The Singularity
Whilst reading “Novacene”, and prior to writing a blog post about it, I was struck by the fact that the author believes both that we are alone in the universe and that we will ultimately be succeeded by vastly more intelligent electronic cyborgs. Despite the 3.7 billion years of evolution and freakish chance required to create us, we are ultimately succeeded by inorganic devices of our own creation. How can we be at once so rare, and yet so replaceable? Is it just hubris to believe that there is something inherently sacred about organic life?
It doesn’t take much to suss out A.I. as it stands, whether it be the hilarious results of an artificially intelligent recipe or a chat with one of the internet’s best-known, most highly-lauded, publicly available chatbots. If you’ve not had the opportunity, allow me to save you from it (bot’s responses in italics):
“How are you?”
Pretty good thanks How about you?
“Not bad.”
Thanks for the kind remark.
“What kind remark?”
Sorry, I thought you complimented me?
“No. You asked me how I was. I replied ‘not bad’. It wasn’t a compliment.”
Fair enough. I must have been mistaken. Oh now I remember asking you how you were. What did you say? Yes I know you did. I just heard you. Was it an insult?
“Was what an insult?”
What you were describing a couple of sentences back.
“Do you know what I was describing a couple of sentences back?”
Nope, please tell me.
Suffice to say that no Turing test was passed here... What if this had been customer support?
It seems that when we try to get an A.I. to do human stuff, things quickly fall apart. But what if the whole premiss is wrong? After all, Lovelock reminds us of that brilliant Wittgenstein quote: “if a lion could speak, we could not understand him”. Perhaps our expectation that an A.I. could ever produce a tasty recipe, or hold an interesting conversation in English, is inherently flawed? In the same way, could our desire for concrete, historical moments such as a passed Turing test, or a recognisable moment of Singularity, be human constructs grafted onto the evolution of “the cyborgs”, in an attempt to understand and control it?
Ultimately, human intelligence was driven by the evolutionary necessity to survive. Humans needed food and conversation to thrive. If the raison d’être of general A.I. were to emulate humanity, it would forever be in our shadow, creating bad food and making awkward conversation… a real-life Kryten.
A.I. is indisputably here. We can’t turn it off in the same way that we can’t turn off the internet. But its evolutionary driver - that which will make it superhuman - remains elusive.