| 1956-1970: Appearance of term (1956); AI used to solve complex mazes and puzzles (non-business cases) (1965); expert systems emerged (1965) | 1971-1996: AI is in exploratory stage; academic interest attracted; computers were expensive; expert systems (using fuzzy logic due to Zadeh, L.) emerged stronger; Windows and Mac appeared; computer prices fell; society became computer-literate | 1997-2009 AI proved superior than best humans in certain games (Othello and Chess); rule-based processing systems emerged (1980-2000) |
| 2010-2020: Machines learned and deeply; the convolutional neural networks (CNN) emerged; audio classification appeared (1988) and image (1989); AI closely related now to business; AI can be used in seamen selection providing a match between ship job & applicant; AI helped by Internet and data and big data; and by increased computer memory (1993); and by the use of GPUs! | 2010-2020: Machines used for predictions in visual recognition & image recognition in advertising; prediction of weather prevailing ahead, wind’s speed, wave’s height; Captain can thus correlate data to the age of ship etc. and he/she may prefer a refugee port than ship’s original route; recognition of: hand written characters (optical character recognition) | 2010-2020: Recognition of defective products and spare parts; AI can be used to collect data on charterers; to predict seamen and ship accidents; AI can analyze information on ship and her equipment, where fuel economies can reach $20m per annum for a fleet of 35 ships; technical problems can also be pre-known; logs on ship performance and on her condition can be monitored 24/7 and detect abnormalities |