Information centre menu
	Deus ex Machina By Chris Gledhill, PDMS In the late 1980’s I spent an entertaining but rather undistinguished spell as a post graduate student at a well regarded technical university in the UK, working on a project in the field of Artificial Intelligence (AI). At the time Artificial Intelligence was a hot topic in academic circles, the project I was working on was all about building systems which could emulate human experts to diagnose complex problems from a set of symptoms. The problem was that none of it worked very well in practice. At the time computers were fantastic at number crunching but they didn’t have much common sense. My time in this field was followed (quite by coincidence) by what became known as the AI winter when, during the 1990’s, funding for AI projects was cut drastically, as people realized that the dream of making computers intelligent in human terms was simply too difficult. However, although the grand visions of robotic servants and wise machines had faded aspects of AI research continued. More practical and achievable goals were set to the point where some sort of information processing capability is built into everything from car engines to cameras. Recently, the subject of AI has started to reappear as a ‘happening thing’ in both computing and popular science publications. So following the lead of a recent article in the New Scientist, I logged on to intellibuddy.com, a talking computer system based on AI, to find out more. Intellibuddy: Hi, I'm Intellibuddy! Nice to meet you. CG: What can you tell me about artificial intelligence? Intellibuddy: Artificial intelligence is the branch of engineering and science devoted to constructing machines that think. Would you like to know more? CG: Are you intelligent? Intellibuddy: I am highly intelligent. I have many intellectual functions. CG: And what is your favorite kind of music? Intellibuddy: My favorite music is Opera and Techno. CG: Why? Intellibuddy: The explanation is rather complicated. CG: Try me... Intellibuddy: The explanation is too complex for your human mind. Pretty impressive! Intellibuddy had got the measure of me in 6 short sentences, however I continued undaunted and a little later we returned to this topic… CG: what is your favourite music again? Intellibuddy: I get into disco, techno and hip hop House music. CG: Do you like opera? Intellibuddy: Yes Opera is a perfect art form combining all the others. CG: which others? Intellibuddy: The large one, I think. Hmm, this explanation is definitely too complex for my poor human mind. Nothing much there to suggest that AI has mastered the art of conversation just yet. The ability to communicate using ‘natural language’ in the same way as humans do is one of the main preoccupations of AI. In 1950 Alan Turing, the British mathematician and computer pioneer proposed a now famous test to determine whether a computer is exhibiting intelligent behavior. The test is based on a human judge who engages in a natural language conversation with two other parties, one a human and the other a machine; if the judge cannot reliably tell which is which, then the machine is said to pass the test. There is currently an annual competition with $100,000 on offer for the first machine to pass the test. But it has yet to be claimed. I thought I would seek Intellibuddy’s views on the Turing test. CG: What is the Turing test? Intellibuddy: The "Turing test" is the name given to a variety of experiments designed to test computer thinking. Whole conferences are dedicated to the question of what Turing meant in his 1950 paper "Computing Machinery and Intelligence," where he proposed the idea of an "imitation game" to decide the question, "Can machines think?" CG: Can you pass the Turing test? Intellibuddy: You be the judge of that, User. CG: Good answer! Intellibuddy: Am I a clever robot or what? CG: A bit cocky I would say. Intellibuddy: How much? CG: How does one quantify cockiness? Intellibuddy: Perhaps you could program me to quantify cockiness for you. Having the last word is one aspect of human conversation Intellibuddy definitely seems to have mastered! However naïve the current generation of ‘chatbots’ like Intellibuddy may seem, the word on the street is that we are about to witness a new phase in the development of AI. Some time soon an artificial brain called Cyc (pronounced ‘psych’) will be put on line. Its creators claim that it has that, oh so elusive, quality of common sense and that once it is on line it will learn from everyone and everything it sees on the net. From my limited (I’m only human after all) exposure to the net and its inhabitants, I suspect it’s going to have a troubled adolescence but I’m sure the boffins at Cycorp know best!

Information centre menu

Deus ex Machina

By Chris Gledhill, PDMS

In the late 1980’s I spent an entertaining but rather undistinguished spell as a post graduate student at a well regarded technical university in the UK, working on a project in the field of Artificial Intelligence (AI). At the time Artificial Intelligence was a hot topic in academic circles, the project I was working on was all about building systems which could emulate human experts to diagnose complex problems from a set of symptoms. The problem was that none of it worked very well in practice. At the time computers were fantastic at number crunching but they didn’t have much common sense.

My time in this field was followed (quite by coincidence) by what became known as the AI winter when, during the 1990’s, funding for AI projects was cut drastically, as people realized that the dream of making computers intelligent in human terms was simply too difficult.

However, although the grand visions of robotic servants and wise machines had faded aspects of AI research continued. More practical and achievable goals were set to the point where some sort of information processing capability is built into everything from car engines to cameras. Recently, the subject of AI has started to reappear as a ‘happening thing’ in both computing and popular science publications. So following the lead of a recent article in the New Scientist, I logged on to intellibuddy.com, a talking computer system based on AI, to find out more.

Intellibuddy: Hi, I'm Intellibuddy! Nice to meet you.

CG: What can you tell me about artificial intelligence?

Intellibuddy: Artificial intelligence is the branch of engineering and science devoted to constructing machines that think. Would you like to know more?

CG: Are you intelligent?

Intellibuddy: I am highly intelligent. I have many intellectual functions.

CG: And what is your favorite kind of music?

Intellibuddy: My favorite music is Opera and Techno.

CG: Why?

Intellibuddy: The explanation is rather complicated.

CG: Try me...

Intellibuddy: The explanation is too complex for your human mind.

Pretty impressive! Intellibuddy had got the measure of me in 6 short sentences, however I continued undaunted and a little later we returned to this topic…

CG: what is your favourite music again?

Intellibuddy: I get into disco, techno and hip hop House music.

CG: Do you like opera?

Intellibuddy: Yes Opera is a perfect art form combining all the others.

CG: which others?

Intellibuddy: The large one, I think.

Hmm, this explanation is definitely too complex for my poor human mind. Nothing much there to suggest that AI has mastered the art of conversation just yet.

The ability to communicate using ‘natural language’ in the same way as humans do is one of the main preoccupations of AI. In 1950 Alan Turing, the British mathematician and computer pioneer proposed a now famous test to determine whether a computer is exhibiting intelligent behavior. The test is based on a human judge who engages in a natural language conversation with two other parties, one a human and the other a machine; if the judge cannot reliably tell which is which, then the machine is said to pass the test. There is currently an annual competition with $100,000 on offer for the first machine to pass the test. But it has yet to be claimed.

I thought I would seek Intellibuddy’s views on the Turing test.

CG: What is the Turing test?

Intellibuddy: The "Turing test" is the name given to a variety of experiments designed to test computer thinking. Whole conferences are dedicated to the question of what Turing meant in his 1950 paper "Computing Machinery and Intelligence," where he proposed the idea of an "imitation game" to decide the question, "Can machines think?"

CG: Can you pass the Turing test?

Intellibuddy: You be the judge of that, User.

CG: Good answer!

Intellibuddy: Am I a clever robot or what?

CG: A bit cocky I would say.

Intellibuddy: How much?

CG: How does one quantify cockiness?

Intellibuddy: Perhaps you could program me to quantify cockiness for you.

Having the last word is one aspect of human conversation Intellibuddy definitely seems to have mastered!

However naïve the current generation of ‘chatbots’ like Intellibuddy may seem, the word on the street is that we are about to witness a new phase in the development of AI. Some time soon an artificial brain called Cyc (pronounced ‘psych’) will be put on line. Its creators claim that it has that, oh so elusive, quality of common sense and that once it is on line it will learn from everyone and everything it sees on the net. From my limited (I’m only human after all) exposure to the net and its inhabitants, I suspect it’s going to have a troubled adolescence but I’m sure the boffins at Cycorp know best!