Mura ALife Labs
Building intelligence that learns cheaply, continually and exactly, from silicon to living neurons.
What we work on
01 Unconventional AI
Language intelligence built on different foundations: systems that store what they read as exact counts, learn from new text immediately, run on an ordinary CPU, and forget any document exactly on request. The question: can learning be cheap, continual and exact, rather than bought with ever larger compute?
02 Computational neuroscience
Learning rules the brain could actually run: each synapse uses only its own activity plus one broadcast signal, with sparse spiking, inhibition and synaptic tagging. We test them in standard neuroscience simulators, on real language, end to end.
03 Neuromorphic computing
Language models that learn and forget on low-power neuromorphic hardware, in the chip's own low-precision format, within one chip's budget, and next on living-neuron hardware.
Almost all of our work is digital: every idea is designed on paper first, then tested on normal CPUs with its success bar written down before it runs, from a small bedroom lab.
People
Prince Siddhpara — Principal scientist and founder (Age 17)
Sneha Shah — Lab assistant
Sometimes temporary appointments join for specific projects.
Contact
Email: contact@mura-alife.com
LinkedIn: LinkedIn Company page
X: @MuraALife
© 2026 Mura ALife Labs · mura-alife.com