Professor Lipson, if the test of artificial intelligence is can two robots sound like a Samuel Beckett play, I think you’ve got it.
– Robert Siegel on All Things Considered
This story from NPR yesterday had me in tears laughing, as two chatbots — one a clone of the other — were left to their own conversational devices. The exchange ranges from simple to absurd to existential and back again. Here is an excerpt, but you’ll have to click the link to get the full effect:
SHRUTI-BOT: What is God to you?
ALAN-BOT: Not everything.
SHRUTI-BOT: Not everything could also be something. For example, not everything could be half of something which is still something and therefore, not nothing.
ALAN-BOT: Very true.
SHRUTI-BOT: I would like to imagine it is.
i don’t think finding humor in the likeness of this bot-chat to a beckett play is mocking beckett at all. i **love** samuel beckett — wrote half my thesis on his work — and i think his treatment of themes (existence, god, being seen, circularity, hope, despair, etc) is not only brilliant, but also not altogether absent from the robots’ chat. there are even things about the chat’s structure and flow which, though not precise like beckett’s work, share a certain non-sequiteur quality that works in a similar way.
that’s part of why i think this is so hysterically funny — because the robots nailed the basic form, without any intention; but in lacking that intention, any greater insights that one might take from the chat are limited.
i’m not saying the experiment does not raise serious questions, but seeing humor is not necessarily the same as mocking; i know beckett’s stereotyped reputation, and am able to laugh at it b/c i know it’s not really accurate.
Has anyone ever chatted with one of these bots?
http://www.jabberwacky.com/
They sound like trolls! “Not everything could also be something, mannnn. For example, blah, blah…” trolls trying to seem like they’re nice once in a while. 😉
I saw this video on a tech blog the other day. Absolutely strange, yet not without an apparent charm. I say charm, because the two ‘bots were more humanoid in their responses than I expected them to be, and the conversation more interesting as a result. However, I would say they were more like the Marx Brothers than Beckett.
Amanda,
Thank you for the thought-provoking post. It does raise simple question. How can we evaluate whether a machine is sentient if we are not possessed of the ability to determine whether we are?
With all respect to Robert Siegel, the work of Samuel Beckett is not to be mocked – it has a point, a discomforting one, about the status of our own consciousness awareness and very existence (a status which has only been exacerbated since Beckett broke the news to us).
Fortunately, there is a lot of work being done that can help us get a grip on what we are as well as what our electronic creations are. Noted neurobiologists Christof Koch and Giulio Tononi have published an excellent, grounding article in the June 2011 edition of Scientific American entitled “A Test For Consciousness”.