Firsts
91 moments when AI did something for the first time, from 1956 to today, and why each one mattered. Hand-curated: the name is the quality filter, so every entry has to be a genuine first, with a source. The story runs in five eras; open one for the full entries.
The first generation of AI researchers worked on machines with kilobytes of memory and a belief that intelligence could be written down as symbols and rules. Programs proved logic theorems, played checkers, parsed small slices of English and modelled single neurons.
- 1956 It proved its first theorems
- 1957 It modeled the brain
- 1958 It got a language
- 1959 It learned to play checkers
- 1960 It learned to minimize error
- 1961 It learned calculus
- 1962 It beat a master
- 1963 It solved IQ tests
- 1964 It read word problems
- 1965 The first expert system
- 1965 The alchemy warning
- 1966 It learned to hold a conversation
- 1966 The money dried up for translation
- 1967 It learned to play in tournaments
- 1968 It learned to find the shortest path
- 1969 It was told to stop guessing
- 1970 It learned to manipulate a virtual world
- 1970 It became an "electronic person"
- 1971 It learned to make a plan
- 1972 It learned to reason in logic
- 1973 The first winter set in
- 1974 The American retreat
- 1975 It learned from evolution
- 1976 It learned to diagnose infections
- 1977 It learned to find ore
- 1978 It learned to configure computers
- 1979 It learned to navigate a room
- 1979 It beat a world champion
AI became an industry in this period, first through expert systems that encoded human knowledge as rules, then through statistical methods that learned patterns from data. It saw a second winter, and the quiet revival of neural networks and backpropagation.
- 1980 It modeled the visual cortex
- 1980 The community gathered
- 1981 It became a national priority
- 1982 It learned to associate
- 1983 It tried to think like us
- 1984 The winter warning
- 1985 It learned to handle uncertainty
- 1986 It learned by its mistakes
- 1987 The second winter begins
- 1988 It learned to calculate the odds
- 1989 It learned to read the mail
- 1990 It learned to translate by counting
- 1991 It planned a war
- 1992 It taught itself to play
- 1992 It drew a line in the data
- 1993 It learned to translate with statistics
- 1994 It won a world championship
- 1995 It drove across America
- 1996 It took a game from the champion
- 1997 It beat the world champion
- 1997 It learned to take dictation
- 1997 It learned to remember
- 1998 It learned to read handwriting
- 1999 It became a pet
- 1999 It learned to optimize policies
- 2000 It learned by watching
- 2000 It learned to walk like us
- 2001 It learned the value of data
- 2002 It learned to grade itself
- 2003 It learned the shape of language
- 2004 It crashed in the desert
- 2005 It crossed the desert
- 2006 It learned to train deep nets
- 2007 The hardware arrived
- 2008 It learned to multitask
- 2009 The data caught up
- 2010 Vision enters the living room
- 2011 It won at trivia
- 2011 It became a personal assistant
Deep learning broke through when neural networks, large datasets and GPUs finally came together. Image recognition leapt forward, machines learned to play games from raw pixels, and the transformer architecture arrived in 2017.
- 2012 It learned to see
- 2013 It mapped meaning
- 2013 It learned to play from pixels
- 2014 It learned to imagine
- 2014 It learned to translate
- 2015 It learned to go deep
- 2016 It learned to play Go
- 2017 It learned to pay attention
- 2017 It learned the rules of the game
- 2018 It learned to read in both directions
- 2019 It learned to write too well
Scale changed what models could do. Systems began learning tasks from a handful of examples, predicting protein structures, drawing images from text and holding conversations that millions of people tried for themselves.
The current frontier is models that reason before answering, use software on a person’s behalf, and increasingly arrive as open weights. Entries here are recent, so they are the ones most likely to be revised as the record settles.
Feed, daily deep-dive and bytes — readable offline, with push alerts for the topics you follow.