Unconventional AI

In a recent episode of the @ThisWeeknAI, our CEO @NaveenGRao joined @jason for a roundtable discussion with @chaselochmiller (@CrusoeAI) and @ml_angelopoulos (@arena). Naveen discusses the long-term limits of scaling AI on today’s computing architectures and why new approaches to hardware and computation will be needed as the field continues to grow. The discussion also touches on infrastructure constraints, evaluation challenges, and how engineers can stand out as the industry matures. These are the kinds of questions we are exploring every day at Unconventional AI. Watch the full conversation below. youtu.be/9o13P7SMAaY?si…

10/10 What other fundamental questions should we be asking about dynamical systems and computation? Join us as we continue to explore the limits of programmability, steerability, and efficiency for tomorrow’s computing. unconv.ai/careers/

9/10 Success! When you plot the learned dynamics through time, the results demonstrate the steerability and programmability of these simple dynamical systems. Yes, this demo is whimsical, but it provides a clear answer: these systems are highly programmable. This answer hints toward a bright future for energy efficient computation by reimaging the very components.

1/10 Reimaging computing using dynamical systems raises a host of fundamental questions, among them: How programmable/steerable is a candidate system? This week, we ran an experiment to test the limits of programmability by asking: is a toy 4-oscillator system expressive enough to dynamically sweep out any arbitrary pattern in phase-difference space? After testing our “[un]” logo, we concluded these systems are highly steerable. Here’s how we did it.

Unconventional AI is designing a new computational substrate for intelligence. This requires extreme codesign across AI algorithms, systems, compute models, analog circuits, and hardware architecture. We are hiring across all of these areas. Learn more about our open roles on our careers page. unconv.ai/careers?cache=…

The biggest problem for AI in 2026 is not prompts or architectures. It is energy, throughput, latency, and scale. Unconventional AI is solving that. youtu.be/VVVOQLLOero @lightspeedvp @NaveenGRao

AI scaling will be energy-limited at the global level in the next 3-4 years. Conventional computing is reaching its limits. It’s time to stop simulating neural networks on digital logic and start building hardware that actually behaves like them. We are Unconventional AI. Learn more about our mission: unconv.ai/introducing-un…

[x(t_0) = x_0] ——————— import time from math import prod print("--- Initiating Dynamical System Simulation ---") # The Attractors Y = [104, 101, 108, 108, 111, 32, 119, 111, 114, 108, 100] # The Analytic Evolution Function x(t): Lagrange Interpolation governing the dynamics f = lambda t: sum(y * prod((t-k)/(j-k) for k in range(len(Y)) if k!=j) for j,y in enumerate(Y)) # Time Evolution: Step through discrete time t, observing state x(t) acc = "" for t in range(len(Y)): c=chr(int(round(f(t)))); acc+=c; print(f"Time t={t:2} | x(t): '{c}' | Trajectory: {acc}"); time.sleep(0.2)