LQG Logo

Developing self-learning artificial intelligence.

At London Quantum Group (LQG), our antithesis is based on a belief that intelligence is not continuous, but discrete. Modelling intelligent systems with continuous mathematics leads to efficient optimizers and compression machines, but logical inference should be modelled as a discrete process.

Current continuous and differentiable mathematical foundations are, paradoxically, still built upon discrete approximations of continuity: whether at the software level, where optimizers take quantized steps through parameter space, or at the hardware level, where all computation rests upon discrete logic gates.

We hold that logical inference is inherently discrete, a process of symbolic transitions. Real world data has an inherent mathematical structure, and we aim to push forward algorithms and hardware that try and exploit the power of building discrete algorithms fueled by number theory on discrete software, which seems to be an emerging intrinsicism.

We have found a number theoretic method to conduct repeated modular division (the mathematical backbone operation for our discretism approach) of large numbers in such a way which beats current SOTA systems by orders of magnitude on their hardware.

We aim to build custom hardware optimised for handling the conversion to a residue number system and conducting operations in that space, as well as building algorithms on top of this new hardware that exploit inherent patterns in numbers to build digital logical inference not just big data correlation.

Building by sharing

The greatest breakthroughs don’t come from isolation but from collaboration. At LQG we believe in sharing, and beyond sharing: understanding. We open our framework through modular tutorials and research tools, inviting others to solve challenges with us.

Building with purpose

We also believe discovery thrives on purpose. Darwin’s theory of natural selection arose from a voyage intended for naval mapping. Likewise, we align our research with real world needs, ensuring that elements of our framework find markets and applications that sustain further innovation.