Artificial Intelligence

We were chatting. I was grasping for a great, recent quote that congealed my thoughts well but I couldn’t find it in my quote collection nor recall anything but the vaguest of phrasings.

What I remembered was that it was written by that famous author who committed suicide, I googled that but that’s sadly too broad a description. So I kept thinking and I also remembered that he was famously very much a fan of that famous swiss tennis player, whose name of course also evaded me. But googling was successful this time, retrieving not Martina Higgins, but ah, yes, Roger Federer. So now I google “federer author suicide” and that finally got me David Foster Wallace. With the name it was a snap to find the quote in my collection, and all of it happened real-timely enough to keep the flow of the IM conversation.

This sort of thing has happened often to me and I’m sure it has to you: googling for vague recall, for completing your thoughts. Instead of closing your eyes and willing an unconscious mind racking you outsorce to Google the unconvering of the tip of your tongue. What stroke me this time was the chaining and the speed (just-in-time-thinking). What got me to write this down was that in a few years such a thing will be so unremarkable I’m sure we’ll wonder how it felt before, if those in transition ever noticed how their mind was being steadily extruded.

The quote?

Believe the hype. Please take a while and go play with it! Its help, as is Wolfram’s tradition, is excellent, the best introduction.

How to describe it? It’s for data what Google was to text, what Wikipedia was to knowledge. It’s to the calculator what Wikipedia was to the encyclopedia, what Google was to the library catalog. It’s the most exciting, hopeful thing to happen to the web, to the world, since both Google and Wikipedia.

And with a mission “to make all systematic knowledge immediately computable by anyone”, it opens up as big and inspiring a project for this generation.

I believe it’s a historic moment and could not let it pass unmarked.

This John Reisig find is fantastic! One more milestone of how “dumb” browsers are steadily becoming more and more powerful…

Ah, long time since I’d been so delighted by so wonderful a metaphor!

Edge’s 2009 Question is out!: What would change everything?. The list of answers by some of the most interesting individuals in the third culture individuals out there is as inspiring and thought provoking (and atrociously designed, interface-wise) as ever. Kevin Kelly’s answer my favorite so far:

Great stuff—it’s people like Kelly that make me miss California ;)

Jeff Bezos had remarkably similar, equally inspiring ideas at a recent TED talk, comparing the web to electricity but Kelly pushes it further, to ”intelligence as electricity”

Hans Moravec’s Mind Children is dated at parts but it is as lyric, visionary, and deep as anything I’ve read. Here 2 excerpts I particularly enjoyed: one on robot bushes (forget many-armed Vishnu, this will be the avatar of the gods), the other on artificial life, spontaneous and deliberate (a great story of a virus put deep inside Unix by Ken Thompson himself and a subtle but intriguing example of spontaneous alife).

I’m experimenting with taking quick page snaps and bundling them into PDFs, so these are rather crude. No matter, it’s early days and I’ll get better at it. PDF/CBR is the new MP3!

As David, to whom I own the Big Dog acquaintance, said: its movements are so fluid, so eerily natural, biological, one just knows the days of the flesh are counted. “It was nice being human.”

This was a couple of weeks ago but I had to write about it because I was so happy through it: Steve Omohundro’ s wonderful talk, AI and Transhuman Morality, organized by the Sillicon Valley transhumanist meetup. I brought Mauro with me and I was very nervous because I didn’t know what to expect. A couple of days ago I had gone to an AI meetup in the same room (in the wonderful TechShop) and it had been confusing and somewhat disappointing: we watched an overly long video, had some haphazard if interesting discussion, and it all ended up abruptly without me being able to make up my mind of the strange event (where these people quacks? mad geniuses? autists? were all meetings this awkward?).

Anyway, we went and I’m happy we did because I enjoyed Steve’s wonderful two-hour presentation so much I was smiling like an idiot the whole time (at one point, I even clutched Mauro to tell him simply, “I am happy”—and it was true). As I said, it was more than two hours long but I honestly didn’t want the presentation to end, particularly when so many of the interventions where, wonder of wonders, relevant and interesting of themselves.

The presentation was divided in 2 halves. The 1st for reviewing what we know of human morality, the 2nd for contemplating what AI morality will be like. Both were fascinating and chock full of surprising, cutting-edge ideas (and book recommendations!), but it was the 2nd where I was truly overjoyed, for, you see, it was when Steve plunged into how an AI’s morality might be structured.

I was struck by how the utility function ethics he considered for AIs were exactly the kind of ethics I had chanced on one day, not long ago, when in my desire to clarify how and for what I wanted to live, I thought, wrote, and rewrote about ethics with the most honesty and rigor I could muster. Heck, we even used the same examples! You have no idea how good it felt to finally find a fellow freak who  not only understood and care about my conclusions but who had arrived to them through entirely different paths (conclusions like how ethics hinge entirely on purposes or goals and how we’re in for an ethical ride when these become much more varied and malleable than they’ve ever been before). Back in Guadalajara I talked about this all the time but no one ever really got it (or much cared).

Ah, this kind of stuff was why I came to the bay area! (Mauro liked it a lot too, saying afterwards he had felt as one should feel after going to mass—full of awe and excitement.)

From Nick Bostrom’s Golden—a fictional interview of Albert, an uploaded dog. His cheeriness and good disposition are attributed to his being a golden retriever. His wisdom I attribute to Bostrom, who’s one fascinating philosopher (don’t miss the fable of the dragon tyrant!).

For those armchair observers of the breathtaking world of quants and structured finance, as myself, Technology Review’s current issue carries a wonderfully didactic and gripping introduction, The Blow-Up: (pesky but FREE registration required).

Focused only in securitization, When it goes wrong, from The Economist (YubNub’s “eco“), is also a good overview and glimpse:

Answers.com (YubNub’s “a“), btw, is invaluable in navigating jargony fields like finance.