Abstraction

As far as I know, the Turning test is inherently anthropomorphic — a human judge engages in a natural language conversation with one human and one machine, each of which try to appear human; if the human can’t tell which is which, you’ve achieved true artificial intelligence. Or some such.  There’s no Turning test for machines mimicking non-humans (such as baby dinosaurs).

Since I’m lately engaged in working with robotics designed to engaged emotionally with people, I’ve been thinking about the level of (artificial) intelligence required to achieve a strong emotional bond between the machine and a person, even if the machine falls short of passing a Turing test.  I suppose the true answer to this question is “none;” for example, my son has developed a reasonably strong attachment to an entire menagerie of stuffed animals (”woofy,” a small dog, is the present favorite).

With stuffed animals in mind, interactions between people and inanimate objects might initially seem to occupy a continuum describing the ratio of human input vs. inanimate input. From this perspective, the case of my son and Woofy occupy an extreme end of the continuum in which my son provides 100% of the intellectual and emotion input, while Woofy provides nothing at all. I think this continuum doesn’t really exist, though — in reality, the human being shoulders the entire intellectual and emotion burden of interaction, regardless of the sophistication of the inanimate object which which he is interacting.

A more accurate continuum might be how easy it is to engage intellectually/emotionally with an inanimate object. One extreme end of the continuum might represent objects with which it’s impossible to get attached. It’s not difficult to thing of objects that fit well on this end — a rotten fish, say. (In fact, I think most objects fit at this end).

As we move away from the extreme end of the continuum, we find objects with which more people can engage emotionally, and at greater intensity. At the opposite extreme might be an artificial human that effortlessly passes the Turning test. Perhaps Woofy lies somewhere in the middle.

Important for entities like the Pleo is the characteristics that effectively move the needle toward the “easy to engage” end of the continuum. Characteristics that improve Pleo’s “endearment quotient,” so to speak. Are some of these characteristics universal, while others are culturally bound, or specific to an age group, or a gender, or living circumstance (such as being bed-ridden) or a combination of these? I don’t think it’s possible to answer these questions in the abstract; rather, Pleo has to submit to a Darwinian experiment of sorts — a submission of personalities and characteristics build by people from around the world. Some of these will fail to improve the “endearment quotient” of Pleo, some will dramatically enhance it.

We’re so close to launching the developer program for Pleo. I’m eager to see how it evolves.

David Sedaris is hilarious, and he had a fantastic essay on air travel in the New Yorker recently. I particularly liked this passage, about traveling in business class. It so perfectly parallels my experience:

“May I bring you a drink to go with those warm nuts, Mr. Sedaris?” the woman looking after me asked—this as the people in coach were still boarding. The looks they gave me as they passed were the looks I give when the door of a limousine opens. You always expect to see a movie star, or, at the very least, someone better dressed than you, but time and time again it’s just a sloppy nobody. Thus the look, which translates to “Fuck you, Sloppy Nobody, for making me turn my head.”

I feel exactly that way whether I’m in coach (most of the time) or business (the occasional upgrade). Walking to my coach seat I’m always thinking about how much more comfortable these slobs in business are going to be for the next 12 hours or so, damn them. Likewise, I feel a vague sense of guilt if I’m sitting in my business class seat. I’m thinking “this trip is going to suck *so much more* for all of you.” Certainly it doesn’t help that I epitomize the sloppy nobody as I sit there in my fleece listening to an iPod while other business travelers whisper importantly into their Blackberrys.

One of the most amazing things I’ve ever seen is a demo of Photosynth done by Blaise Aguera y Arcas at TED 2007. Photosynth is a kind of sophisticated web robot that can find images (e.g. by scouring sites like Flickr), automatically determine which images are of the same object, and knit those related images together into an amazing mosaic through which users can navigate.

It’s impossible to really understand how incredible this technology is by reading about it. Take a look at the demo for yourself, and visit the Photosynth site.

What really caught my attention, beyond the technology itself, is the social networking implications for Photosynth. It’s enables one to look at an object (say, Notre Dame, as in the TED demo) through the eyes of thousands of people simultaneously; it potentially represents our collective view of the world. Amazing.

I work for a company based in Ireland, and I travel to Dublin a lot. It’s a global lifestyle filled with parties and laughter, a jet-setting wonderland that only a very few get to enjoy. I know you’re jealous! And well you should be.

Take comfort. There’s always pictures through which you can live vicariously. Like this one, taken from my hotel window this morning. Ah, Ireland in January!

This quarter’s topic at the Silicon Valley Junto is “This I believe,” a theme borrowed from the recently-reinstated NPR program. “This I Believe” was originally produced in the 50s by Edward R. Murrow, and was wildly popular in its day. A book of transcripts of essays from the program was a huge best seller. Because Murrow refused to support the show with commercials, he was able to include many scandalous essays on the program that would otherwise never air because frightened sponsors would kill it. One of these was an Eleanor Roosevelt essay in which she admitted to doubts regarding the existence of God (if I remember correctly, her gist was that it’s more appropriate to focus on solving worldly problems rather than perservating on post-death access to a gated community with lots of gilding and de rigeur facial hair). But I digress.

Anyway, I’m out of town and won’t be able to participate in the Junto, but it made me think. What would I write? What, exactly, do I “believe?” It’s a deceptively difficult question, at least for me. Certainly there are things that I think are true, but I find it difficult to boil these ideas down into one that captures me. Something that I can point to and say, “this is what I’m about.”

The trouble is that the things I believe are either trite (I believe my employer should provide me with all the free Diet Coke I can drink) or are pretentious (who the hell cares what I believe politically or religiously or whatever). Then there’s the temptation to write about things I “believe” that are really about trying to establish myself as a good person.  “I believe in treating people equally and with kindness” is abstractly true, but the true-truth is I’m often a grumpy asshole for no good reason. And don’t get judgmental about that, mother fucker, because you can be an asshole sometimes too.

I imagine this all has to do with the depth of thought one puts into things. That is, the depth of one’s thinking about an issue is proportional to the importance one places on the issue. It’s disconcerting to be faced with one’s shallowness.

My sporadic blogging oflate is due to a major change — I’m transitioning jobs, having finished my last day with my previous employer Friday only to get on a plane Saturday, fly to Ireland, and begin my first day with my new employer (based in Dublin) on Monday.

It’s been an amazing experience, not least because the trip has occasioned my first experience driving on the “wrong” side of the road. I’ve been nervous about it for some time; I imagined that I’d arrive in Ireland and my jet-lagged mind would turn the wrong way at the first intersection I encountered.

Well, so far so good. I rented my car, got lots of insurance, pushed and shoved my (modest) suitcase into the smallest car I’ve ever seen, much less driven, and started driving around on the left side of the road.

It’s an odd experience for two reasons. First, I’m not used to *thinking* about driving. It’s something for which I usually rely on muscle memory, saving all other cognitive processes for grumpily judging other drivers.

Second, the muscle memory is still very much there. I’m experiencing a kind of dual-mind in which I’m intellectually telling myself “stay left,” but every unconscious process I have is simultaneously pulling me the other direction. This duality really freaked me out when I was driving down a road and someone pulled onto the road from an intersection ahead of me. The other car turned to travel toward me, and my intellectual mind said “they are turning into the right lane,” but my body erupted in alarms “YOU’RE IN THE WRONG LANE! GET OVER!” My little car began swerving slightly to and fro as I fought with myself, but I made it through without driving off the road into the trees.

Only four more days to go, then I have to try and switch back to driving the “right” way. Ugh.

Anyone representing a product weaves the conversation along a continuum that includes absolute accuracy, accuracy in terms of planned developed, accuracy in terms of potential development, and just plain inaccuracy.

Software is abstract. From a product perspective it doesn’t exist in the sense that, say, an automobile exists, or a laundry detergent.  Software is ephemeral—so we can’t touch it, we can’t directly measure it, and we can’t see it.  This makes marketing and managing software products different from managing “traditional” products; it’s much more challenging, especially if the product is new or if it performs behind the scenes (such as a middleware product). The challenge means that product management or marketing is a rewarding career. It also means you’re probably going to screw up a lot.

This blog is about the things I’ve learned over the course of more than a decade of screwing things up.