Also: AI, machine learning, generative AI, IndiaAI Mission, deep learning
Definition
The field of building computer systems that perform tasks usually needing human intelligence, such as recognising speech and images, translating languages, making predictions and generating text, mostly by learning patterns from large amounts of data.
The field of AI was named at a workshop at Dartmouth College in 1956, building on Alan Turing's 1950 question of whether machines can think. Early AI relied on hand-written rules and logic. Progress was uneven, with periods of reduced funding known as "AI winters" when promises outran results. From the 1990s, machine learning, in which systems learn from data rather than explicit rules, gained ground. In 1997 IBM's Deep Blue defeated chess champion Garry Kasparov.
The breakthrough of deep learning came in the 2010s, when neural networks with many layers, trained on large datasets using powerful graphics processors, achieved large gains in image and speech recognition. In 2016 DeepMind's AlphaGo beat a top Go player. The transformer architecture, introduced in 2017, led to large language models, and the public release of generative AI chatbots from late 2022 brought AI into everyday use. India published a national AI strategy through NITI Aayog in 2018 under the theme #AIForAll, and approved the IndiaAI Mission in 2024 to build computing capacity, datasets, skills and indigenous models.
Create a free account to read 5 full pages a month, or upgrade for unlimited access. The definition above, the A–Z index and the pop-up definitions in lessons stay free.
