1933: John Hopfield is born

No landmark computing event is reliably dated to 14 July itself, so today we mark the closest well-documented anniversary in the AI story: physicist John Hopfield was born on 15 July 1933 in Chicago. It is a one-day gap worth being honest about, but his work sits so close to the heart of modern machine learning that it earns its place in this calendar.

Hopfield trained as a solid-state physicist, studying things like excitons in crystals long before he turned his attention to brains and computation. That shift, in the late 1970s, was unusual: a physicist bringing the tools of statistical mechanics to the question of how networks of simple units might store and retrieve memories.

In 1982 he published "Neural Networks and Physical Systems with Emergent Collective Computational Abilities" in the Proceedings of the National Academy of Sciences. It described what we now call the Hopfield network: a recurrent system of interconnected nodes that settles into stable, low-energy states, much like a physical system relaxing toward equilibrium. Crucially, those stable states could be used as memories, letting the network reconstruct a complete pattern from a partial or noisy cue.

• 1982: Hopfield's PNAS paper introduces associative memory via neural networks
• Mid-1980s: the work helps revive interest in neural networks after the first "AI winter"
• 2024: Hopfield shares the Nobel Prize in Physics with Geoffrey Hinton for foundational work enabling machine learning with artificial neural networks

The paper mattered far beyond its immediate results. Alongside the backpropagation work that followed a few years later, it helped pull neural networks back from the margins of computer science and into serious research, laying groundwork for the connectionist wave that eventually grew into today's deep learning.

That lineage was recognised in 2024, when Hopfield and Geoffrey Hinton jointly received the Nobel Prize in Physics "for foundational discoveries and inventions that enable machine learning with artificial neural networks." Researchers have since shown that modern, continuous versions of Hopfield networks are mathematically close cousins of the attention mechanism inside Transformer models, the architecture behind today's large language models. A physicist's idea about energy and memory, born from a mind shaped in mid-century Chicago, still quietly underpins the chatbots and recommendation systems of 2024 and beyond.