AI startup Counterintuitive is pioneering the development of reasoning-native computing, aiming to move AI beyond pattern recognition to genuine comprehension. This breakthrough could lead to more human-like systems capable of thinking and making decisions.
Counterintuitive Chairman Gerard Rego highlighted the ‘twin trap’ problem facing current AI systems, focusing on the lack of reliable numerical foundations and memory in architecture. These problems limit the stability, efficiency, and true intelligence of AI, leading to non-determinism and a lack of reasoning behind decisions.
With a team of experts in mathematics, computer science, physics, and engineering, Counterintuitive is working on deterministic reasoning hardware, causal memory systems, and software frameworks to redefine computing based on reasoning, not mimicry. Their artificial reasoning unit (ARU) aims to create a new era of computing that prioritizes memory-driven reasoning and transparent, auditable systems.
