 |
|
|
AUDIENCE COMMENTS |
|
|
|
|
An opportunity for the audience to have their say on In Our Time. |
|
 |
|
|
|
ARTIFICIAL INTELLIGENCE
Martin Wheatman - AI
One point, which possibly marred an otherwise excellent programme, was that it was stated that an AI would be easily detected in a Turing Test as the interrogator would merely ask if the respondent was a machine, and it would reply that it was either a computer or person. Unfortunately, intelligence has nothing to do with honesty, which a point Turing actually addressed in his 1950 paper. The two requirements that an artificial intelligence would need would be to both be able to converse - to respond in English, and to construct a reality within which the AI inhabits. This ontology need not, however, "reflect the reality that we experience".
Jim Russell - Artificial Intelligence
Congratulations on three successive programmes that totally seized my full attention and made me grateful that I can download an .mp3 file for later and more detailed study, even if they can take a day or two to appear. On hearing Mervyn's introduction to 'Artificial Intelligence' my mind immediately turned to Professor Joad of the 'Brains Trust' of many years ago and his apparently almost automatic reaction to any question, "It all depends on what you mean by, (in this instance) 'think'". It seems to me that in the context of intelligence, 'think' must include the ability to initiate thought, which probably requires self awareness or consciousness, at which point the waters of philosophy close over my head as I find myself trapped in a swirl of circular and contradictory concepts. Perhaps another correspondent can help
Claire Tancell - Artificial Intelligence
It would seem that the brain is an interpretive and reactive interface between physiology and environment. Its processes are chemical and thought is therefore a purely physiological phenomenon, requiring cell growth and chemical stimuli. It would be possible to create a machine (a computer) that would superficially mimic some of the responses of a brain, but this would not truly be thought, just as one could create a machine that carried out photosynthesis, but was not a plant. I think that computers could be regarded as thinking space, rather than tools for specific jobs (ie a whole typwriter, or a whole brain). The corollary of this is that maybe one day, it will be possible to fuse living brains with computers to literally expand the brains capacity for thought (memory storage, information retrieval etc) but not perform the whole thought process.
Ed Iglehart - Thinking machines, AI and mind
I was struck by the motivation of those engaged in AI. It seems to me that to try and create and control intelligence or mind from outside (or top down) as it were, is perverse. In Nature, these things arise from within, through self-assembly, and are 'emergent', and not predictable from the 'sum of the parts'. To try and assemble from outside also implies a desire to control not only the process of creation but also the operation of the created, although the expressed intent is to 'understand'...Is this not a bit hubristic?
Robert - Artificial Intelligence
Facinating program this week, glad I managed to catch it. Although it was a shame I was listening online, as the whole thing went dead after about 26 minutes - I hope the mp3 version will be complete so that I can catch the bits I missed. Particularly interesting this week was the mention of emergent behaviour - surely this must be critical to intelligence itself, as within a single brain you have billions of totally unintelligent cells, each doing a very simple job of passing on messages, and only the overall structure and combined effect of masses of these results in something that can be considered intelligent.
|
|
|
|
|
