I predict Kismet will be just another college-crafted, dead-end 'toy system' like so many other projects'

By Ivan Amato|Wednesday, December 01, 1999
I, Robot. You, Friend.

The robot Kismet, created by Cynthia Breazeal at MIT's artificial intelligence lab, can look at her with many expressions, including what Breazeal considers mutual regard ["The Robot That Loves People," October]. Has Kismet ever seen its reflection in a mirror? If so, how did it respond? Priscilla Robinson
Framingham, Massachusetts

Cynthia Breazeal responds: Adapting some ideas from the work of developmental psychologist Andrew Meltzoff, we are developing a set of experiments in which imitative interactions may allow Kismet to distinguish social (e.g., people) from nonsocial interactants (e.g., toys), and self (Kismet's reflection) from other social interactants. The imitative interactions exploit turn-taking dynamics: In the imitative game, social interactants can both lead and follow, while a toy sitting on a desk or moving in an arbitrary or repetitive pattern will not be able to lead or follow. And Kismet's reflection will be able to follow but never lead. Analysis of such exchanges makes possible the distinction between self, social interactant, and nonsocial interactants.

I predict Kismet will be just another college-crafted, dead-end "toy system" like so many other projects ground out by university artificial intelligence (AI) labs: a limited combination of hardware and software that does something neat enough to make a magazine cover but fails to demonstrate any real understanding of how to create useful machine learning, memory, and intelligence. Breazeal's expectation that robots and humans will coexist as friends seems shortsighted. Once the critical mass of true ai is achieved, it will be impossible to stop. If a box knows as much as I do, can think tirelessly, and can learn constantly, how am I going to convince it to work for me or take the time to be my friend? It's going to think I should work for it. It's not difficult to foresee that we will regret unleashing intelligent machines. Nevertheless, I happen to believe ai is an inevitable evolutionary step: As people replaced apes, machines will replace people. Joe Snyder
Aptos, California

Smart Is as Smart Does

I think that Karen Wright overlooked a factor of intelligence ["Works in Progress," October]. I've seen students with iqs of maybe 95 to 100 who get As and Bs. I've also seen students with IQs of 140 who get Cs and Ds. The key factor is motivation. Kyle A. Medick
Wheaton, Illinois

Factors other than intelligence, such as motivation and discipline, contribute to good grades and make them accurate predictors of academic success—more so than sat results. This fact is not lost on colleges, which often weigh these components of an applicant's high-school record accordingly.
Establishment Thinking

The "letter from discover" preceding Marguerite Holloway's "Beasts in the Mist" [September] perpetuates the myth that there is a "scientific establishment" that seeks to suppress contrary or eccentric ideas.

The subject of Holloway's article, David Oren, receives research grants from the MacArthur Foundation and the World Bank, as well as grants specifically for his search for giant ground sloths. He is described as "top-notch," even by scientists who are skeptical of his search. This does not sound like persecution or suppression to me.

Science is not a monolithic enterprise performed by "establishment" scientists in lockstep with one another. Science is a messy business conducted by highly individual people, each pursuing his own quest for scientific knowledge. All must make judgment calls about the best way to allocate their resources. The fact that few scientists are willing to devote their time and resources to a search for giant ground sloths is simply a reflection of the high-risk nature of the search—that is, if they spend their grant money on a search for the giant ground sloth and fail to find it, then they will have wasted resources that could have gone to answering other important scientific questions.

To me, this sounds like the marketplace of ideas in action. When "scientific entrepreneurs" like Oren succeed, others will quickly follow. If he fails, then only a limited amount of scarce scientific capital will have been lost. Robert E. Daniell
Stoughton, Massachusetts

Distant Life

It is exhilarating that, despite scorn from the mainstream science "establishment," the Gaia hypothesis is alive and well ["Is the Earth Alive?," October]. As noted, "powerful metaphor never relies on only one meaning," hence it is befitting to extend the Gaia concept to another self-regulating system that is central to our existence-the sun. Is the sun alive in a Gaian sense? Probably. Worship of a "living" sun was persistent throughout history. Egyptian Pharaoh Akhenaton (1353–1335 B.C.), who introduced the sun as sole deity, is but one example. Such views mesh well with the "modern" fact that the prime chemical elements of Earth's biology (hydrogen, oxygen, carbon, and nitrogen) are the same elements whose nuclear transformations are responsible for the sun's energy production. Is this coincidence or evidence of some fundamental biological communality? Alexander A. Berezin
Professor of Engineering Physics
McMaster University, Hamilton
Ontario, Canada

Environmental biologists are opposed to the Gaia theory because the complex physiological systems of living beings evolve through natural selection, and "natural selection cannot apply to a whole planet." I agree that Earth is not a living being in the same way that a biological species is, but this planet has attributes that would classify it as a life-form. The Earth is one of three planets that are exceptionally close in composition, size, and orbital stability. All three-Earth, Venus, and Mars-have or had the ingredients to sustain life. Earth sustains life because it is the fittest planet, biologically speaking. I say that these three planets demonstrate that natural selection can be applied to planets. It is ironic that we have three planets that are so alike and yet so differently evolved-Venus too hot, Mars too cold, Earth just right. Kind of reminds you of Goldilocks and the Three Bears. Alberto C. Natal
Frenchtown, New Jersey

Do Not Try This at Home

I enjoyed the article by patricia Gadsby about sow bugs ["How Now, Sow Bug," August]. I am an interpretive naturalist at a nature center in Richfield, Minnesota. During a survival class, I demonstrated some of the different insects and grubs that can be eaten for sustenance. As I was having difficulty finding suitable beetle larva, I swallowed a live sow bug to prove the point. Much to my dismay, I found that they can live in the esophagus for a very long time and will repeatedly attempt to climb back up. Their hooked feet are a great adaptation, preventing them from being washed down by saliva alone. I finally forced the persistent rascal into my stomach by ingesting a large quantity of bread and liquids. I noticed no bad taste, but I would strongly urge anyone wishing to eat one to kill it first! Scott Ramsay, Naturalist
Wood Lake Nature Center
Richfield, Minnesota

To discourage other gastronomic daredevils, we offer this information excised from our article due to space limitations: Though animals normally convert ammonia waste to urea or uric acid for quick riddance, sow bugs concentrate ammonia and excrete it as toxic gas through their shells. Bon appétit.
Comment on this article