Nand Kishor Contributor

Nand Kishor is the Product Manager of House of Bots. After finishing his studies in computer science, he ideated & re-launched Real Estate Business Intelligence Tool, where he created one of the leading Business Intelligence Tool for property price analysis in 2012. He also writes, research and sharing knowledge about Artificial Intelligence (AI), Machine Learning (ML), Data Science, Big Data, Python Language etc... ...

Full Bio 
Follow on

Nand Kishor is the Product Manager of House of Bots. After finishing his studies in computer science, he ideated & re-launched Real Estate Business Intelligence Tool, where he created one of the leading Business Intelligence Tool for property price analysis in 2012. He also writes, research and sharing knowledge about Artificial Intelligence (AI), Machine Learning (ML), Data Science, Big Data, Python Language etc...

3 Best Programming Languages For Internet of Things Development In 2018
281 days ago

Data science is the big draw in business schools
454 days ago

7 Effective Methods for Fitting a Liner
464 days ago

3 Thoughts on Why Deep Learning Works So Well
464 days ago

3 million at risk from the rise of robots
464 days ago

Top 10 Hot Artificial Intelligence (AI) Technologies
305991 views

Here's why so many data scientists are leaving their jobs
79902 views

Want to be a millionaire before you turn 25? Study artificial intelligence or machine learning
74442 views

2018 Data Science Interview Questions for Top Tech Companies
73767 views

Google announces scholarship program to train 1.3 lakh Indian developers in emerging technologies
60630 views

Chatbots Artificial Intelligence FB Chatbots Tech Skills Chatbot Marketer

How The Facebook Chatbots Story Got Overblown: 5 Key Lessons For Marketers

By Nand Kishor |Email | Aug 21, 2017 | 4308 Views

If you've been paying attention to social media for the past few weeks, you've likely caught wind of a mildly disturbing story about a pair of Facebook-developed chatbots that invented their own language. The story has been shared millions of times, from different publishers' perspectives, and has led to both serious discussions and amused jokes about the impending "AI apocalypse."

But here's the thing: the story got completely overblown, and most people sharing the story never learned what actually happened. As fascinating as this is from a cultural and scientific perspective, it's even more fascinating from the perspective of a content marketer.

So if you're currently running a content campaign, or any marketing/advertising/PR campaign, take a look at what happened here-and what you can learn from the incident.

The Blowup
Before we look at the facts, let's look at how the story was presented by some of the top-shared publications. One of the top headlines: "Facebook shuts down robots after they invent their own language." Similarly: "Facebook Shuts Down AI Robot After It Creates Its Own Language."

The title certainly captures audience attentions and imaginations, illustrating a story of deceptive, human-like robot intelligences that schemed to create a covert language that their human creators couldn't understand. None of this is explicitly stated in the title, but the alarmist tone and personification of the chatbots in question seem to imply such a narrative.

What Actually Happened

So what actually happened here? It doesn't take much digging to find out. In fact, Facebook reported about this way back in June, when it actually happened. They've been completely open about all their AI experiments, even releasing all available data on their experiments to the public as soon as it becomes available. It's only now, nearly two months later, that publishers picked up on this potential story.

Facebook researchers did set up an experiment with twin AI-driven chatbots, designed to communicate with each other and learn from each other to communicate as efficiently as possible. And did they invent their own language? Kind of.

If you read the actual back-and-forth messages, the indecipherable text isn't a "language" so much as it is gibberish with bits and pieces of English in the mix, such as "I can can I I everything else."

The experiment wasn't shut down because this wasn't a major concern, nor was it even a surprise to researchers. In fact, it's entirely natural for AI bots to discover and create shorthand to communicate with fewer pieces of text, the same way we abbreviate phrases like "LOL" and "omw." After headlines started rolling, researchers heavily stressed their objections to the way the story was being presented.

Key Lessons

So let's get to the meat of what happened here. What can content marketers take away from this fantastical blowup?

  • People don't read beyond headlines. First, this story makes it apparent that most social media users simple don't read anything other than headlines. In fact, research has already shown this to be the case, with 6 in 10 users sharing articles they haven't even clicked on. What does that mean for you? It means a headline can mean far more to the success of your content than the content itself. The story of an almost-routine scientific AI experiment is way less exciting than a close call with a superior machine intelligence.
  • Timing isn't always a necessity. In content marketing, we often think about the timing of our work, especially when it's related to current events. We want to respond to news as quickly as possible, and churn out work at a faster and faster pace. This story shows us that timing isn't a necessity; after all, this story gained traction almost two months after it was initially reported. You can find gems to report on dating weeks to months back, and you can dig up old information to make it fresh again, so don't rule something out just because it isn't hot off the presses.
  • Original reporting is rare (and a huge commodity). Facebook's original report on the experiment received almost no traction, but is full of highly detailed, important information. The only problem is that it failed to present that information in a way that made the community excited to receive it. Original research (such as information from surveys or experiments), is rare, and having it presented in an entertaining, approachable way is even rarer. It's up to you to fill that void in your industry.
  • Fact checking is vital. If you're reporting on something else, or basing your article on someone else's presentation, make sure to get the facts straight. Though I haven't performed a root cause analysis, my guess is this specific incident started with one exaggerated report, followed by a few dozen other reports that mirrored what it said without checking the facts themselves. Before you assume your source is correct, follow up on the details.
  • Emotional appeals work (but be careful). One of the biggest reasons for this article's success was its play on human fears of an AI takeover. Marketing to people with strong emotions (including fear) can be a powerful way to gain more attention-but again, you have to ground your headlines in reality if you want to protect your reputation.

As both a content marketer and a tech enthusiast interested in AI, I'm simultaneously appalled at how this story blew up, and endlessly amused at the situation. Fortunately, as a content creator, you and I have the power to control how stories like this develop, how they're told and retold, and how we engage with our audiences overall.

Let's take this opportunity to arm ourselves with more responsible research and presentation methods, and improve our ability to persuade readers-as ethically as possible.

Source: Forbes
Previous Story Next Story