~$ cat keeg.si a thank-you wall for the geeks who built AI · .si = SI, the new name for AI ~$ whoami --built-ai researchers, engineers, grad students ~$ ls geeks/ | wc -l 35 ~$ thank --all # tap ♥ on any card ~$
They’re calling AI “SI” now. Fine. But before anyone takes the credit: decades of researchers and engineers built it — neurons, proofs, datasets and late-night GPU runs. This is a thank-you wall for them.
1943
First mathematical model of a neural network
Described neurons as simple logic units in “A Logical Calculus of the Ideas Immanent in Nervous Activity” — the ancestor of every artificial neural network.
1948
Information theory
Founded information theory with “A Mathematical Theory of Communication”, and co-authored the 1955 Dartmouth proposal that named the field of artificial intelligence.
1950
The imitation game (Turing test)
Asked whether machines can think in “Computing Machinery and Intelligence” and proposed the imitation game, now known as the Turing test.
1955
The term “artificial intelligence”
Lead author of the 1955 Dartmouth proposal that introduced the term “artificial intelligence”. ACM Turing Award 1971.
1955
Co-founding the field
Co-authored the Dartmouth proposal and spent a career shaping the field. ACM Turing Award 1969.
1955
Co-writing the Dartmouth proposal
IBM engineer and co-author of the 1955 Dartmouth proposal that named artificial intelligence.
1956
Logic Theorist
Built the Logic Theorist (with Cliff Shaw), an early program that proved mathematical theorems. Shared the ACM Turing Award 1975 for basic contributions to AI, the psychology of human cognition, and list processing.
1958
The perceptron
Introduced the perceptron, one of the first neural networks that learned from examples.
1959
A program that learned to play checkers
Wrote a checkers program that improved by playing, described in “Some Studies in Machine Learning Using the Game of Checkers”.
1966
ELIZA, the first chatbot
Created ELIZA, an early program that held text conversations — and later became one of AI’s most thoughtful critics.
1982
Hopfield networks
Showed how a network of simple units can store and recall patterns as associative memory. Nobel Prize in Physics 2024, shared with Geoffrey Hinton.
1986
Backpropagation goes mainstream
Showed in Nature how back-propagating errors lets multi-layer networks learn useful internal representations.
1988
Bayesian networks & causal reasoning
Gave AI a rigorous way to reason under uncertainty — Bayesian networks — and later about cause and effect. ACM Turing Award 2011.
1994
Large-scale AI systems
Pioneered large-scale AI systems — Feigenbaum with expert systems, Reddy with speech and vision — and showed AI could be practically useful. Shared the ACM Turing Award 1994.
1997
Long short-term memory (LSTM)
Invented the LSTM, a recurrent network that can remember information over long sequences.
1998
Convolutional networks (LeNet)
Developed convolutional neural networks for reading handwriting, published with Bottou, Bengio and Haffner in “Gradient-based learning applied to document recognition”.
2009
ImageNet
Built ImageNet, a large labelled image database that became the benchmark that kick-started modern computer vision.
2012
AlexNet
Trained a deep convolutional network on GPUs that won the ImageNet challenge by a wide margin, kicking off the deep learning boom.
2014
Generative adversarial networks (GANs)
Introduced GANs: two networks trained against each other, one generating samples and one judging them — a foundation of generative AI.
2016
AlphaGo
Combined deep neural networks with tree search to master the game of Go, published in Nature.
2017
The Transformer
Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser and Illia Polosukhin introduced the Transformer in “Attention Is All You Need” — the architecture behind today’s large language models.
2018
Turing Award for deep learning
Shared the 2018 ACM Turing Award for decades of work that made deep neural networks practical.
2018
BERT
Showed that pre-training a Transformer on lots of text, then fine-tuning it, beats task-specific models across language benchmarks.
2020
GPT-3 and few-shot learning
Showed in “Language Models are Few-Shot Learners” that a very large language model can pick up new tasks from a few examples in the prompt.
2021
AlphaFold
Predicted protein structures with high accuracy using deep learning. Jumper and Hassabis shared half of the 2024 Nobel Prize in Chemistry (the other half went to David Baker).
2024
Nobel Prize in Physics
Shared the 2024 Nobel Prize in Physics with John Hopfield for foundational work on artificial neural networks — after co-authoring both the 1986 backpropagation paper and AlexNet.
2024
Turing Award for reinforcement learning
Laid the foundations of reinforcement learning, starting with a series of papers in the 1980s, and wrote its standard textbook. ACM Turing Award 2024.
2025
Kyoto Prize for the theory of learning machines
Since the 1960s, worked out the mathematics of learning machines: a theory of adaptive pattern classifiers (1967), self-organising networks of threshold elements, the natural gradient method (1998) and the field he named information geometry. Kyoto Prize in Advanced Technology 2025.
2025
GPU hardware that let machine learning scale
Shared the 2025 Queen Elizabeth Prize for Engineering for leading the GPU hardware and architecture advances that made it possible to scale machine-learning algorithms. The prize also went to Bengio, Hinton, Hopfield, Huang, LeCun and Li.
2025
Reasoning learned through reinforcement learning
Showed in Nature that a large language model’s reasoning can be incentivised through pure reinforcement learning, without human-labelled reasoning examples. Self-reflection, verification and changing strategy emerged during training.
2025
AlphaEvolve: a coding agent that discovers algorithms
Built AlphaEvolve, an evolutionary coding agent in which language models keep rewriting an algorithm’s code and automatic evaluators keep score. It found a way to multiply two 4×4 complex-valued matrices with 48 scalar multiplications, the first improvement over Strassen’s algorithm in that setting in 56 years.
2025
AlphaProof: formal maths proofs by reinforcement learning
Built AlphaProof, an AlphaZero-inspired agent that learns to find formal proofs in the Lean language through reinforcement learning. With AlphaGeometry 2, it reached a silver-medal score at the 2024 International Mathematical Olympiad. The paper was published in Nature in 2025.
2026
AlphaGenome: reading the regulatory code of DNA
Built AlphaGenome, a single model that reads 1 Mb of DNA and predicts thousands of functional genomic tracks down to single-base resolution. It matched or beat the strongest external models in 25 of 26 variant-effect evaluations.
2026
Multi-agent systems
Received the 2026 IJCAI Award for Research Excellence for decades of work on multi-agent systems: how autonomous software agents coordinate with each other and team up with people. He had won the IJCAI Computers and Thought Award in 1999.
2026
Computers and Thought Award
Received the 2026 IJCAI Computers and Thought Award, given to outstanding young AI scientists, for work across computer vision, generative AI and robotics that combines neural and symbolic representations so machines can understand and act in the physical world.
No geeks match these filters.
In September 2026 the US President proposed renaming AI to “SI”. We think it’s only fair that the people who actually built it get a name too.
“I will only encourage AI or, SI (SUPER INTELLIGENCE)!”
“Welcome to the new world of super intelligence - SI”
UN General Assembly speech, as reported by BBC News (opens in new tab) ·
“the executive branch shall use the terms “Super Intelligence” and “SI” in place of “Artificial Intelligence” and “AI””
Executive order “Inaugurating the Era of Super Intelligence”, whitehouse.gov (opens in new tab) ·
Quotes are verbatim from the linked pages. This site is about the people, not the politics.