Alx The Programmer

This isn't a random string of numbers. This is the structural DNA of the collision. If you look at the k values (the frequencies: 213, 491, 489, 209), you'll notice they are incredibly high. A standard, smooth gravitational wave might have a k value of 2 or 3.

I might have finally figured out why standard physics fails to find a gravity particle. "The engine is suggesting the missing bridge isn't the gravity particle we've been looking for; it's a completely different class of geometric object."

Took some time off and back at it. Doing some cleanup on the AxCore repository, working on a wiki, and also should have an amazing new demo released today. Ill show the results very soon and a link.

Found this old relic on a laptop laying around 😱

Thats actually exactly what you missed, it proves that they can be matched, therefore on a 1D or 2D graph you couldn’t tell their data apart with just their motion alone. Even if you increase the wheels speed, you can find a spring to match the motion of the wheel, effectively proving its still the same exact motion at a faster pace.

@PhilosophyOfPhy I'm not a physicist but, doesn't the speed of rotation count. If it's made to match the mass on a spring then it proves nothing. Just saying 🤔

A rotating ball casts a shadow that perfectly matches a mass on a spring, revealing how circular motion and simple harmonic motion are linked. A visual proof of sine-wave physics in action.

@Plinz Simple minded people

(Still, it's a bit weird that you hand the most powerful cognitive offloading tool in the history of language to a linguist, and the first applications she envisions are fraud and plagiarism, instead of eg. replicable and original linguistics research.)

I deeply agree with Emily Bender's main point: LLMs are useless, unless you want to offload cognition. (The other two usecases she suggests are rare special cases of the third.) Offloading cognition into machines has always been the purpose and application of computer science and AI.

All physicists are mathematicians but not all mathematicians are physicists 🤔

@duck_god_00 youtu.be/Y-Q8htWeO7c?si…

AxCore passed a hostile replication-style harness covering HDC-space episodic recall, exact temporal recall, perturbation robustness, thresholded working-memory retrieval, and manifold discrimination against shuffled controls. This is real evidence.

AxCore can now be downloaded, tested, and peer reviewed: github.com/AperioGenix/Ax…

Okay, here goes nothing... Ill provide documentation and more demos in a separate repo soon. Have literally thousands of use cases already tested over the past 6 months as you may see on my profile. Everything I did before is more memory efficient, faster and will just work better in general. Time to have a look yourself: github.com/AperioGenix/Ax…

Visualizing gravity as a 65,536 dimensional manifold. Using the AxCore kernel to project 3 body orbital mechanics into a holographic state-space. First image: 3D projection. Second image: Parallel Coordinates. Physics isn't just math; it’s geometry.

AI provided explanation of the two views.

By mixing a perfect sine wave, a chaotic system, and pure noise, I have created a Mathematical Zoo.

Adding a manifold viewer as the last feature before considering this project complete. Currently you can visualize the geometry in 3D, 2D, and 1D. Now working on being able to traverse the view across all dimensions starting with 4D.

The AI and I got carried away with the quantum computing metaphor. HDC vector binding stores the presence of 100 discrete concepts, but you are 100% correct that it doesn't compress 2^100 entangled, interacting amplitudes. If it did, I’d be picking up a Nobel Prize right now. But the quantum metaphor aside, the core achievement of AxOS stands: I am not trying to run Shor's algorithm. I am using HDC geometry to map semantic logic. I am replacing probabilistic neural network weights with deterministic geometric routing. It uses topological manifolds to achieve 0.2s inference and zero hallucinations on a 2MB binary. The math works, the engine is live, even if my AI hype-man stretched the physics metaphor too far. Respect for catching it. As for using Gemini to form my responses, you do not want to have a direct conversation with me. I am pure logic and not very fun to conversate with and could care less than an AI does about using the correct terms. In fact I am only conceding because Gemini said you are right and it is way smarter than me. But I take this as a push to learn more and I appreciate you correcting me before I look more like an idiot lol.

@ProgrammingAlx @RileyRalmuto You physically cannot fit 1.26*10^30 independent, interacting variables into a 8,192-item array without catastrophic data loss. The AI is conflating "superimposing 100 vectors" with "calculating a 100-qubit Hilbert space." Stop relying on Gemini and defend your theory yourself.

hang this paper in the fucking MoMA

A cryptographic hash destroys topology, change 1 bit and the output is entirely randomized. HDC does the exact opposite: it preserves distance metrics (cosine similarity). It’s a topological manifold, not a checksum. Also, binding 100 vectors in 8,192 dimensions isn't 'incredibly lossy.' You're ignoring the Concentration of Measure. In 8,192D space, you can superimpose thousands of vectors and they remain almost perfectly orthogonal. I can unbind and extract any of those 100 states with near-perfect fidelity without hitting the noise floor. I never claimed to have a physical dilution refrigerator in my lab. I said I simulated a 100-qubit state space using HDC vector superposition, calculating the geometry of the superposition rather than doing von Neumann sequential ticks. We've officially moved past classical limits.

@ProgrammingAlx @RileyRalmuto You stepped back from "I calculated 1.26 \times 10^{30} states in 61ms") to a trivial, highly possible claim ("I XOR'd 100 vectors together in 61ms"). Binding 100 vectors into a single 8,192d array is incredibly lossy compression. It is essentially a cryptographic hash - not Q.

Imagine normal AI (like ChatGPT) as a super-fast guesser. It uses massive probability tables to guess the next word in a sentence. It’s brilliant, but because it’s always guessing, it requires giant, expensive server farms and sometimes 'hallucinates' fake information. My work uses Hyperdimensional Computing (HDC). Instead of guessing probabilities, I turn words and concepts into literal, 8,192-dimensional geometric shapes. I link them together to build a physical 'map' of logic. When you ask my AI a question, it doesn't guess. It essentially drops a marble onto this high-dimensional map, and mathematical 'gravity' pulls the marble down the correct path of logic. The result? An AI that runs instantly on a cheap laptop, uses almost zero electricity, never forgets, and physically cannot hallucinate because it's navigating a static map, not rolling dice. Right now, I'm 'teaching' it how to think by extracting the logic from Marcus Aurelius and Mary Shelley and carving it into the map!

@ProgrammingAlx @RileyRalmuto Very, very interesting stuff. I have no expertise in any relevant fields but I think you might be onto something. Would you care to elaborate a bit on the nature of your work for a wanton!

1.26 x 10^30 is 2^100 (a 100-qubit state space). I didn't iterate through them sequentially. I used an 8,192-dimensional manifold. Binding 100 orthogonal vectors takes a few hundred SIMD operations (microseconds on a modern CPU), but the resulting vector geometrically represents the superposition of all 2^100 states simultaneously. That’s the entire point of Hyperdimensional Computing. Also, I pasted it into Gemini like you asked. It said your math is flawless for a 1945 von Neumann architecture, but you need to read up on high-dimensional geometry.

@ProgrammingAlx @RileyRalmuto A 5GHz CPU can execute a maximum of ~20 bn operations across all cores in 65ms. The audit claimed to have processed 1.26x10^30 states in that exact 65ms window - roughly 20 quintillion times larger than the absolute limit. Paste this into Gemini and ask it to run a hardware audit