in the “Scientific Revolution” by Thomas Kuhn. As I mentioned many times, texts and even crappy code became very, very cheap, just like a processed junk-food or a low-effort street-food slop. This is the “shift” and the end of so-called “knowledge work” as we know it. At least this is the end of the pretentious “knowledge work”, when one just pretends to be an expert is social settings, using very straightforward verbal and non-verbal cues to signal their “knowledge” and “expertise”, just as a priest would do in the not so distant past. ...
I understand a lot of complex things, maybe because I spent my whole life trying to understand and explain things around me, since 4 year old, when I used no name every single car on the road in a small Ukrainian Steel and mining town where I was born. Understanding cannot be “outsourced” or even safely “delegated”. One will always end up with a sort of “tragedy of commons”, when sterilization and modern technologies produced packaged foods which slowly but surely kill you. This is what happens when you “delegate” your own understanding. ...
https://www.youtube.com/watch?v=LCEmiRjPEtQ and, of course, the No.1 spot on the Chuddie safe space https://news.ycombinator.com/item?id=44314423. Karpathy is shilling “Cursor” and other cloud-based mettered AI services (which have to pay back their debts). Probably has an interest in it and some other meme AI startups. Nothing to see here. We should some day know which marketing “genious” came up with this “winning strategy” – to metter every single token (byte) and try to sell this to corporations. Corporations do not want to be mettered like that, they want to metter normies, the way Cellular operators do, and they never use any normies plans themselves. ...