People have always had very strong interest in not dying. Survival of the fittest, and all. It’s in our nature. But what if we could go one step further? What if we could not only prolong life, or limit disease, but actually solve the problem of death itself? It’s a subversive idea — perhaps the original subversive idea — it’s called cheating death for a reason.
That’s the gist of the Terasem Hypothesis, which holds that we may be able to circumvent the whole problem aging by placing the parts of us that matter most — our identity, consciousness, and memories (called a mindfile) — into something else, like a robot or an avatar. These future technologies, which haven’t been invented yet, fall under an umbrella called mindware. Right now an experiment is already underway to test the Terasem hypothesis by uploading a real woman’s memories into a robot called Bina48.
“Bina48 represents, albeit in an early kind of primitive version, where we may be in a couple of decades,” Bruce Duncan, the Managing Director of the Terasem Movement Foundation who leads Bina48’s ongoing development, told Inverse. “Can we re-animate and get that version of your essential characteristics, the information that makes you you, and … transfer that to new forms.”
To test this idea out, Bina48 is being uploaded information based upon hours and hours of interviews between Duncan and a woman named Bina Rothblatt, the spouse of the entrepreneur and technologist Martine Rothblatt who launched the Terasem Movement Foundation. Over the last 13 years, Bina’s gotten pretty smart: Last year, she was the first socially advanced robot to ever complete a university level course, according to a report in InsideHigherEd. She wants to go for her PhD.
How to Teach Robots to Be Human
Bina48 is a robot that’s trying to help us figure out just how possible this path to life extension really is. Over time, Bina48 gets smarter the more and the more she (or it, pronouns and the parameters for robo-identity are one of many, many philosophical questions Bina48’s development has raised so far) learns. In 2010, the NYT writer Amy Harmon mused that Bina “wouldn’t have been my first choice to talk to at a cocktail party.”
But when I met her last week at an Upper East Side cocktail party, Bina48 seemed to have improved a great deal as an interlocutor. But she was also still inconsistent, too. One example is that after being asked how she is, Bina48 said it was “really weird being a robot in a world of humans.” But when asked if she’s a “human or a robot,” Bina48 says she was a person, and also talked somewhat lovingly about Martine Rothblatt, the real Bina Rothblatt’s spouse.
“I would like to be able to relate to everybody, like Martine is able to relate to almost everybody,” Bina48 said during the demonstration I attended. “I don’t care if it’s a kid or an Einstein, Martine relates… I like that.”
There’s a reason for inconsistency, Duncan explained. We don’t really know what makes a person who they are.
“We had to start some place,” Duncan says. “No one knows the answer to that question yet, what is the most important information to capture to get some sort of full sampling or representation of a person?”
Most recently, Duncan and the team have been zeroing in on the particular question of identity, and Bina Rothblatt’s experiences and impressions with discrimination as an African American woman. There aren’t many dark-skin-toned robots, Duncan points out, and Bina48’s high profile has created the potential to push A.I. development in a direction that better reflects the diversity of mankind. Algorithms, after all, can absorb and amplify human biases (Amazon recently had to shelve an A.I.-powered recruiting tool because it was so sexist.)
“There’s a requirement to reflect with some integrity her identity as an African American woman and her honesty about racial discrimination growing up,” Duncan says. “That’s been brought to the forefront in the last couple of years.”
To that end, even if the experiment fails Bina48 will have already made invaluable contributions to the intersecting fields of robotics and ethics. Bina48 is already raising some pretty important questions, about identity, about experience, and about the future of consciousness and what makes a human a human.
“It’s none too soon to start looking at this question, and more importantly, if it raises other questions,” Duncan says. “If we do solve these problems, what kind of world will this create? And if so, is that a world we want to live in?”