MetroLutheran

Machines help Lutheran researcher ponder what is human

Maybe God should keep an eye on Anne Foerst. She could be trouble.
The German pastor helps create intelligent robots — and thinks God
approves. “Whenever we rebuild ourselves,” says Foerst, “we celebrate God.”
Indeed, she thinks God helps in her work. “I single-handedly invented the
field of robot theology,” claims Foerst, who teaches at St. Bonaventure (N.Y.)
University. “You can’t do that without praying for a lot of guidance.”
Foerst, who spoke at Augsburg College in Minneapolis earlier this year, calls
her work in artificial intelligence (AI) both scientific and pastoral. Designing
machines that resemble humans teaches us about ourselves — and about our
creator. What is a person? Why do we reject some? “I definitely see my job as
a pastoral job,” says Foerst. “I try to make people behave better towards one
another.”
AI research, says Foerst, “broadens and softens how complicated denial of
personhood is.” The work, she adds, “makes you very humble and in awe of
the complexity of God’s creation when we try to rebuild ourselves because we
are so darn complex.”
Foerst’s work quickly draws attention, but doubters aren’t hard to find. Larry
Crockett, himself an Episcopal priest and computer scientist at Augsburg,
follows AI — but is wary of Foerst’s claims.
Her work is a “novel combination,” says Crockett, who calls himself a
“skeptic” about AI and robotics, doubting that research in those areas will
create touchpoints with religion.
Nevertheless, Crockett says Foerst is “engaging and very bright” — a
“compelling figure and given to some delightfully provocative claims.”
In any case, keeping an eye on the fast-developing field of robotics may be
wise. Robots already help with manufacturing and surgery. They can
recognize speech. They can move around and avoid obstacles. One can load a
dishwasher. Another can make coffee. A few read human facial expressions.
Robots may someday be companions for the elderly or children.
How should we treat robots? As machines? Animals? Human? If we make
robots in our image, are robots then also in the image of God? What happens
if one doesn’t want to be turned off or taken apart?
Christian theology gives humans dignity and intrinsic value. Science, on the
other hand, sees humans as products of evolution following psychological,
biological, and behavioral rules. The two distinct views are related — but
don’t mix them, suggests Foerst.
The Lutheran minister supported herself through eight years of higher
education in Germany by repairing computers. In the 1990s, Foerst served as
director of the God and Computers project at Massachusetts Institute of
Technology’s Artificial Intelligence Laboratory. She was theological advisor to
scientists who built a pair of robots named Kismet and Cog.
Building machines that resemble humans prompts designers to make
assumptions about humans — the same question, says Foerst, that
theologians explore: What is human?
Are the robots farther along than we think? Sometimes it seems so. Kismet is
essentially a head on a table, deliberately designed to be less intimidating
than its bigger sibling, Cog. Kismet has appealing features and “sucky” baby
lips — Foerst’s word.
During a media demonstration a few years ago, Kismet wasn’t responding.
Foerst pleaded in a stagy whine: “Kismet, you don’t like me any more.”
The result was startling. Kismet is programmed with four basic vocalizations
— sounds for soothing, scolding, approval, and a call for attention.
In response to Foerst’s whimper, Kismet abruptly “started babbling in
soothing tones toward me,” says Foerst, “and I felt better.”
Foerst thinks robots can teach us about rules that govern human
relationships. We feel both fear and attraction for these machines. Foerst
wonders if that volatile mix sheds light on tolerance and prejudice. Robots
for her are tools that help her think about why humans accept or reject
others.
Trying to build robots in our own image has a spiritual side that transcends
science. Humanoid robots can teach us more about ourselves, help us to
redefine what it means to be human, and make that definition more inclusive,
she maintains.
Is a robot a person? A chimpanzee that has been raised as a child in a human
family? A human fetus?
Cog is a machine, mounted on a table, vaguely human albeit legless. You can
still see Cog in an online video. Cog reaches for a ball, takes it from a
researcher, and then gives it back. In this video Cog looks like nothing so
much as a baby exploring its world.
Researchers hoped that Cog and Kismet would eventually speak, but the
robots never developed to that point. “Kismet could vocalize but had
absolutely no understanding of what it was saying,” says Foerst.
Cog and Kismet now are in a museum. They didn’t meet expectations. Does
that remind you of anyone? Adam? Eve? Us?
If robots reach a stage of development at which we can’t just turn them off —
how should we regard them? As pets? As humans? At what point does a
creature deserve to be treated as intrinsically valuable?
Researchers have identified the part of the human brain involved with
religious experience, notes Foerst. The theoretical ramifications for AI are
profound: “We know,” she adds, “how to produce a religious experience in a
robot.”
But dare we? And if we do, what happens when a robot that has managed to
learn to speak on its own says, “Tell me about Jesus”?
Or asks to be baptized? Or wants communion?