John Markoff, writing for The New York Times, reports on the discovery of a long lost paper by M.I.T mathematician Norbert Wiener. Written more than sixty years ago, the final paragraphs of the paper resonate loud and clear as we now flirt with the realities of artificial intelligence and, one hopes, consider carefully what it is we wish for from our machines.
“Moreover, if we move in the direction of making machines which learn and whose behavior is modified by experience, we must face the fact that every degree of independence we give the machine is a degree of possible defiance of our wishes. The genie in the bottle will not willingly go back in the bottle, nor have we any reason to expect them to be well disposed to us.”
See full article here.
‘s funny you should bring this up, David. It so happens I was reading a short story (or rather – the foreword to a non-existent book) from the late, great Stanisław Lem the other day, that treats of just such a problem (namely: the fiasco of the US’s attempt to produce an ultimate strategy computer, due to the fact that once such computers exceeded human intelligence levels they decided that philosophy is much more interesting than politics). There’s a fragment in there that sums it up pretty nicely, that I’ll take the liberty of translating:
“The loftiest intellect cannot be the lowliest slave.”
It’s easy to come up with BS slogans like that. But you’d have to define intellect first, because computers as of today are much much better at many things that were previously considered highly intellectual work (complex symbolic math, bookkeeping, etc.).
In fact, it’s the least intellectual work (movement, object identification, obstacle avoidance), stuff that even fruit flies are able to do effortlessly, that is what is difficult to do in a computer and requires the highest levels of both in research efforts and computational complexity.. Really not surprising that stuff that evolved over billions of years is harder to reproduce then something that has evolved over thousands of years (human analytical capability).