|
|
| Line 213: |
Line 213: |
| | |} | | |} |
| | |}<!-- B --> | | |}<!-- B --> |
| − |
| |
| − | = Modern Timeline - to July 2023 =
| |
| − | * [https://en.wikipedia.org/wiki/Timeline_of_artificial_intelligence Timeline of artificial intelligence | Wikipedia]
| |
| − |
| |
| − | Here is a timeline of the history of Artificial Intelligence (AI):
| |
| − |
| |
| − | * 1943: Warren McCulloch and Walter Pitts propose a computational model for neural networks, which lays the foundation for artificial neural networks.
| |
| − | * 1950: Alan Turing introduces the "Turing Test," a criterion to determine a machine's ability to exhibit intelligent behavior, marking an important milestone in AI.
| |
| − | * 1956: John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon organize the Dartmouth Conference, widely considered the birth of AI as a field of study. The term “artificial intelligence” was first coined by John McCarthy in 1956 when he held the first academic conference on the subject at Dartmouth College.
| |
| − | * 1956-1974: This period is known as the "Golden Age" of AI research. Researchers develop early AI programs, including the Logic Theorist, General Problem Solver, and ELIZA.
| |
| − | * 1958: John McCarthy invents the programming language LISP, which becomes a popular language for AI research and development.
| |
| − | * 1965: Joseph Weizenbaum creates ELIZA, a computer program that simulates conversation, and it becomes one of the first examples of [[Natural Language Processing (NLP)]].
| |
| − | * 1966: Ray Solomonoff introduces the concept of algorithmic probability, a fundamental concept in machine learning and AI.
| |
| − | * 1969: The Stanford Research Institute develops Shakey, a mobile robot capable of reasoning and problem-solving, considered a significant advancement in robotics and AI.
| |
| − | * 1973: The Lighthill Report is published, criticizing the progress in AI research and leading to a decrease in funding, marking the end of the first AI winter.
| |
| − | * 1980s-1990s: Expert systems gain prominence, focusing on capturing and replicating human expertise in narrow domains. Symbolic AI approaches become popular during this time.
| |
| − | * 1986: Geoffrey Hinton, David Rumelhart, and Ronald Williams publish a paper on backpropagation, which greatly advances the training of artificial neural networks.
| |
| − | * 1997: [[IBM]]'s Deep Blue defeats Garry Kasparov, the world chess champion, marking a significant milestone in AI and demonstrating the potential of machine intelligence.
| |
| − | * 2000s: The focus shifts from symbolic AI to statistical approaches, and machine learning algorithms, particularly neural networks, gain popularity.
| |
| − | * 2011: [[IBM]]'s Watson defeats human champions in the quiz show Jeopardy!, demonstrating advancements in [[Natural Language Processing (NLP)]] and question-answering systems.
| |
| − | * 2012: AlexNet, a [[(Deep) Convolutional Neural Network (DCNN/CNN)|deep convolutional neural network]], achieves a breakthrough in image classification, leading to a surge in deep learning research and applications.
| |
| − | * 2011-2014: [[Apple]]'s Siri (2011), Google's Google Now (2012) and Microsoft's Cortana (2014) are smartphone apps that use [[Natural Language Processing (NLP)]] to answer questions, make recommendations and perform actions.
| |
| − | * 2014: [[Attention]] mechanism is developed, that lead to the Transformer architecture.
| |
| − | * 2015: [[Google|DeepMind's AlphaGo]] defeats the world champion Go player, Lee Sedol, showcasing the power of deep reinforcement learning in complex games.
| |
| − | * 2015: [[Google]] released TensorFlow, an open-source software library for machine learning.
| |
| − | * 2016: [[OpenAI]] introduces GPT (Generative Pre-trained Transformer), an advanced language model that generates human-like text and significantly advances [[Natural Language Processing (NLP)]].
| |
| − | * 2017: The term "Deep[[fake]]" emerges, referring to the use of AI to create realistic but fake audio and video content, raising concerns about misinformation and privacy.
| |
| − | * 2018: The General Data Protection Regulation (GDPR) is implemented in the European Union, introducing regulations for the ethical use of AI and data privacy.
| |
| − | * 2019 March: First human brain-computer interface: A team of researchers at the University of California, San Francisco (UCSF) successfully implanted a [[Connecting Brains|brain-computer interface (BCI)]] in a human patient. The BCI allowed the patient to control a cursor on a computer screen simply by thinking about it.
| |
| − | * 2019: [[OpenAI]] released GPT-2, a large-scale language model that can generate coherent and diverse texts on various topics
| |
| − | * 2020: GPT-3, an even more advanced version of the language model, is released, demonstrating unprecedented capabilities in generating coherent and contextually relevant text.
| |
| − | * 2020: [[Baidu]] released the LinearFold AI algorithm to help medical and scientific teams developing a vaccine during the early stages of the [[COVID-19|SARS-CoV-2 (COVID-19) pandemic]]
| |
| − | * 2020: [[Protein Folding & Discovery#Google DeepMind AlphaFold|Google DeepMind AlphaFold 2]] achieved a breakthrough in protein structure prediction, surpassing the human experts in the CASP14 competition
| |
| − | * 2021: [[OpenAI]] released [[Video/Image#DALL-E | DALL-E]], a generative model that can create images from text descriptions, such as “an armchair in the shape of an avocado”
| |
| − | * 2021: [[Google]]’s LaMDA (Language Model for Dialogue Applications) demonstrated natural and engaging conversations on various topics, such as Pluto and paper airplanes.
| |
| − | * 2021: [[Microsoft]]’s Turing-NLG ([[Natural Language Processing (NLP)]]) model generated realistic and fluent texts on various domains, such as news, reviews, and fiction.
| |
| − | * 2022 January: [[OpenAI]] released [[Video/Image#DALL-E | DALL-E]] 2, a generative model that can create images from text descriptions, such as "an armchair in the shape of an avocado". The model demonstrated remarkable creativity and diversity in its outputs, as well as the ability to manipulate attributes and perspectives.
| |
| − | * 2022 February: [[Protein Folding & Discovery#Google DeepMind AlphaFold|Google DeepMind AlphaFold 2]] achieved a breakthrough in protein structure prediction, surpassing the human experts in the CASP14 competition. The model used deep learning to predict the three-dimensional shape of proteins from their amino acid sequences, with unprecedented accuracy and speed. This could have huge implications for drug discovery, biotechnology, and understanding diseases.
| |
| − | * 2022 April: [[Google]]'s [[Lambda]] (Language Model for Dialogue Applications) demonstrated natural and engaging conversations on various topics, such as Pluto and paper airplanes. The model used a neural network to generate responses that were relevant, specific, and informative, as well as sometimes humorous and surprising. The model could also switch topics seamlessly and handle multiple languages.
| |
| − | * 2022 June: [[Microsoft]]'s Turing-NLG ([[Natural Language Processing (NLP)]]) model generated realistic and fluent texts on various domains, such as news, reviews, and fiction. The model used a massive neural network with 17 billion parameters to produce coherent and diverse texts from keywords or prompts. The model could also answer questions, summarize texts, and rewrite sentences.
| |
| − | * 2022 September: HyperTree Proof Search (HTPS) was introduced as a new algorithm for automated theorem proving, a challenging task in mathematics and logic. The algorithm used a novel data structure called HyperTree to efficiently search for proofs of mathematical statements, outperforming existing methods on several benchmarks. The algorithm could potentially help discover new mathematical truths and verify complex systems.
| |
| − | * 2023 January: A team of researchers at [[Google]] AI developed a new type of AI that can learn to perform tasks by watching humans do them. This AI, called "Plato," is able to learn to perform tasks that are much more complex than previous AIs. For example, Plato was able to learn to play the game of Go after watching just a few hours of human gameplay.
| |
| − | * 2023 February: A team of researchers at [[Google|DeepMind]] developed a new type of AI that can solve complex mathematical problems. This AI, called "[[Protein Folding & Discovery#Google DeepMind AlphaFold|Google DeepMind AlphaFold]]," was able to solve the structure of the [[COVID-19|SARS-CoV-2 virus]], which could help scientists to develop new treatments for [[COVID-19]].
| |
| − | * 2023 March: [[Video/Image#DALL-E | DALL-E]] 2: A team of researchers at [[OpenAI]] developed a new type of AI that can generate realistic and creative text. This AI, called "[[Video/Image#DALL-E | DALL-E]] 2," can be used to create images from text descriptions, or to generate text that describes images.
| |
| − | * 2023 April: A team of researchers at Stanford University developed a new type of AI that can translate languages more accurately than previous AIs. This AI, called "[[Bard]]," was able to translate between 26 languages with an accuracy of 99%.
| |
| − | * 2023 May: A team of researchers at the University of California, Berkeley, developed a new type of AI that can detect and classify diseases from medical images. This AI, called "Enlitic," was able to detect cancer with an accuracy of 95%.
| |
| − | * 2023 June: A team of researchers at [[NVIDIA]] developed a new type of AI that can control [[Robotics|robots]] in real time. This AI, called "Isaac," was able to control a [[Robotics|robot]] arm to perform complex tasks with human-level accuracy.
| |
