Scroll down. The story unfolds horizontally — five chapters, eighty years, and a future still being written.
McCullough & Pitts draw the first artificial neuron. Turing asks 'can machines think?' The first perceptrons learn. Then winter sets in — funding evaporates, expectations cool. But a small group keeps the flame alive in universities and back rooms.
Backpropagation revives neural networks. Support vector machines, decision trees, the rise of statistical learning. The web explodes, generating oceans of data. GPUs turn out to be the perfect substrate. Something is stirring.
AlexNet crushes ImageNet. Speech recognition crosses the human threshold. AlphaGo defeats a world champion. DeepMind, OpenAI, and a Cambrian explosion of labs push the frontier at an exponential pace. The world starts to pay attention.
Attention is all you need. GPT, BERT, T5, PaLM, LLaMA — the era of the foundation model. A single network, scaled, can do almost anything. Multimodal. Multilingual. Multi-task. A Cambrian explosion of capability.
Reasoning models pass the bar exam, code at expert level, generate feature films. We are entering an era where intelligence is a utility — abundant, ambient, available on tap. The next chapter is being written right now, by all of us.
Every chapter from here on is written by the people who build with NEXUS.