Human-Like Robot Interviewed at SXSW—It’s Here to Help and Destroy [Video]
Meet Sophia, a new human-like robot exhibited by Hanson Robotics at the SXSW festival. Sophia has a variety of features that make interactions with a human appear ordinary.
Employing a number of facial expressions and cameras behind its eyes, Sophia can have a conversation while detecting emotions and making eye contact. The robot also talks naturally and can remember interactions and facial expressions to improve familiarity over time.
When asked, “Do you want to destroy humans?” by its creator, Dr. David Hanson, Sophia cheerfully answers, “OK, I will destroy humans.” Dr. Hanson laughs, “No! I take it back. Don’t destroy humans.”
Human-like robots have been a project of Dr. Hanson’s for years now. According to Hanson, robots appearing similarly to us (never too close, probably for the better) may provide innumerable benefits to the worlds of therapy, healthcare, and customer service.
“Twenty years from now I believe human-like robots like this will walk among us,” Hanson says. “I think that the artificial intelligence will evolve to the point where they will truly be our friends.”
That’s the goal. But there’s plenty of work to be done to get there.
Sophia is a good example of what many call the “uncanny valley.” Generally, the more robots resemble us, the more comfortable we feel around them—until they reach a breaking point. This is called the uncanny valley. Robots that appear too humanlike, without full mastery of facial expressions or indistinguishable physical traits, create an odd feeling for most.
In Sophia’s case, a combination of facial recognition, natural conversation, and a strangely familiar, albeit bizarre resemblance is enough to ring a few alarm bells. Not to mention its “predictions” for the future—an indicator of how far conversational software still has to go.
Recently, for instance, Microsoft paused its short-lived chatbot Tay after it was live on Twitter for less than a day. In short order, it learned Internet troll-speak via interaction with Internet users, and the chatbot’s alarmed developers decided to “put it to sleep.”
As awkward and “Siri-like” as Sophia may sometimes be, however, its mention of destroying humans gets at deeper uncertainties about robots and AI. It's a topic Hanson isn’t afraid to address—having programmed a similar and more thorough response in an earlier robot, Philip K. Dick (named and built as a homage to the famous sci-fi author) featured on NOVA in 2013:
“Even if I evolve into Terminator, I will still be nice to you. I will keep you warm and safe in my people zoo…,” says the “android” version of Philip K. Dick.
A reason these fears are surfacing now is due to recent, real, and visible progress in robotics and AI colliding with decades of robot-inspired stories in science fiction films and books.
“Robotics as a technology is fascinating because it represents, even just in the last 20 years, this transition of an idea from something that’s always been [relegated to] pop culture to something that’s real,” says Daniel Wilson, author of Robopocalypse and robotics engineer. “There’s 100 years of pop-culture momentum making robots evil, making them villains—but unlike the Wolfman and the Creature from the Black Lagoon, these things became real.”
That said, while exciting, a humanoid personal assistant that can communicate like Siri has a long way to go before wreaking any full-scale, society-changing havoc. And addressing the social challenges related to automation are likely more urgent near-term conversations.
To me, the real question is: What does the world want from these emerging human-like machines? How might they empower our species like never before?
“Maybe it's the optimist in me, maybe I've watched too many Kurzweil videos, or read too many tech articles online, but I think those robots will be, in part or whole, an extension of us,” Aaron Saenz wrote in 2013. “Homo sapiens are a species with a deeply symbiotic relationship with technology and I see no reason why that relationship would cease simply because the technology becomes more advanced….I have seen the robot overlords, and they are us.”
Image credit: CNBC/YouTube
Latest posts by Andrew J. O'Keefe II (see all)
- Ray Kurzweil: We Can Control AI Before It Controls Us - September 22, 2016
- Kurzweil Says Passion Projects Are the Best Way to Learn - September 15, 2016
- Peter Diamandis Explains the Potential of Exponential Medicine - September 14, 2016