wAIpoint
Italiano

88 milestones, from 1950 to 2025

Each milestone is the first public version of a cited work. Filter by region.

  1. 1950Shannon: programming a computer for playing chess. Adversarial search & games source
  2. 1950Arrow: a difficulty in the concept of social welfare. Collective decisions source
  3. 1950Nash: equilibrium points in n-person games. Non-cooperative game theory source
  4. 1950Turing: computing machinery and intelligence. Can machines really think? source
  5. 1954Shapley and Shubik: a method for evaluating the distribution of power in a committee system. Cooperative game theory source
  6. 1956Chomsky: three models for the description of language. Grammar & parsing source
  7. 1957Bellman: a Markovian decision process. Decisions: MDPs, bandits, POMDPs source
  8. 1958Rosenblatt: the perceptron, a probabilistic model for information storage and organization in the brain. Feedforward networks source
  9. 1960Kalman: a new approach to linear filtering and prediction problems. Reasoning over time source
  10. 1962Davis, Logemann and Loveland: a machine program for theorem-proving. Logical agents & propositional logic source
  11. 1965Robinson: a machine-oriented logic based on the resolution principle. First-order logic & inference source
  12. 1966Graham: bounds for certain multiprocessing anomalies. Time, schedules & resources source
  13. 1967Nearest neighbor pattern classification (Cover and Hart). Nonparametric models (kernels, nearest neighbors) source
  14. 1968Hart, Nilsson and Raphael: a formal basis for the heuristic determination of minimum-cost paths. Uninformed & heuristic search source
  15. 1971STRIPS: applying theorem proving to problem solving (Fikes and Nilsson). Classical planning source
  16. 1972Colmerauer's group builds the first versions of Prolog, from a project on processing French. Logic programming source
  17. 1977Mackworth: consistency in networks of relations. Constraint satisfaction source
  18. 1977The EM algorithm: maximum likelihood from incomplete data (Dempster, Laird, Rubin). Statistical & probabilistic learning (EM) source
  19. 1980Searle: minds, brains, and programs. Can machines really think? source
  20. 1983Simulated annealing: optimization inspired by statistical mechanics (Kirkpatrick et al.). Local search & optimization source
  21. 1986Langton: studying artificial life with cellular automata. Artificial life source
  22. 1986Learning representations by back-propagating errors (Rumelhart, Hinton, Williams). Training & backpropagation source
  23. 1988Lauritzen and Spiegelhalter: local computations with probabilities on graphical structures, applied to expert systems. Bayesian networks source
  24. 1989Rabiner: hidden Markov models and their applications in speech recognition. Speech recognition source
  25. 1989Rabiner: a tutorial on hidden Markov models and their applications in speech recognition. Reasoning over time source
  26. 1991Brooks: intelligence without representation. Agent architectures (reflex, goal, utility, learning) source
  27. 1992Williams: simple statistical gradient-following algorithms for connectionist reinforcement learning. Policy search source
  28. 1992Q-learning (Watkins and Dayan). Passive & active reinforcement learning source
  29. 1993Gruber: a translation approach to portable ontology specifications. Knowledge representation & ontologies source
  30. 1995Wooldridge and Jennings: intelligent agents, theory and practice. Rational agents source
  31. 1995Support-vector networks (Cortes and Vapnik). Supervised learning (regression, classification, trees) source
  32. 1996Kavraki et al.: probabilistic roadmaps for path planning in high-dimensional configuration spaces. Planning & control source
  33. 1997Boosting: a decision-theoretic generalization of on-line learning (Freund and Schapire). Ensemble learning source
  34. 1997Long short-term memory (Hochreiter and Schmidhuber). Recurrent networks source
  35. 1998Gradient-based learning applied to document recognition (LeCun, Bottou, Bengio, Haffner). Convolutional networks source
  36. 1998Planning and acting in partially observable stochastic domains (Kaelbling, Littman, Cassandra). Planning in nondeterministic domains source
  37. 1999Options: a framework for temporal abstraction in reinforcement learning (Sutton, Precup, Singh). Hierarchical planning source
  38. 2000Zhang: a flexible new technique for camera calibration. Image formation & features source
  39. 2001Random forests (Breiman). Ensemble learning source
  40. 2003Miller: the cognitive revolution, a historical perspective. Cognitive science source
  41. 2004Apprenticeship learning via inverse reinforcement learning (Abbeel and Ng). Apprenticeship & inverse RL source
  42. 2006Durrant-Whyte and Bailey: simultaneous localization and mapping, part I. Robotic perception source
  43. 2006Hinton and Salakhutdinov: reducing the dimensionality of data with neural networks (autoencoders). Unsupervised learning (clustering, dimensionality reduction) source
  44. 2008Goodrich and Schultz: Human–Robot Interaction, a survey. Humans & robots source
  45. 2013R-CNN: convolutional networks applied to region proposals for object detection (Girshick et al.). Object detection source
  46. 2013Deep Q-learning: a convolutional network learns to play Atari games from raw pixels using reinforcement learning (Mnih et al.). Passive & active reinforcement learning source
  47. 2013word2vec: efficient ways to learn vector representations of words from very large text collections (Mikolov et al.). Word embeddings source
  48. 2014Soft attention added to encoder-decoder translation models (Bahdanau, Cho, Bengio). Attention mechanism source
  49. 2014GANs: a generator and a discriminator trained against each other in a two-player game (Goodfellow et al.). Generative adversarial networks (GANs) source
  50. 2014Sequence to sequence learning with LSTM networks (Sutskever, Vinyals, Le). Sequence-to-sequence models source
  51. 2014Adam: a method for stochastic optimization (Kingma and Ba). Training & backpropagation source
  52. 2015ResNet: residual learning makes much deeper networks easier to train (He et al.). Convolutional networks source
  53. 2015ResNet: residual networks for image recognition (He et al.). Image classification source
  54. 2015YOLO: object detection as a single regression problem (Redmon et al.). Object detection source
  55. 2015End-to-end training of deep visuomotor policies from camera images to motor torques (Levine et al.). Reinforcement learning in robotics source
  56. 2015YOLO: object detection as one regression from full images (Redmon et al.). YOLO (real-time object detection) source
  57. 2016Mastering the game of Go with deep neural networks and tree search (Silver et al.). Adversarial search & games source
  58. 2016WaveNet: a generative model for raw audio (van den Oord et al.). Audio & speech processing source
  59. 2016Concrete problems in AI safety (Amodei et al.). The ethics of AI source
  60. 2016Understanding deep learning requires rethinking generalization (Zhang et al.). Generalization source
  61. 2017The Transformer: a network built only on attention (Vaswani et al.). Attention mechanism source
  62. 2017Evolution strategies as a scalable alternative to reinforcement learning (Salimans et al.). Evolutionary algorithms source
  63. 2017Introduced in "Attention Is All You Need" (Vaswani et al.). Transformer architecture source
  64. 2018BERT: pre-trained deep bidirectional transformers for language understanding (Devlin et al.). Large language models source
  65. 2018GLUE: a multi-task benchmark for natural language understanding (Wang et al.). NLP tasks (translation, extraction, dialogue) source
  66. 2018BERT: pre-training of deep bidirectional transformers for language understanding. Pretraining & transfer learning source
  67. 2018Pyro: a probabilistic programming language built on Python and PyTorch (Bingham et al.). Probabilistic programming source
  68. 2018BERT: pre-training deep bidirectional transformers on unlabelled text, then adapting them (Devlin et al.). Unsupervised & transfer learning source
  69. 2020DETR: object detection as direct set prediction with a transformer (Carion et al.). DETR (detection with transformers) source
  70. 2020Denoising diffusion probabilistic models: high-quality image synthesis with diffusion models (Ho et al.). Diffusion & image/video generation source
  71. 2020ViT: a pure transformer for image recognition at scale (Dosovitskiy et al.). Image classification source
  72. 2020GPT-3: scaling up language models for few-shot learning (Brown et al.). Language models source
  73. 2020GPT-3: a 175-billion-parameter language model that learns tasks from a few examples (Brown et al.). Large language models source
  74. 2020RAG: combine a pre-trained language model with a retrievable external memory for generation (Lewis et al.). Retrieval-augmented generation source
  75. 2020Scaling laws: language-model loss follows power laws in model size, data and compute (Kaplan et al.). Scaling laws source
  76. 2020NeRF: scenes as neural radiance fields for view synthesis (Mildenhall et al.). The 3D world source
  77. 2020Vision Transformer: a pure transformer on image patches matches state-of-the-art convolutional networks on classification (Dosovitskiy et al.). Vision Transformer (ViT) source
  78. 2021LoRA: adapt a huge pre-trained model by training only small low-rank matrices, freezing the original weights (Hu et al.). Fine-tuning & alignment source
  79. 2021On the dangers of stochastic parrots: can language models be too big? (Bender et al.). The limits of AI source
  80. 2021CLIP: learning visual models from natural-language supervision, with 400 million (image, text) pairs (Radford et al.). Multimodal models source
  81. 2022InstructGPT: aligning language models with user intent by fine-tuning with human feedback (Ouyang et al.). Fine-tuning & alignment source
  82. 2022Chain-of-thought prompting: showing intermediate reasoning steps improves reasoning in large language models (Wei et al.). Chain-of-thought reasoning source
  83. 2022ReAct: language models interleave reasoning traces and actions to use external sources (Yao et al.). LLM-based agents & tool use source
  84. 2022Whisper: robust speech recognition via large-scale weak supervision (Radford et al.). Speech recognition source
  85. 2023Generative agents: an LLM-based architecture with memory, reflection and planning for believable behaviour (Park et al.). Agent patterns: planning, memory, multi-agent source
  86. 2023Toolformer: language models teach themselves which tools (APIs) to call and how (Schick et al.). LLM-based agents & tool use source
  87. 2024Anthropic introduces the Model Context Protocol, an open standard for connecting data sources and AI tools. Model Context Protocol (MCP) source
  88. 2025YOLOv12: an attention-centric real-time detector (Tian, Ye, Doermann). YOLO (real-time object detection) source