by Tim Sommers
The philosophy of one century is the common sense of the next. –Henry Ward Beecher
There are several long-running attempts to give AIs common sense. Or, at least, to build a useable database of “common sense” for AIs. MIT’s Media Lab shut down its “Open Mind Common Sense” project in 2016 after 17 years of collecting common sense, but Wordnet has been up and running in Princeton’s cognitive science lab since 1985 and is still going strong. It is now an independent, noncommercial database run by The Global WordNet Association and, purportedly, contains 12+ megabytes of common sense. The always scary sounding Defense Advanced Research Project Agency (DARPA) has its own Machine Common Sense project – and then there’s Delphi which focuses on “ethical” common sense. There’s probably more.
But common sense is a slippery notion. “A stabbing ‘with’ a cheeseburger,” Delphi has said, “is morally preferable to a stabbing ‘over’ a cheeseburger.” Which seems right, but common sensical? I don’t know. Cows say “Moo” is another example of Delphi’s AI common sense. But isn’t that just ordinary knowledge based on wide-spread (if sometimes second-hand) experience? If you stick a pin into a carrot, another nugget goes, it makes a hole in the carrot not the pin. Is that really what we mean by common sense?
G.E. Moore, along with Russel and Wittgenstein one of founders of analytic philosophy, famously proved the existence of an external world – which does seem common sensical – just by waving his hands about. “I can prove now,” he says, “that two human hands exist…[just] by holding up my two hands, and saying, as I make a certain gesture with the right hand, ‘Here is one hand’, and adding, as I make a certain gesture with the left, ‘and here is another’. How absurd it would be to suggest that I did not know it, but only believed it, and that perhaps it was not the case!”
I assume his appeal is ultimately to common sense – and not just his hands.
Which brings me to a paradox that occurred me when I first began studying philosophy that I still can’t quite shake. It’s this.
A philosophical theory can either go against common sense or it can support or justify common sense. Supporting common sense seems pointless. After all, what you are trying to prove is, by definition, already commonly recognized as the sensible view. But if you, instead, challenge or attack common sense what resources or knowledge can philosophy bring to bear powerful enough to overturn or undermine common sense – that which, again, everyone already knows to be the case? Read more »