AI & Visionary Thinking

In the 1980s and 1990s, most of the AI world had moved on from neural networks. These systems—designed to mimic the brain's structure—were dismissed as inefficient, shallow, and impractical. The dominant trend was symbolic AI: logic-based systems engineered to follow programmed rules. It was neat, controllable, and explainable—but it couldn't learn or adapt.

Geoffrey Hinton thought differently. He believed that the brain-like structure of neural networks held the key to true intelligence. While others abandoned them, he quietly refined his ideas, waiting for computing power and data to catch up. In 2012, that moment arrived.