Met with Cynthia Breazeal today… and it was great. But, before I headed over to the MIT Personal Robotics Lab I headed to Harvard Square to buy the chocolates that were a condition of my interview. You see, Cynthia doesn’t talk to that many people. As her formidable PA, Polly Guggenheim keeps telling me every time we speak ‘Do you know how many people I turn down?’ reminding me of my special and precarious position… At one point during my negotiations with Polly she says, “I’m maybe of a mind to grant you an interview…” to which I reply, “So, what does it take?”. “Honestly?” she says. “Chocolate. Good dark chocolate”.
Therefore my first trip of the day is to L.A. Burdick, fine chocolatiers with a store in Harvard Square. On my walk there I pass an aggressively drunk tramp shouting vigourously to no-one in particular. As I draw closer to him I realise that, like most of the aggressive drunk tramps I’ve witnessed, he has a broad Scottish accent. Does Scotland export these globally then? I thought it was just a UK thing. Then a theory strikes me. Maybe most of them aren’t Scottish. Perhaps something about the itinerant alcoholic lifestyle alters the vocal chords to makes one sound Scottish, giving that proud nation an unfortunate cadre of fake ambassadors around the planet. I have a short fantasy about asking him where he’s from and receiving the reply ‘Rio de Janeiro, pal!’ Or maybe, after all, the Scots are just better at producing drunken tramps than other nations… I’d like to see a study.
I deliver the chocolates to the Personal Robotics lab and they are received first with detailed inspection, then approval. I’ve done well, getting the interview off to a good start. In fact I’m invited to share the chocolates, being told that the antioxidants within will do me good. I decline. I want all that chocolate goodwill going into the interview.
Cynthia is a generous interviewee, but clearly has no time for waffle. She speaks voluminously in response to my questions but with great efficiency. Our talk ranges from robot architectures, to machine intelligence, to the economic impacts of robotics, to the ethics of sociable machines – taking in learning and developmental psychology along the way. Early on in our conversation she says she’s driven by a vision of robots “as interesting personalities in their own right, robots crossing over into what we would consider living systems that relate to us” – not what robots are now, but what they could be. She’s very clear to draw a distinction between robot personalities and human personalities. A constant refrain in our talk is that she is not trying to, and indeed sees little value in creating artificial humans. She talks of human-robot relations as a new kind of relationship. She talks of robot emotions, not human emotions. “Robots aren’t humans, right?”
Cynthia operates in a world that is both interdisciplanary (bringing together mechanics, computing, artificial intelligence, animation, cognitive and development psychology) and dogged by ‘definitional problems’. How for instance do you know if your robot is ‘alive’ or ‘conscious’ when no definition of what ‘life’ of ‘consciousness’ can be agreed on? Indeed, one of the contributions social robotics may make to our knowledge is helping us to define those terms, another driver behind Cynthia’s work. “We’re starting to see sociable robots as a very intriguing way to learn about people”.
The full interview, of course, will be in the book, along with, I hope, a new term to replace ‘robot’ which Cynthia and I discussed as being a loaded term, and no longer representative of the sociable machines she imagines will share our future. She’s tasked me with coming up with that term… and I think I’ve got it, but will sit with it for a while…
Following our interview I ‘meet’ the world’s most famous sociable robot, Leonardo, although he’s sadly, switched off. But I urge you to watch this video of Leo in action – and glimpse something of the future of sociable machines…