In early 2009, news broke that an experimental Toshiba robot that had been “programmed to love” was malfunctioning. The story, paired with the above image of a half-Frankenstein, half-Asimo automaton running amok with a captive damsel-in-distress, graced some of the internet’s most-read tech blogs and news sites. Gizmodo, IGN, the Next Web, and others all ran stories about Kenji, the robot-turned-stalker that refused to stop hugging a terrified female intern.
It was a slightly chilling and totally titillating narrative, and it played directly into our science fiction-fed imaginations—the ones still avidly populated by increasingly sentient and sure-to-short-circuit robots. It was also a complete hoax.
Yet the robo tall tale continues to quietly circulate around the web four years later, masquerading as blogged truth. This April Fool’s Day, when all of our feeds are brimming with gags, fictions, and jokey might-as-well-be-truths, it’s worth taking a look at how a minor, haphazard hoax has grown a rather long tail.
Word that Kenji the love-struck robot was still lurking around online was delivered directly to my inbox. I have a Google Alert set for ‘robots’, as I’d like to receive any news of the forthcoming robot apocalypse as promptly as possible. Typically, the stuff Google sends me are the ordinary robotics stories of the day: Stories about scientists’ newly invented robots, about robots that assist in new medical procedures, about advances in military robotics, about robots that are taking people’s jobs.
But amongst the usual suspects was an item about a machine built to love. Honestly, it didn’t stand out too much as overly preposterous, given the sensationalist leanings of blog headlines—even though it was titled Robot Programmed to Fall in Love with a Girl Goes Too Far .
The tab was open on my browser for a couple days before I got around to reading it, and it was clear within sentences that it was pure fiction:
“Researchers at Toshiba’s Akimu Robotic Research Institute were thrilled ten months ago when they successfully programmed Kenji, a third generation humanoid robot, to convincingly emulate certain human emotions,” it read. “At the time, they even claimed that Kenji was capable of the robot equivalent of love. Now, however, they fear that his programming has taken an extreme turn for the worst.”
Right. Because there’s a single scientist alive who would casually claim to have pinpointed “the robot equivalent of love.” The story was published at a site called Reality Pod, which I’d never heard of. It seemed like a pretty well-visited site that, while chock-full of traffic-whoring listicles and sensational headlines, didn’t appear to stoop to the outright hackery of non-Onion fake-news sites. It wasn’t supposed to be a joke.
The article also wasn’t new. For some reason, it was making the rounds again, after initially being published in October 2010. It had 54,000 shares on Facebook. I hit refresh—remember, it’d sat idle a day or two. Now it had 61K. A couple of the commenters assumed it was a joke, but most bought it as news.
The hoax turned out to date back at least to 2009, when Gizmodo, The Next Web, and IGN were all suckered by the exact same story of robot love gone awry. The source of each (besides IGN, which acknowledged none) was a blog called Geekologie, which had run a post titled ‘ Wow, I’m Shocked: Robot Programmed to Love Goes Too Far, Commences Stalking ’.
But even Geekologie had a deeper source—a broken link to a site called MuckFlash. And that now-defunct, amateur Wordpress blog was Kenji’s creator. Geekologie excerpts the since-deleted story as follows:
The trouble all started when a young female intern began to spend several hours each day with Kenji, testing his systems and loading new software routines. When it came time to leave one evening, however, Kenji refused to let her out of his lab enclosure and used his bulky mechanical body to block her exit and hug her repeatedly. The intern was only able to escape after she had frantically phoned two senior staff members to come and temporarily de-activate Kenji.
Dr. Takahashi admits that they will more than likely have to decommission Kenji permanently, but he's optimistic about one day succeeding where Kenji failed. "This is only a minor setback. I have full faith that we will one day live side by side with, and eventually love and be loved by, robots," he said.
What a sad and lonely man the fictional Dr. Takahashi must be, the internet was supposed to muse.
MuckFlash appears to be an early hoax news website; it’s tagline is “Breaking and Broken. Odd, funny news from all over. Source what you can, enjoy the rest.” A typical example of another one of its fairly widely cited, completely made-up stories is ‘Woman Gives Birth to a Block of Cheese’.
But its robot love story might have brought too much heat too quickly— MuckFlash appears to have stopped posting new stories right around February 2009, shortly after the robot love story took off. Its Twitter feed went dead not too much later.
I haven’t had any luck in tracking down the admin of MuckFlash, who appears to have operated the blog out of Utah, so we don’t really know why, exactly, he was inspired to whip up a story about a stalker love-bot named Kenji. I couldn’t find any live email address, and efforts to contact the admin at the Twitter handle proved fruitless.
So the true source of the story is sadly left to conjecture. Toshiba’s “Akimu Robotic Research Institute” isn’t real, though the electronics firm does have a well-known robotics division. One detail raised my eyebrow—Kenji was the name of the first man believed to be killed by a robot. According to the Economist, Kenji Urada was a 37 year-old factory worker in Japan who neglected to switch a robot off properly when he went to repair it in 1981. The hydraulic arm kept working as he drew near, and it pushed him into a grinding machine.
As for the rest, the MuckFlash blogger was probably just trolling, exploiting our sci-fi expectations of man’s hubris with robot research—of course if you program a machine to love, things are going to go wrong. And people will click on the headline. Yet it’s a little crazy that anyone fell for this story—even at the apex of the content-hungry blog age of the late 00s (and I was there, man)—especially since it was pushed with that absurd picture.
But they did. And still do, kind of, even today. Geekologie, the Next Web and IGN still have the story live on their sites, without any update or correction. The minor item about robot love gone awry was simply ingested into the blogosphere. Only Gizmodo actually updated when commenters pointed out that the robot pictured was a well-known medical robot, and that the details didn’t add up.
Clearly, Kenji still lives on. It’s getting reposted at cheap content-harvesting sites, it’s getting re-‘liked’ and re-‘shared.’ If you search Google for ‘robot programmed to’ hoax-related results populate the field.
Maybe Kenji’s still around because his is the story that we know in our guts is coming one day, whether it is or not—it’s both what we desperately want and don’t want to believe robots are capable of. It’s maybe why editors didn’t bother to correct it, or weren’t even ever alerted to the falsehood, or why ‘Kenji robot’ and the ‘Toshiba Akimu Robotic Research Institute’ still auto-complete in Google search terms, or why the story still gets excerpted.
Because it doesn't seem so crazy, the notion of a humanoid robot getting attached to a human, and the events detailed in the story itself are just innocuous enough to seem believable. If the bot were to have been charged with murder, say, this thing would have been roundly and immediately debunked.
As it is, the poorly-written, clearly apocryphal anecdote limps around on the periphery of the internet, almost true, scoffed at or somberly noted when shared or discovered in a search with those unfortunate keywords. As with the ever-growing mass of successful April Fool’s posts, fake news items, and press release reprints that are beginning to clog up the corners of the internet, Kenji reminds us again that the internet is fertile ground for lies that we want to believe. Especially about robots.
Inlaid image: Flickr, CC