The Turing test was devised as a bar for whether a computer could fool a person that they were talking to another person
The stories are a series about humans, robots and morality and centres around the technophobia (trying to prove it was misplaced)
Top computer scientists, including John McCarthy met to discuss AI and its potential. Mix of approaches including programming to mimic human behaviour, neutral networks (emulating brain cells) to learn.
Stanley Kubrik's film about a sentient machine: HAL 9000 that controls flight which claims to be "fool-proof and incapable of error". It captured the public fear of a robot takeover
'Shakey the robot' was able to reason and make its own decisions and developed a map of where it was. It was too slow and had to remap every time it moved forward and could be prevented from working by a moving object in its field of view. It took 6 years to develop.
Millions spent up to the 1970s with next to no actual development, Professor James Lighthill gave a damning speech in the UK saying simple tasks like face recognition will always be beyond their capability and the most avanced they could get was playing an "experienced amateur" level of chess. AI budgets were slashed worldwide
New investment for AI, less ambitious but creating expert systems
The first film shows society being overthrown by Skynet AI, building on fear
Caused by over-expectation of short-term outcomes for the technology, which didn't happen and caused budget slashes
Expert systems couldn't work like a human as they weren't build like them (in different chunks all doing different functions), drove back to research of neural networks
IBM's Deep Blue supercomputer took on the world's chess champion and beat him
Vacuum that navigates the room looking for dirt, it learns the layout of the area and will automatically go around and clean it
Built by Boston Dynamics, the BigDog was built to navigate terrain that vehicles couldn't. The company also makes a bomb disposing robot that can 'sniff' explosives, over 2000 of these were deployed in Afghanistan
The AI powered vehicle was able to travel across the desert autonomously for 131 miles
Released with 80% accuracy, able to allow users to search using their voice. It improves itself with the vast amounts of data it receives. As of 2015, Google says that it is 92% accurate
New humanoid robots used technology such as neural networks to allow them to observe their surroundings. 20 of these danced in perfect harmnoy af the 2010 world expo in harmony
IBM's Watson participated in Jeopardy, it was trained for 3 years to allow it to recognise questions. Watson defeated its opposition easily using vast amounts of mined data
Looking through millions of Google Images and YouTube videos, it learned what a cat looked like
64 years after the Turing Test was devised by Alan Turing. The outcome was the chatbot 'Eugene Goostman' was 'taught for the test' using tricks to fool judges.