The People Behind AI

Asked 'Can machines think?' in 1950 and proposed the Imitation Game. His Turing machine established the mathematical foundation for all computation. Died at 41, decades before his ideas were proven right.

Coined 'artificial intelligence' in 1955, organized the Dartmouth Workshop that launched the field, and created LISP. His proposal predicted AI would take one summer to solve. It took seventy years.

Co-authored the backpropagation paper, spent 30 years proving deep networks work, won the Turing Award and Nobel Prize, then left Google to warn about AI risks. The person most responsible for building the technology is now one of its most prominent critics.

Developed convolutional neural networks — the architecture that made image recognition work. Now leads AI at Meta and is the field's most vocal advocate for open-source models and against AI doomerism.

Pioneered recurrent neural networks and word embeddings that led to modern language models. Unlike his Turing Award co-laureates, Bengio has become increasingly focused on AI safety, arguing for international governance before the technology outpaces control.

Former Y Combinator president who became the face of the AI revolution. Oversaw ChatGPT and GPT-4, survived a boardroom crisis, and is now leading OpenAI's transition from nonprofit to capped-profit — defining the tension between AI idealism and commercial reality.

Left OpenAI to co-found Anthropic around the thesis that safety research and frontier capability must coexist. Built Claude to compete with GPT while emphasizing constitutional AI and interpretability.

Co-founded NVIDIA to make gaming graphics chips. Those same chips now power every frontier AI model. Bet the company on AI compute before anyone else saw the market — making NVIDIA one of the most valuable companies in history.

Left Google over a paper on bias in large language models. Founded DAIR to study AI harms independently. Her work centers on how AI systems encode inequality — and who bears the cost when they fail.

Deep learning's most persistent critic. Argues LLMs can pattern-match but not reason, and scaling won't fix it. Whether he's right that we need a fundamentally different approach is one of the field's most important open questions.

Helped popularize the term 'artificial general intelligence' and has been building toward it since before the current deep learning boom. Founded SingularityNET to decentralize AI development on blockchain — rejecting the idea that AGI should be controlled by a handful of corporations. Believes AGI is only a few years away and will be a net positive, assisting with climate change and economic inequality. His approach combines cognitive architectures, evolutionary learning, and probabilistic logic — fundamentally different from the 'scale the Transformer' consensus. Whether he's right that AGI requires architectural diversity rather than just more compute is one of the most consequential open questions in the field.

Co-founded DeepMind in 2010 with the explicit goal of 'solving intelligence.' Built AlphaGo, AlphaFold (which predicted the structure of nearly every known protein), and now leads Google's combined AI research. Represents the thesis that AGI emerges from game-playing, scientific reasoning, and neuroscience-inspired architecture — not just language modeling at scale.