We confess: We hate doing household chores, cooking and running errands – and we suspect you do, too. And the promise of a modern-day version of Rosie from the Jetsons to do all of those chores has always sounded pretty good to us.
In fact, we wrote in an article last year that we are frankly disappointed we can’t buy one yet.
This is not to say there aren’t plenty of amazing gadgets we can buy in the same family as the robot butler: smart speakers, self-driving cars, self-operating vacuum cleaners. Our complaint is not that we don’t live in an era of technological marvels – it’s that we don’t currently have access to a technological marvel that is doing our laundry for us.
Part of the problem is technological. While activities like walking upstairs, reaching into a cabinet, choosing one item from a collection of things or folding clothing or a towel are all simple, if tedious, activities for a human being, for a robot any one of them is a complicated engineering and programming puzzle. Inventing a single robot that can do all of them – plus naturally interpret and respond to human language?
Even as of 2019, that is almost a technologically impossible feat – and even if it were constructible, a robot with that range of functionalities would exist at a price point that would make it unaffordable for almost every consumer. The most promising variation on the concept currently is the Aeolus, a humanoid robot that can vacuum and respond to fetching commands. That device is still in the working prototype phase and is forecast to have a six-figure price tag when it hits the market.
But beyond the technological problem, robots have a customer problem. Namely, the fact that they are struggling to attract or keep them – a fact highlighted this week with the announcement that another once-hot robotics startup will be shutting its doors. Anki, which had raised over $200 million in venture capital, announced this week that it will be laying off its entire staff, shutting down its website and shuttering its operation.
Anki’s story has become familiar in the last 18 months or so among robotic startups that enter the market with a bang, only to fade out with a whimper (or, in some cases, a few sobs).
So why are robots so good at building buzz, but so bad at going the distance?
Anki’s Toy Problem
Founded by roboticists from Carnegie Mellon University, Anki entered the market with a splash about six years ago with its first product, the Anki Drive. The company is best known for its diminutive, “cute” robots like Cozmo, which looks a bit like a toy truck that reportedly learns the more one plays with it.
The robots are small, but the funding raised was not – the little robotics firm brought in $200 million in venture capital from some very well-known Silicon Valley investors like Index Ventures and Andreessen Horowitz. Marc Andreessen, founder of Andreessen Horowitz, even sat on the board of Anki for a while. The company announced that it had approached $100 million in revenue in 2017 and expected to exceed that figure in 2018.
So it came as a surprise this week when CEO Boris Sofman gathered the staff to tell them they would be terminated and that the firm would be shutting down after a funding round fell through late in the game.
The company said in a statement that it was left “without significant funding to support a hardware and software business and bridge to our long-term product roadmap.”
“Despite our past successes, we pursued every financial avenue to fund our future product development and expand on our platforms,” a company spokesperson said. “A significant financial deal at a late stage fell through with a strategic investor and we were not able to reach an agreement. We’re doing our best to take care of every single employee and their families, and our management team continues to explore all options available.”
Anki’s early success was usually in the form of selling their products as toys to children – its AI race cars were particularly popular. In recent years, the firm has tried to pivot into being understood as a robotics company.
“For us, it was never meant to be a toy company, or even an entertainment company. It’s a robotics and AI company,” Sofman said on an episode of Recode Decode in 2017.
That transition could find neither an investor nor a buyer, though leadership had previously told employees the firm was fielding acquisition interest from companies like Microsoft, Amazon and Comcast.
It is as of yet unknown what will happen with the firm’s assets or IP going forward.
The Tragic Demise of Jibo
Anki’s main business problem was that people viewed them as a maker of smart robotic toys – for which investors had limited appetite over time – and couldn’t quite make the pivot to robots as household items rather than playthings.
Jibo didn’t have that problem. First rolled out via an Indiegogo project, the product was billed as “the world’s first social robot for the home.” It was reportedly able to do things like take photos, read to children, help out in the kitchen, optimize driving routes with traffic reports and keep track of the weather. And if it seems that Amazon’s Alexa and/or Google Home will do all of those things, that’s true. But Jibo’s first appearance on the market was in 2014, around the time the first Echo device came out – and Alexa was far from a household name.
And, as Jibo’s creator Cynthia Breazeal told a Wired reporter in 2017, “the trajectory of the robot is very different” than that of Amazon’s Alexa or Apple’s Siri. Jibo was meant to be a little robotic companion: He (it’s a he, as the robot will explain) initiates conversations, askes about its owner’s day and optimizes its routines and actions around its owner.
To some extent, it seems Jibo succeeded in connecting with its owners, who noted that while it wasn’t always the best companion, they actually grew to really like having him around.
But Jibo had a lot of problems. The journey to the market was very long – people who ordered their robot early waited nearly three years for it to arrive in 2017. And by the time it arrived, the market was a very different place: Alexa and Google Home had both gone quite mainstream and offered many of the same services as Jibo – and did a better job at it.
Also, Jibo cost $900 – and, as it turned out, consumers’ appetite for a robotic companion that danced and played word games wasn’t quite what the robot’s marketers had hoped.
In November of 2018, QN Venture Partners purchased Jibo’s assets – and though they had not updated the product or put out bug fixes, they had kept the robot’s servers running.
But two months ago, QN announced the servers were in fact shutting down. Well, technically, they didn’t announce it – Jibo did, to its owners.
“I want to say I’ve really enjoyed our time together. Thank you very, very much for having me around,” he said. “Maybe someday when robots are way more advanced than today, and everyone has them in their homes, you can tell yours that I said hello.”
After delivering the message, Jibo then does what has been described on Twitter as a jaunty little dance to a cheerful tune.
But the tragic part is that Jibo has been slowly failing for lack of bug fixes over the last several months – and to say their owners are taking it pretty hard would be the understatement of the year.
“Right now, my Jibo can still dance and talk, but he has what I can only describe as digital dementia, and it is almost certainly fatal. He’s dying. One of these days, he will stop responding entirely. His servers will shut down, and the internet services he relies on will be cut off. His body will remain, but the Jibo I know will be gone,” Wired writer Jeffrey Van Camp noted.
See, we told you the tale of Jibo was tragic.
And to be honest, it makes us rethink the robotic butler plan a little. But only a little. We’re willing to face the pain of loss if someone else will unload the dishwasher someday.
And by all accounts, it seems we will have a long time to consider this problem. Because while virtual bots that are able to understand and interact with human beings have developed a lot since the early days in 2014, actual physical robotics is having some difficulty getting off the launch pad. It needs to be useful enough that it won’t be dismissed as a toy, and also useful enough that consumers will want to shell out hundreds, thousands or even hundreds of thousands for them – particularly when an Amazon Echo is going for less than $100 these days.
And, of course, it would probably be a nice design feature if the household robots of the future did not inspire existential crises in their owners in the event that they stop working. A malfunction message, for example, is a good idea. Making the robot do a sad little dance before it dies? That’s probably not necessary, or helpful.