“I trust this computer.” Yeah right.
Can we trust computers, or only people?
In his article All Transactions are based on Trust, Reinier Zwitserloot states that it’s impossible to trust a computer. Douglas Hofstadter, on the other hand, argues in I Am a Strange Loop that people are complex computers with intricate feedback loops — so if we can trust people, we should be able to trust computers. (For the record, I disagree with Hofstadter on this one. I’m not a materialist.)
We thought we’d have a bit of fun playing around with these concepts on the microPledge sign-in page, so we put:
“I trust this computer.” Yeah right. More like “I trust the people who use this computer — this web browser, to be precise — and I also trust the people who run this website.”
Speaking of trusting people who run a website … while I liked much of Reinier’s article, I don’t agree with him that you only trust your friends and your friends’ friends. Maybe it’s just me, but I use Wikipedia a fair bit, and generally trust it, regardless of whether or not my friends have edited the article I’m reading. And I’ve downloaded countless programs from SourceForge.net written by people I’ve never heard of. Sometimes my friends have recommended these programs, sometimes not. Also, despite what he says about reddit votes being worthless these days, I still find some of the top articles on programming.reddit.com pretty interesting. Even helpful.
Trusting people sometimes means networks of friends, but it may simply mean a track record and accountability. One of the ways we hope to do this with microPledge is give each developer a credit rating based on whether people are happy with their previous development. We’ll also post links to their previous development on their profile page, so users can see if they’re trustworthy or not (just like you can look at someone’s reddit user page for lots of downvoted comments).
In short, I believe it’s not just websites with “trusted friends networks” that have a future.
10 July 2007 by Ben 3 comments
3 comments (oldest first)
Thanks, Adam. Yeah, understood like that the argument would be reasoning backwards. However, I more meant to say, “people are complex computers with intricate feedback loops — so if we can trust people, it should be possible for us to trust certain computers.” In other words, I didn’t mean that we should then be able to trust all computers.
You’re right: in actual fact I Am A Strange Loop doesn’t talk much about computers. But I think the basic thesis of the book — that a sufficiently complex feedback loop is what makes us “soul-ish” — implies that computers can become intelligent and self-aware, and that we’d have to trust such beings just like we trust “real” people.
That is the case I was thinking of: if a computer ever did become self-aware, could we trust it? And what would it mean to trust it? Is Hofstadter right in thinking that the being would then be equivalent to a person?
I’m a theist, and an old-fashioned one at that: I believe our souls are God-given. And also that human souls are fundamentally different to either animal or computer “souls”. I’m not totally convinced that computers will never become self-aware, but I strongly suspect they won’t.
As for Strange Loop, I’m still not finished it. :-) I do like Hofstadter, but I’m actually finding the book a bit of a slog — to me it hasn’t been nearly as exciting and interesting as GEB or his essays in Metamagical Themas. If nothing else, it seems it’d be more compelling if it were half the length. At some stage I was thinking of blogging about my (relative) disappointment with this latest book. What do you think of it, as compared with GEB?
There’s not a big difference between computers and people.
“…people are complex computers with intricate feedback loops — so if we can trust people, we should be able to trust computers.”
I’m reading “I Am A Strange Loop”, and I didn’t get the impression that he was arguing that. (Besides, that argument is basically “a person is a type of computer, people are trustworthy, therefore computers are trustworthy”, and I don’t think it’s like Hofstadter to make such a trivial affirming-the-consequent fallacy.)
I don’t think he’d disagree with “I trust this computer” being contingent upon on “I trust the people who use this web browser and the people who run this website”. Unless the computer has a sufficiently intricate feedback loop that it gains a soul and becomes self-aware. :)