AI vs. The Robot Job-Stealers Narrative

Tuesday, May 9 2017 Doug Randall Categories:

AI may be in the spotlight –but it has some narrative headwinds to overcome


I recently attended TED in Vancouver, where Narratives about artificial intelligence (AI) were buzzing. The fervor and excitement behind the technology is extremely powerful and, I think, warranted. But I also found myself feeling cautious. Tech leaders have been especially bombastic recently about the wonders of AI while dismissing the concerns the technology raises. Eric Schmidt of Alphabet recently said in an interview:


“You’d have to convince yourself that a declining workforce and an ever-increasing idle force, the sum of that won’t generate more demand. That’s roughly the argument that you have to make. That’s never been true…in order to believe it’s different now, you have to believe that humans are not adaptable, that they’re not creative.”


That confidence might not resonate with those who are warier of the threats of AI. Protagonist has recently analyzed conversations around AI using our Narrative Analytics platform. Most of what we found was positive techno-friendly Narratives, but there are also very real, deeply held beliefs about AI as a threat to humanity, human jobs and human privacy that need to be addressed.


In today’s world it doesn’t matter whether specific types of AI present a real threat to human jobs or privacy; if they’re perceived to be dangerous the businesses behind them could be in real trouble. Negative affiliations could result in anything from investor slowdown, to active boycotts to slowed adoption during critical growth periods.


In Silicon Valley and tech markets, growth rates are particularly important and AI companies often experience surges of enthusiastic early adopters. When AI companies are strategizing for continued growth and allocating resources, they also need to think about how adoption trends might change when their product reaches the broader market. Negative narratives could create significant–even damning–headwinds if they aren’t accounted for and addressed directly.


But how powerful are these Narratives really? How closely are they tied to each business’s specific audience? Many companies know that these fear-based narratives exist, but disassociate them from their own customers or product. This is where Narrative Analytics becomes critical.


Understanding the Impact of Narratives of Distrust

Even without Narrative Analytics, most companies in the AI space can readily allow that Narratives like “robots will take our jobs” or “AI is a threat to humanity” exist somewhere out in the world. What they don’t know is how much that Narrative rests with their target audience or whether those Narratives are being applied to their own brand. Narrative Analytics exposes both.


As of last year, 10 percent of Americans considered AI a threat to humanity and six percent considered it a threat to jobs, though the former number was declining and the latter was rising. With that in mind, businesses should be aware of the very real risks of being labelled a job-killer. That Narrative might not be as broadly held as some others, but is one of the most evocative. Fear and anger over outsourced or inaccessible work opportunities played a major role in the last presidential election. It’s clearly a topic that resonates with people on a deep level.


Imagine if some Government official tweeted about a certain business costing Americans jobs by using AI. The backlash would be immediate, powerful and almost certainly unanticipated by the company. It would be a highly emotional response from an audience the company might not have been prepared to engage with. The effect could be devastating, but it wouldn’t have happened in a vacuum. Large-scale analyses of social engagement across the country (rather than small and biased sample groups) and media tone reveal risks and opportunities in AI Narratives before they come to a head.


Our analysis revealed a cautionary finding: the less tech-friendly Narratives about AI have significantly higher levels of engagement than the more positive tropes. That means they’re more likely to spread quickly once they are triggered.


So what can businesses implementing AI do? First, understand the Narrative Landscape, then take action by addressing negative beliefs head-on. Of the seven percent of Americans who fear AI is feeding the surveillance machine, most are in finance, marketing or healthcare. So businesses looking to sell into those fields should emphasize privacy in their marketing. Americans view robotics and machine learning as the least sophisticated iterations of AI (while natural language processing is respected the most), which is something companies looking to prove their technical acumen should emphasize their NLP capabilities to establish trust.


Sixty-nine percent of the Narrative Impact around artificial intelligence is positive: it’s innovative, it’s exciting, it’s transforming business as we know it. AI-using companies can and should be able to participate in that shared glow. With Narrative Analytics, they can do so without losing sight of the other belief systems at play which make up 21 percent of the Narrative Landscape, so that if and when those issues arise, they are ready to address them.