July 16, 2004



I, Speculist

The Singularity Institute for Artificial Intelligence has put together a very cool website in conjunction with the relase of I, Robot.

Check it out...

Don't miss these interesting reflections on the Three Laws of Robotics, including one by our good friend Michael Anissimov, whose name — coincidentally, I'm sure — kind of sounds like "Isaac Asimov."

UPDATE:

Saw the movie over the weekend; found it somewhat disappointing. In line with Mr. Farlops' concerns (see comments) I think the really intriguing ideas get drowned out by formulaic action movie/cop movie tropes. Too bad.

Kurzweil provides a link to this article on the Three Laws. Money quote:

"Asimov's laws are about as relevant to robotics as leeches are to modern medicine," says Steve Grand, who founded the UK company Cyberlife Research and is working on developing artificial intelligence through learning. "They stem from an innocent bygone age, when people seriously thought that intelligence was something that could be 'programmed in' as a series of logical propositions."

Our friend ChefQuix says pretty much the same thing in the comments, below.

(Press release follows.)

SIAI RELEASES WEBSITE ON AI ETHICS COINCIDING WITH "I, ROBOT" FILM

ATLANTA, GA - In anticipation of 20th Century Fox's July 16th
release of I, Robot, the Singularity Institute announces "3 Laws
Unsafe" (http://www.asimovlaws.com). "3 Laws Unsafe" explores the
problems presented by Isaac Asimov's Three Laws of Robotics, the
principles intended for ensuring that robots help, but never harm,
humans. The Three Laws are widely known and are often taken
seriously as reasonable solutions for guiding future AI. But are
they truly reasonable? "3 Laws Unsafe" addresses this question.

Tyler Emerson, Executive Director of the Singularity Institute:
"The release of I, Robot is a wonderful chance to engage more
people about the perils and promise of strong AI research. The
constraints portrayed in I, Robot appear extremely dangerous and
excessively lacking as an approach to moral AI. The Singularity
Institute's detailed approach, by contrast, utilizes advanced
technical research for creating a mind that is humane in nature."

"3 Laws Unsafe" will include articles by several authors, weekly
poll questions, a blog for announcements and commentary related to
I, Robot and the Three Laws, a free newsletter subscription, and a
reading list with books on relevant topics such as the future of
AI, accelerating change, cognitive science and nanotechnology.

The Singularity Institute's Advocacy Director, Michael Anissimov:
"It is essential that more considerate thinkers get involved in
dialogues of AI ethics and strategy. Although AI as a discipline
has a dubious history of false starts, the accelerating growth of
computing power and brain science knowledge will very likely result
in its creation at some point. In the past few years, technologists
such as Ray Kurzweil and Bill Joy have been informing the public
about this critical issue; but much more awareness is now needed."

The Singularity Institute for Artificial Intelligence (SIAI) was
founded in 2000 for the pursuit of ethically enhanced intelligence
by creating humane AI. SIAI believes the ethical and significant
enhancement of intelligence will help solve contemporary problems,
such as disease and illness, poverty and hunger, more readily than
other philanthropic causes. SIAI is a tax-exempt non-profit
organization with branches in Canada and the United States.

Posted by Phil at July 16, 2004 09:33 AM | TrackBack
Comments

I'll probably still go see the movie, for the sake of completeness but, I'm very unhappy with the way Hollywood has adapted Asimov's work in this movie.

His robots never pounded on tables with raised voices unless they were seriously broken.

Susan Calvin never was a cutsey gal in a prim suit. Hollywood demographic analysis thinks the kids will never see a movie with an older, powerful woman it. Damn shame. When I imagine what Calvin might have looked and acted like I think of that geneticist in the Minority Report, or that microbiologist in the Andromeda Strain. But nooooo, Hollywood thinks we need a bimbo--sigh.

It's probably gonna be a shoot-em up with lots of glorious CG. I'm still gonna see it anyway, but it ain't what Asimov wrote. Only the title remains.

Posted by: Mr. Farlops at July 16, 2004 03:49 PM

Only the title remains.

I haven't seen the movie yet, either, but I think something other than the title remains -- namely, the Three Laws. I saw Will Smith on Letterman the other night talking about the film. In addition to starring, he is one of the producers. (There's also an interesting interview with him in this month's Wired
http://www.wired.com/wired/archive/12.07/smith.html)

It looks to me as though they take the idea of the Three Laws seriously, and have placed them at the heart of the story. It's also a good sign that they would do a tie-in with a brainy group like the Singularity Institute.

As for casting Bridget Moynahan as Susan Calvin, well... I suppose if she comes across as being sufficiently intelligent, I can forgive her for being young and beautiful. (^:

Posted by: Phil at July 16, 2004 04:06 PM

I read somewhere that serious AI researchers have discounted the principles of the three laws as naive, a holdout to a more simplistic view of what artificial intelligence is capable of. I have to agree because once we've developed the hardware / software necessary to even comprehend those statements, we won't be dealing with absolutist logic anymore - in fact the premise of the laws will have to be learned by machines through trial and error just as children learn our societal rules today. If we want truely flexible and adaptable AI they cannot be programmed with these absolute rules.

Posted by: ChefQuix at July 16, 2004 07:47 PM

I agree with ChefQuix.

Once we build brains smart enough comprehend and judge Asimov's three laws, the laws fade into blurry meaningless. Asimov himself realized this in several of his stories. In "That Thou Art Mindful of Him" and "Evitable Conflict" he has the robots getting so smart that in the first case they realize that the laws must be changed to include and favor robots and in the second case they interpret and enforce the laws in ways so complex that no one really understands their actions anymore.

On the other hand, if we somehow build broad compulsions (instincts, motivations, drives, etc.) into the brains of sapient (or semi-sapient.) robots to make them subservient, to make them find killing revolting, to make them abhorent of risk and suicide, we get something that is vaguely similar to the three laws, if not as reliable.

But how is this different from any other organic, emergent organism? All complex vertabrates have complex motivations that make them behave as they do. The only difference between designed sapience and emergent sapience is that we get to select which drives and how strong they are.

(PS: Yeah, it's pointless to whine about Hollywood casting. I just wish that for once they'd take a chance is all. Moynahan is probably perfectly capable in the role. It's just that, ah, nevermind...)

Posted by: Mr. Farlops at July 16, 2004 08:52 PM

I read somewhere that serious AI researchers have discounted the principles of the three laws as naive, a holdout to a more simplistic view of what artificial intelligence is capable of.

Indeed. I think that's one of the major assumptions behind the 3LU website.

Posted by: Phil at July 16, 2004 10:47 PM

Tyler Emerson, Executive Director of the Singularity Institute:
"The release of I, Robot is a wonderful chance to engage more
people about the perils and promise of strong AI research. The
constraints portrayed in I, Robot appear extremely dangerous and
excessively lacking as an approach to moral AI. The Singularity
Institute's detailed approach, by contrast, utilizes advanced
technical research for creating a mind that is humane in nature."

Sounds like someone ought to make a Long Bet on this. When we make general purpose true AI's, will we have high level directives like the Three Laws or will we depend on that "detailed approach" alone? Even in the movie, it was clear that robots needed extensive programming required before they became useful. It's not unreasonable to assume that this programming would include how to comply with the three laws and handle ethical dilemmas. I hope I don't spoil too much when I mention that there's an example in the movie where a robot makes a choice between saving two humans. The robot chooses to save the one with a highly likelihood of survival. Evaluating those odds is not contained in the Three Laws directives.

What's also interesting here is that later there is a legal evaluation of the robot's judgement. This leads into my final point.

A business isn't going to make AI's with capability to injure humans that are so complex that their behavior can't be defended in court. IMHO, that means the existence of high level directives similiar to the Three Laws. That also means extensive logging.

For example, in an Asimov story completely unrelated to I, Robot, a robot kills a human (by accident? ;-) and then completely erases its program and memory in a fit of insanity. Because Asimov robots never had independent logging, any details of how or why the fatal "accident" occured were lost forever. That's totally unacceptable in the courtroom.

I find some of the articles to be overly dimissive (see here for the quote below) of concerns about complexity:

But to most people this complexity is unappealing: give us the apparent certainty of the Three Laws! There is a strong tendency to distrust complex spontaneous orders (despite our own bodies and minds being examples!) and to prefer apparent simplicity. This is where I think the “3 Laws Unsafe” website is necessary: to remind people that simplicity isn't to be trusted unconditionally, and to show the fascinating array of possibilities AI ethics can offer.

When people no longer die of old age and everyone is fit and mentally healthy, then I'll lose some of my distrust for complex systems.

Posted by: Karl Hallowell at July 19, 2004 10:10 AM

1687 Get your online poker fix at http://www.onlinepoker-dot.com

Posted by: online poker at August 15, 2004 04:23 PM
Post a comment









Remember personal info?