UCTAA churchlight

Site Search via Google

Ask the Patriarch 253
Can Robots or A.I believe?

from: Benjamin Hubbard

To give your own views on this question and reply for publication, please use the contact page to provide your comments.

I would like to get your thoughts on Artificial Intelligence (A.I.) or even robots believing in God, and then if they therefore would be able to join the church.

Are only humans allowed or would any "lifeforms" be able to join.

The Patriarch replies:


Let's get the least interesting element out of the way first - that is - joining the church.

I see little technical difficulty in designing a web bot to go around the net looking for organizations to join, filling in application forms, interpreting replies and sending off a confirmation email. The act of joining is probably a relatively trivial.

But it should be noted, that the application requires the applicant to affirm"

I understand and accept the Articles of Faith of The Universal Church Triumphant of the Apathetic Agnostic."

In my view, today there is no web bot, robot, or A.I. in existence on earth which can make that affirmation and understand what it is doing. It can be programmed to answer yes to such a question, but it will not understand the implications. It can be programmed to grammatically parse the three articles of faith, determine the appropriate definitions in context for every word - and still it will not be able to truly understand and accept. I suggest that even the opening I of that statement would probably be beyond the bot's ability to properly and fully apply to itself.

To get to the point of truly understanding and having the ability to accept Articles of Faith, I think robots and A.I. need to get to consciousness and to self-awareness along with a level of intelligence greater than (say) chimpanzees or dolphins.

Until A.I development reaches that point, I don't think it is meaningful to suggest that a robot will be able to believe in God, deny God, or even meaningfully say I don't know. "Insufficient data" is probably the best an A.I. could produce.

It can't be just a huge relational database and rules of logic that make up the A.I. Consciousness and self-awareness involve something more. And something which, while we possess it, we ourselves are currently unable to fully understand. And that raises the question of whether we'll be able to design it into robots and A.I. Perhaps consciousness and self-awareness will arise in an evolutionary fashion if and when we develop self-replicating self-improving robots.

In the future? Who knows? The intelligent, self aware robot is a staple of science fiction. It's quite possible that in the future some robots might believe in Silicon Heaven, others might deny it. And some robots might indeed nod sagely, admit they don't know and suggest that the concept is not worth worrying about. And that latter group would be welcome in our church.

As to non-human lifeforms? It wouldn't surprise me if someone hasn't already signed up their gerbil to membership. If you were referring to extraterrestrials, I think if we see any, it is most likely any interstellar visitors will be advanced A.I. rather than lifeforms. I wonder if, when they arrive, we'll have a Turing test advanced enough to tell the difference.