4 ms·

For what it's worth, I burned out sometime in the late 90s just after video cards came out around 1997, because I could see how they would drive video game desi

For what it's worth, I burned out sometime in the late 90s just after video cards came out around 1997, because I could see how they would drive video game design towards a narrow branch of DSP/rasterization and building derivative games with cookie cutter libraries. At the time, I was more interested in getting away from all of that and exploring stuff like ray tracing and artificial life. My interest in video games never fully returned in the same way. And if I'm being truthful, I've been burned out ever since, approaching 25 years now. But after two lost decades, I've noticed a shift lately. I think that people are getting fed up with centralized control of the supply chain and are maybe looking towards things like FPGAs as a way to get to a level above the current imperative programming paradigm, rather than working at the level below it in the glorified DMA transfer and shader paradigm we have now (vertex buffers). For something as esoteric as quantum computing, I think it would help the world a great deal if someone would go back to first principles and explain it how it is. Not with complex notation, but with counting principles from probability and statistics. I had a great teacher in college who taught it just exactly that way, by imagining small groups of apples and bananas and how we could count them and detect larger patterns to get to stuff like Bayes' theorem. So for example, we can imagine cubits as coin flips that we haven't looked at yet. If there are relationships between multiple unseen coins, then we can map out that potential solution space like a big tree. Then look at the tree with random sampling of the superposition to infer where the answer probably lies. Which is a lot faster than iterating through all all of the permutations with a computer. But I suspect that convergence will end up taking roughly the same amount of time as brute force, due to noise or some fundamental aspect of the universe we don't understand yet. But I'm probably wrong about that, because I haven't gotten clarity on how any of this stuff actually works, because nobody bothers to write it this way from first principles. Even if the author isn't going to be doing quantum mechanics in academia, it would help the world a great deal to share what they've learned so far!

selimthegrim 4y ago
What you’re looking for exists with stuff like Gleason’s theorem and noncontextuality, but it’s not what you would train up in if you wanted to be “productive”.
- zackmorris 4y ago
  I'll use your comment to correct myself, I meant to write "qubit" heh. Thanks yes, after posting my comment, I was researching qubits and Hilbert spaces and stumbled onto ergodic processes, where the properties of a process can be inferred from random sampling: https://en.wikipedia.org/wiki/Ergodic_process https://en.wikipedia.org/wiki/Ergodic_process I believe that combining it with the Born rule (which gives the probability that a measurement of a quantum system will yield a given result) might lead to Gleason’s theorem: https://en.wikipedia.org/wiki/Born_rule https://en.wikipedia.org/wiki/Born_rule https://en.wikipedia.org/wiki/Gleason%27s_theorem https://en.wikipedia.org/wiki/Gleason%27s_theorem This is all so far outside of my wheelhouse though that I'm almost certainly not really saying anything. Which is the fundamental issue I take with quantum mechanics. It introduces too much new terminology for simple concepts because experts appreciate the short notation to be productive, as you said. I've been programming for three decades but would never try to explain it in Lisp. I'd start with Javascript, but demonstrate where each conceptual flaw occurred so that the student derives much of Lisp before I get to it. The heart of my cognitive dissonance is: I need to know what separates theory from quantum systems in real life. If we know what we're trying to do, then we should be able to build it in a simulator and just pause it periodically and examine the state. Is the limit of that the speed of our computers? Is it that quantum mechanics plays out at the speed of light? Is it that entangled particles are really connected regardless of distance, because without hidden variables they have to be, so can't be simulated? So far the best I can do is, imagine we're breaking N bit cryptography. We have one key, and know a mathematical relationship to the key we're trying to find, but the search space is huge. Maybe every time we sample, some percentage of the qubits reveal something, but we don't know which ones. Maybe we sample more, and see that certain ones tend to stay the same. Maybe when we get down to a tractable number of qubits remaining (like 32), we just brute force search the remaining bits classically by iterating over the (2^32) permutations? Apparently there's just no path in layman's terms to get from my improper description to what really happens. Which isn't true in computer science. If I understand a topic, it's borderline trivial to explain with metaphors. Ok I think after writing this out, I'm looking for an open source implementation of quantum mechanics that doesn't use established terminology. Maybe it looks like a wiki with every use case and bit of errata answered. So like, everywhere that complex numbers appear, we could just use independent vectors and formulas to get the angle and magnitude. Same for matrices, forget about eigenvectors, just explain it in a way that people understand, like center of mass or the direction with the least rotational inertia. No integrals, just summation and approximation. Yes, we'd lose the abstractions that let us find new associations, but after we have working mental models, I think we'd grasp them automatically by our curiosity. Instead of hitting this woo woo wall that makes it impossible to talk about with enthusiasts like me!

selimthegrim 4y ago

What you’re looking for exists with stuff like Gleason’s theorem and noncontextuality, but it’s not what you would train up in if you wanted to be “productive”.

zackmorris 4y ago
I'll use your comment to correct myself, I meant to write "qubit" heh. Thanks yes, after posting my comment, I was researching qubits and Hilbert spaces and stumbled onto ergodic processes, where the properties of a process can be inferred from random sampling: https://en.wikipedia.org/wiki/Ergodic_process https://en.wikipedia.org/wiki/Ergodic_process I believe that combining it with the Born rule (which gives the probability that a measurement of a quantum system will yield a given result) might lead to Gleason’s theorem: https://en.wikipedia.org/wiki/Born_rule https://en.wikipedia.org/wiki/Born_rule https://en.wikipedia.org/wiki/Gleason%27s_theorem https://en.wikipedia.org/wiki/Gleason%27s_theorem This is all so far outside of my wheelhouse though that I'm almost certainly not really saying anything. Which is the fundamental issue I take with quantum mechanics. It introduces too much new terminology for simple concepts because experts appreciate the short notation to be productive, as you said. I've been programming for three decades but would never try to explain it in Lisp. I'd start with Javascript, but demonstrate where each conceptual flaw occurred so that the student derives much of Lisp before I get to it. The heart of my cognitive dissonance is: I need to know what separates theory from quantum systems in real life. If we know what we're trying to do, then we should be able to build it in a simulator and just pause it periodically and examine the state. Is the limit of that the speed of our computers? Is it that quantum mechanics plays out at the speed of light? Is it that entangled particles are really connected regardless of distance, because without hidden variables they have to be, so can't be simulated? So far the best I can do is, imagine we're breaking N bit cryptography. We have one key, and know a mathematical relationship to the key we're trying to find, but the search space is huge. Maybe every time we sample, some percentage of the qubits reveal something, but we don't know which ones. Maybe we sample more, and see that certain ones tend to stay the same. Maybe when we get down to a tractable number of qubits remaining (like 32), we just brute force search the remaining bits classically by iterating over the (2^32) permutations? Apparently there's just no path in layman's terms to get from my improper description to what really happens. Which isn't true in computer science. If I understand a topic, it's borderline trivial to explain with metaphors. Ok I think after writing this out, I'm looking for an open source implementation of quantum mechanics that doesn't use established terminology. Maybe it looks like a wiki with every use case and bit of errata answered. So like, everywhere that complex numbers appear, we could just use independent vectors and formulas to get the angle and magnitude. Same for matrices, forget about eigenvectors, just explain it in a way that people understand, like center of mass or the direction with the least rotational inertia. No integrals, just summation and approximation. Yes, we'd lose the abstractions that let us find new associations, but after we have working mental models, I think we'd grasp them automatically by our curiosity. Instead of hitting this woo woo wall that makes it impossible to talk about with enthusiasts like me!