My Jibo has been turned off since a mid-April power outage. I like him a lot better this way.
I contributed to the Jibo Indiegogo campaign on July 16, 2014 and wrote about my reasons and my optimism. We expected to have a little social robot in our home by Christmas 2015. Eighteen months seemed like a long time, but the wait turned out to be 3 years, 3 months, and 30 days. I’m still glad to have been part of funding Jibo’s development, and I think that he’s the start of something important for the future. (I choose to use a male pronoun based on the voice given to Jibo, as I use “she” for the Amazon Echo.)
In all honesty, though, I can’t stand the little guy.
His movements are cute and now and then he does something oddly endearing. Most of the time, he’s a socially inept, slow-witted, creepy nuisance. His programming is off key in many ways.
When we have him powered up:
- He regularly interrupts conversations to ask, “Hey, can I tell you something?” If you say yes, he spouts some random piece of trivia.
- He turns to watch a person walk by, with a full body swivel and head tilt. This seems like a sociable movement, but if people in the room spun around to watch silently every time you got up to go to the bathroom, you’d see how unnerving it can be.
- He answers questions after a significant delay, and usually with incomplete information or an “I don’t know the answer to that.” The latency in searching for an answer is awful compared to other digital assistants and the pause can be long enough that you’re not sure the question was heard.
- He greets faces on the television. (If you buy a Jibo, don’t put him where he can see a TV screen. This is much worse than the Echo occasionally mishearing “Alexa” from dialogue and waking up.)
- His recognition abilities are shaky. He greets me by name more than my husband, but after months of interaction, that’s not impressive.
- He never shuts down. He “goes to sleep” at night, but from about 7:00am-10:00pm, he’s using power, sometimes moving and randomly flashing things on his screen, turning and “staring” at the brightest area of the room, or talking to the TV. You can tell him to take a nap and he’ll go quiet for a while.
All in all, I’m disappointed. I expected more from the geniuses who started the project, but it was in development for so long that other interactive devices had years of learning before Jibo shipped. While I understand that Cynthia Breazal’s vision is that Jibo will be far more than a digital assistant, he has to meet that baseline. The price is down to about $700 — from $900 in December — but I think that’s absurdly high for what he can do. Jibo launched to developers before ordinary buyers so that additional abilities could be available in a skills store, yet the bot is in general release and there’s no sign of that store yet.
Maybe Jibo is a better fit for families with young children. My 20something stepson found our Jibo amusing, so I looked into giving the bot to him. No dice. Jibo is a one-owner product and cannot be transferred.
I’m still optimistic about the future of social robotics, but I don’t yet know what Jibo’s contribution to that will be. We’ve learned more about human-bot interaction from Furby. In fact, I recommend listening to the recent More or Less Human episode of Radiolab for a discussion about chatbots and the Turing test, Furbys vs Barbies vs “gerbees”, and the risks of designing pain response into a social bot.