The Pros & Cons of Artificial Intelligence vs Human Intelligence.
The real issue is not whether the answer to a question is provided by artificial or human intelligence, but whether the answer is true, relevant and complete or merely plausible and convincing. Or to put that more bluntly – is the answer true, relevant and complete or just Bullshit?
A digression to apologise if you think that Bullshit is a rather “naughty expletive”. It’s actually the title of a 2005 best-selling book by Harry Frankfurt, a distinguished former Professor Emeritus of Philosophy at Princeton University NY. Frankfurt explains:
“…. the bullshitter …. is neither on the side of the true nor on the side of the false. His eye is not on the facts at all, as the eyes of the honest man and of the liar are, except insofar as they may be pertinent to his interest in getting away with what he says. He does not care whether the things he says describe reality correctly. He just picks them out, or makes them up, to suit his purpose.”
Bullshit is clearly NOT a new phenomenon, but the rapid growth in the development of – and reliance on – computers and social media for generating and distributing information has created an absolute explosion in the popularity of Bullshit. And particularly amongst the gullible looking for quick answers which suit them, and those looking to exploit the gullible for their own benefit – from social media “influencers” to fraudsters and politicians !
A well-designed sharp surgical knife is a wonderful tool in the hands of an experienced surgeon restoring a patient to good health. But in the hands of a malicious individual it is a weapon of horrible death. Unfortunately, alongside much evidence of the potential for good of Artificial Intelligence, we are already seeing plenty of evidence of how Artificial “Bullshit” is being used to the detriment of those who are taken in by it.
The essence of the issue is:
“how can people, who are asking questions because they lack the knowledge or expertise to answer a troubling question or concern for themselves, be confident that the answer (whether from human or artificial intelligence) is true, relevant and complete and not just “Bullshit” (human or artificial) ?
My professional background was in academic/research science – experimenting, collecting and analysing information to try to find credible explanations for some of “science’s little conundrums”. So I’m used to searching the scientific literature and the internet to see if any other researchers have also been searching for answers to the same or similar questions. If they have – and have all come up with the same or similar answers – it is probable that the answer is “true”. But if some of the answers are widely different, the probability of at least some of them being just Bullshit is quite, high and further research and re-examination of the data/evidence is necessary to determine which (if any) of the answers is the right one, and not just Bullshit.
Much the same thing applies to “Artificial Intelligence”. Always be sceptical about AI answers to question. They can be very helpful – but the fact that they appear to be plausible and convincing doesn’t prove that they are right. So don’t be afraid to challenge them and compare the answers for consistency. Try asking the same question in a different way, eg: using different words (synonyms) in different orders to see if you still get the same answer; or try asking the same question in different AI systems.
But there’s an “elephant in the room” problem with that ! As AI systems get larger – accessing unimaginably huge amounts of data using computers occupying vast buildings and consuming as much water and electricity as a small town – the ability of “ordinary” users to access several such computers to compare answers to the same complex question diminishes considerably. Which leave the questioner with little choice about the answer – believe it or leave it!
I’ve been experimenting with that recently and, I have to say, that so far I have been finding the consistency of the answers very unconvincing.
So while I don’t deny the potential for good of AI – both now and what it might develop to in the future – I remain very cautious about putting too much faith in the “truth” of its answers and the honesty and integrity of the people who exploit it. To my mind, relying on one Artificial Intelligence source as a true, relevant and complete answer to your question is potentially risky. And if you immediately believe that answer simply because it is plausible and convincing – and particularly because it was the answer you were hoping for – the risk that you join the community of the gullible to be led astray by a “Pied Pipper playing a Bullshit tune” increases significantly. The notion that computer software is always infallible is very dangerous – you only have to look at the recent Post Office debacle to see that.
(PS: “For the avoidance of doubt” - this reply WAS written by hand NOT by AI)