Nancy Fulda, herself a lettered student of artificial intelligence, joins us to talk about writing artificial intelligence believably. We fire questions at her so that you don’t have to!
We talk about what’s current, what’s coming, and what it is that we’re all expecting. We also cover some of the things that writers get wrong (at least insofar as they knock the cognoscenti out of the story.)
Liner Notes: Here’s the article Howard mentioned, “Evolving a Conscious Machine,” from the June 1998 Discover. He got the details almost 100% wrong, but the gist of it was still there.
Podcast: Play in new window | Download (Duration: 19:41 — 36.0MB)
Subscribe: RSS
Go to the Internet and look up Bayesian learning, neural networks, and genetic algorithms. Yes, it’s more of a reading prompt.
Rainbows End, by Vernor Vinge, narrated by Eric Conger
(note: Howard got this wrong — no apostrophe at all! And yes, a lantern got hung upon that particular missing bit of punctuation.)
Man that reading prompt, you guys are MEAN. Stuff can pop a person’s head if they go in cold :P
“We anthropomorphize our toasters!” is a great line in context.
Nancy made a fascinating observation: Programs like ELIZA are like stage magic; the more you know about how it works, the less magical it becomes. E.g. if you know how a Chess supercomputer works, you may still admire the engineering behind it, but you understand the actual process as rote data processing.
In my experience, this does not happen with the human mind: After reading about new research that – often fundamentally – enhances our understanding of the human brain, mind, decision making, moral judgement, perception, or behaviour, I always come out with an even greater sense of wonder. You might say there is even more magic in it.
Timothy’s Test: An AI is a “true” AI, if itbecomes more magicalthe more you know about how it works .
This is not only very anthropocentric, but also quite subjective. But any test for AI “trueness” will be, because our concept of intelligence is.
More Nancy please! :)
That writing prompt… writing excuses does not mess around. Looking forward to listening to this one tomorrow. My own AI story went nowhere for various reasons.
Did you intentionally switch from 96kbps to 256kbps? It made the mp3 file almost three times the size of your previous podcasts.
@Doug: I’ll check with Jordo. It’s possible this was an export error (for “likely” values of “possible.”)
@Nizlemia: We loved recording with Nancy. There are more Nancy episodes in the mixing queue, and I’m sure we’ll have her back in the future.
Ooh, I have to go listen to see if the audio quality got better. My wife and I often listen to this together in the car, and while we love the ‘cast, we often have a lot of trouble with the audio. Generally the problem is that Brandon (my favorite) is too loud if we turn up the audio enough to hear Mary (my wife’s favorite).
Thanks for the podcast. AI’s always throw me off because I think of something cool, then next thing I know a friend or someone comes to me and goes ‘Look at this cool new tech I have!’ and suddenly my creations don’t sound so creative anymore.
I’d love for this topic to be revisited the next time Nancy is on the podcast. That was fascinating.
ok, I’m not even into the episode and I have to drop you guys a line.
the Uplifted Kitten was classic… I had to rewind and listen again.
love the show as always
And for the textually oriented sapient beings out there…
A Transcript!
http://wetranscripts.livejournal.com/84112.html
Great episode! Would anyone have a citation, or avenue to search for the story Howard told of the computers programming TI 555 timers? Would love to read more about that.
I can’t believe neither the movie Her, nor IBM’s Watson came up!
Nancy commented that she can’t figure out a way to make electronic work without a clock. Biological life appears to be that way, too.
The CLOCK (Circadian Locomotor Output Cycles Kaput) gene controls the persistence and period of our circadian rhythm, which in turn influences reproduction, mood, and sleep. We don’t have just one circadian clock, but rather the one in our brain helps synchronize the “63 different systems” (to borrow a Howardism) in our body that have their own clocks (and yes, it is possible, if something goes wrong, you have your liver on London time when the rest of you in on New York time). And the Circadian rhythm is just one of several clock systems in our bodies.
Generally speaking, the more I learn about biology and electronics, the more they appear to be fundamentally the same.
Thanks for this great episode! I have been struggling with a story featuring an AI protagonist, and this has given me a couple of ideas and avenues to go down. Thanks for all your hard work.
Oh thank goodness Terrell brings up an issue I have with Writing Excuses. In fact, I emailed the show a few months back but never heard a thing. Mary’s so gosh darned quiet, but usually has the best things to contribute, making it frustrating to appreciate the whole program. Can you turn up the gain on Mary’s microphone?
http://mag.newsweek.com/2014/02/21/military-robots-increase-complexity-relationship-soldiers.html
Support for Mary’s point about inevitable anthropomorphisation…
This episode prompted me to not only write but comment here for the first time. First, Nancy Fulda is an impressive writer and intellect. I hope to hear her voice on this ‘cast again. (and again)
The last five minutes of the show directly benefited me by showing me another critical aspect of thinking about the AIs in my own novel-in-progress. Thank you!
Great show.
Rock on.