Mind Simulation Lab

Mind Simulation Lab is an independent AI research lab pursuing an alternative path to Artificial General Intelligence by addressing core limitations that continue to constrain modern AI systems. These include the elimination of hallucinations, the resolution of catastrophic forgetting, and the development of a transparent and explainable world model – an increasingly central objective in contemporary AI research.

For more than a decade, its fully bootstrapped team has been developing a new class of AI architecture: systems with structured world models, transparent reasoning, and persistent memory. These systems learn incrementally in a single pass, retain knowledge over time, and operate with extreme computational efficiency even on devices as lightweight as a Raspberry Pi.

While much of the industry focuses on scaling compute-intensive generative models, MSL takes a fundamentally different approach, constructing AGI from first principles. Its mission is to develop a reliable, multimodal intelligence capable of reasoning, planning, and interacting with the real world without the unpredictability of current generative systems.

MSL’s technologies reach end users through STEOS, the lab’s product ecosystem. Originally launched as an AI voice generation platform, STEOS now serves over 2 million users and hundreds of B2B clients with a growing suite of AI-powered tools.

Based in Yerevan, Armenia, Mind Simulation Lab is expanding globally, because there is no reason frontier AGI has to be built in Silicon Valley.