Intelligent machines
INTELLIGENT MACHINES, which can learn about their surroundings, express emotion, use reason, play and think, have been the plot source for loads of science fiction movies.
In the The Matrix, humans go to war against artificially intelligent machines that are powered by the humans grown in pods. In Stanley Kubrick's mystifying masterpiece, 2001: A Space Odyssey(pictured above) super computer HAL has an emotional breakdown on a mission to Jupiter. In The Terminator, Arnold Schwarzenegger is an intelligent robot assassin sent from the future to destroy John Connor, the leader of the human resistance, before he is even born. And in A.I . Artificial Intelligence, Haley Joel Osment plays a robot boy, programmed to feel love, who finds it difficult to comprehend his non-human existence.
In all of these films, machines are able to think, feel, learn, predict, communicate, adapt, plan, reason and generally mimic the emotional and intellectual behaviour of humans.
Could such an intelligent machine be invented? The answer is most likely yes: intelligent computer programs are already in use every day. But creating a fully functioning artificially intelligent being is a bit more complicated.
The term artificial intelligence (or AI) was coined by American scientist John McCarthy in 1956. Back then, scientists wanted to create computer software programs that thought like humans – which could solve algebra problems, play and learn logical games like chess, solve puzzles and even engage in conversation. The focus was on creating intelligent software rather than intelligent machines and relied on the idea that humans used step-by-step deduction to solve problems. We now realise that the human thought process is a lot more complicated than that.
AI software has advanced hugely in the last 50 years and is now found in many household objects including your mobile phone, PlayStation or Xbox and even in cars. Banks use artificial intelligence programmes to invest in shares and manage work rotas.
Voice recognition software in your phone or PC, learns the way you speak, enabling you to communicate with your computer or phone in a new way. Even your email learns what spam is and puts it in a special folder.
Simple toys like Tamagotchis or the Furby utilised AI software to gain knowledge and learn the more they were used.
Many combat or sports games for the Xbox or PlayStation get more difficult the more you play them and seemingly learn your style of play – giving the game a “real” sense of intelligence. You can even download apps for your iPhone where you can create your own virtual personality (MyBot Creator for those interested).
But creating a fully functioning intelligent machine (or robot) is a different matter, mainly because intelligence is a very difficult thing to define. It is multifaceted and multilayered, and takes many forms, including emotional, logical, musical and linguistic.
The robot would need to learn, problem-solve, reason and plan – all difficult to achieve. It would also need to be able to express emotions and interact socially with humans and other intelligent robots.
This is the most difficult feature to achieve and potentially the most risky. As well as love, artificially intelligent machines could be programmed to feel anger, fear or even hatred. Potentially (but not very likely) we could see the scenarios portrayed in the aforementioned movies played out.
Our understanding of human consciousness is far from complete and we would need a much greater comprehension of its origins before we could artificially create an emotional machine.
So while we may have “intelligent” phones, games and computer programs, we don’t yet have machines that can behave like humans, think like us, feel emotions, learn or communicate like humans.
What we do have though are loads of brilliant movies and books, which can stir our imaginations and make us wonder “what if . . . ?”
The Frog Blog:blog.scccscience.com
