The Staccato Conundrum - Can AI transform us into master storytellers and why should we care? (Part 1)
Shajee Rafi
Aviation AI Evangelist | Airlines | MRO | Operations | Digital Transformation | Product Leadership | Innovation | Data Landscape | Business Architecture | Strategy
"The single biggest problem in communication is the illusion that it has taken place." George Bernard Shaw
In 1990, Elizabeth Newton, a post-grad psychology student at Stanford University, conducted a fascinating experiment, with what were, quite thought-provoking outcomes. In the style of academia, it was innocuously called the Tappers & Listeners experiment, a title which didn’t quite shout out aloud, what we now know it to be, an exposé of the business executives’ love affair with vague strategy statements, a phenomenon they called, “the curse of knowledge.” Hitherto obscure, at least to my limited knowledge, I first heard of it while reading the authors Chip and Dan Heath’s 2006 HBR article about this savoire faire curse.
For her PhD dissertation, Newton invited some of her fellow students to participate in the study. Each of them was randomly placed in one of two categories: “The Tappers” or “The Listeners”. The tappers were given a list of twenty-five popular and widely known tunes, such as "Happy Birthday to you" and "Jingle Bells". They had to tap out the tune with their fingers against a wooden desk, and the listeners were tasked with naming that song. This apparently was no mean feat, the difficulty of the task at hand was corroborated by participants’ dismal scores at the end of the experiment. Of the hundred and twenty times a tune was tapped, the listener could guess the tune correctly only thrice, a rather paltry success rate of only 2.5%.
Although a striking result, but that wasn’t the punchline. Before the tappers began their staccato renditions, Newton asked them to predict the listener’s success rate. The tappers, in an act of supreme confidence in their percussion talents, on average predicted a 50% certainty, that they would be able to get the listeners to guess the tune correctly.
So while the participants were quite confident that they would be able to get the listeners to guess correctly one out of two times, the reality was that listeners could guess the tune correctly only once in forty tries. Mere overconfidence or is there a deeper reason for such unexpectedly poor performance?
Like most things in life, there indeed is a reason. As the tapper percusses out the tune, she can hear the song playing in her head. Her fingers seem to be tapping the tune in perfect harmony with the record that’s playing in her mind. And she, for the life of her, can't fathom why the listener can’t hear, what to her is crystal clear!
Let’s look at this performance from the listener’s point of view, shall we? He, for starters, doesn't have the tune playing in his head, and without that accompaniment, he has no idea of what's transpiring. He tries as hard as he can to decipher the data deluge, which to him sounds like some weird Morse-code-like tapping, trying to match the rhythm to the thousands of songs that he knows, and failing 117 out of 120 times. This results in utter confusion, wrong answers and yes, frustration. The big question is, who is responsible for this dismal performance, the tapper or the listener?
As leaders, managers, teachers, husbands, wives, children and parents, we fall prey to the tapper's trap daily, actually for most mere mortals, many times a day! We communicate intent that seems very clear to us but our intended audience may have no idea about what we want them to do. Has it ever happened to you that you called one of your direct reports to get some urgent and important work done, and when they got back the next day - having kept their nose to the grindstone late into the night to finish that uber important task - you were disappointed? The output left a lot to be desired, they hadn't quite accomplished what you needed. You probably felt paroxysms of frustration at that time that your team member 'didn't quite get it' and your slam dunk would have to wait a tad longer. Another valuable insight, had Newton included it in her study, would surely have been, that tapping harder or tapping repeatedly, doesn’t make it any easier for the listener, to guess correctly! Another approach needs to be employed if we are to drastically improve the success rate.
Next time you take your communication skills for a spin, do remember that the problem is with the tapper - not the listener. Just because you have the symphony playing in your head because of the knowledge that you have accumulated through years of hard labour, does not mean that the listener has the benefit of similar familiarity. When it comes to communication, we all tend to think we’re pretty good at it. Truth is, even those of us who really are good aren’t nearly as good as we think we are. This overestimation of our ability to communicate is magnified when interacting with people we know well.
Academics at the University of Chicago, Booth School of Business, put this theory under the microscope and what they discovered is astounding. Kenneth Savitsky, professor of psychology at Williams College in Williamstown, Massachusetts, devised an experiment resembling a parlour game to study the phenomenon. Two sets of couples sat in chairs with their backs to each other and tried to discern the meaning of each other’s ambiguous phrases. In all, 24 married couples were part of the study. The researchers used common everyday conversational phrases to see if the married couples were better at understanding phrases from their partners, as compared to from people they did not know. The astounding part of it all, which defies at least what I thought was common sense, the researchers discovered that people who knew each other well understood each other no better than people who’d just met! To add insult to injury, when dealing with people they knew well, participants often overestimated their ability to communicate. Nicholas Eply, the co-author of the study, said something quite insightful,
“Our problem in communicating with friends is that we have an illusion of insight, ….. Getting close to someone appears to create the illusion of understanding more than actual understanding.”
When communicating with people or groups, we think that we know well, we make conjectures about what they understand—guesses that we don’t usually make with strangers. This predilection to oversell to your own self, your ability to effectively communicate and how well we’re understood by those on the other end, is so widespread that the psychologist community even has a name for it: closeness-communication bias.
“The understanding, ‘What I know is different from what you know’ is essential for effective communication,” as per the Booth study lead Savitsky, “but that insight can be elusive. Some may indeed be on the same wavelength, but maybe not as much as they think. You get rushed and preoccupied, and you stop taking the perspective of the other person.”
Most people agree that communication is an important pillar of leadership but there is less of an agreement about the monetary value of storytelling. In 2009, two men by the names of Rob Walker and Joshua Glenn ran a social experiment to prove exactly this point. They purchased cheap knick-knacks and asked some of the era’s most gifted creative writers to “invent” stories about the objects. They then proceeded to use the stories and the objects together in eBay auctions to see whether the invented tales enhanced the monetary value of the object. The hypothesis that they were trying to prove was that stories are such powerful drivers of emotional value that their effect on a given object’s subjective value can actually be measured objectively. Simply put, does a compelling narrative transform insignificant objects into very significant ones?
One of the items, this simple globe paperweight basically came with its very own soap opera attached. It was accompanied by a handwritten letter to a friend with the tale of a broken marriage, world travel, remorse, solitude and at times, questions of amour-propre. And in just this one single object, a heartbreaking reminder of multiple romantic let-downs, the paper weight had a great story to tell. What is not quite evident from the words is that this trinket that was bought for a paltry $1.49, because of its artful and emotional albeit artificial augmentation, went under the gavel at a whopping 13,155% premium with an eventual sale price of $197.50. In case you were wondering, this wasn’t just a fluke outlier, similar premium outcomes repeated for every one of the 199 other items too. Two hundred trinkets were purchased at local thrift shops for a total spend of $129. These 200 items sold for a total of $3,613 at a staggeringly tidy gain of 2,700%. If this is not a significant enough monetary upside of good storytelling, then I don’t know what is!
Thus far we have seen, that contrary to popular credence, people might not understand your intent as well as you think. The closer they are to you, the higher the chances that assumptions on your part would reduce the efficacy of the communication that ensues and finally we most certainly have observed the significant monetary upside of telling a good tale. Considering my ongoing infatuation with all things AI, I couldn’t help asking myself the question, “Can we utilise AI to level the playing field, to remove ambiguity and to turn every communication into a work of a master storyteller.” This isn’t a new idea, applications like DialogFlow (formerly known as Api.ai), that was recently acquired by Google, have been trying to instil this intelligence about semantic nuances into their software for years but I haven’t come across any use cases where Generative AI and Natural Language Processing were used as tools for impactful communication. If I am not aware of it, that does not mean it isn't happening already or couldn't, as we'll discuss, the tools of the trade are being sharpened as we speak.
The National Health Service in the UK has taken to using a chatbot to do tedious corporate team-building. They believe that if you and your colleagues are not as good at communicating with each other as you'd like, then perhaps CoachBot could help. Developed by the London-based HR company Saberr, it asks simple questions about workplace dynamics and provides the team with reports. CoachBot helps measure engagement and supposedly drives action utilising small nudges to remind the team to act on what they've learnt. A unit within the UK’s National Health Service is trialling it, as are ten other companies, including Unilever and Logitech. Imagine a world where all your team members just “get” what you are saying. Your words inspire, motivate, energise and engage the ones that you communicate with, every time, all the time. If communication is that important, then a million-dollar question, quite literally is, what makes an excellent communicator?
If you have ever tried to in the past or intend to anytime in the future, communicate with impact and make a lasting difference, then you might want to take heed. Recent scientific research is beginning to shed a lot more light on how stories change our attitudes, beliefs, and behaviours, on how a cocktail of chemicals, inspired by words, set the stage for a neural ballet in which stories radically alter the activity of people’s brains.
Paul J. Zak is the founding director of the Centre for Neuroeconomics Studies and a professor of economics, psychology, and management at Claremont Graduate University, his lab discovered that a neurochemical called Oxytocin is a fundamental component of the “it’s safe to approach others” signal in the brain. Oxytocin is produced when we are trusted or shown kindness, and it has an effect of stimulating the act of cooperation with others. It does this by enhancing the sense of empathy in our neural circuitry, which is our ability to experience others’ emotions. More recently his lab has been trying to “hack” the Oxytocin system to motivate people to engage in cooperative behaviours. To do this, they experimented with tales shot on video, rather than face-to-face interactions, would cause the brain to make Oxytocin. By taking blood samples before and after the videos, they found that video stories with a character focus do reliably cause Oxytocin synthesis. They further discovered that the quantity of Oxytocin released in the brain correlated with how much people were willing to help others, for example, donating money to a charity associated with the narrative.
As they delved deeper into the topic in further studies they discovered that, in order to create a desire to help others, a story must first grab attention by developing tension during the video. If the story is able to espouse tension successfully then it is likely that viewers will identify with the emotions of the characters in it, and even after it has ended, continue copycatting the feelings and behaviours of those characters. This explains the feeling of power you have after you leave the cinema having watched James Bond save the world, and your motivation to work out, while humming the tune of the eye of the tiger, after watching Rocky Balboa fight!
These discoveries about the previously little understood neurochemical dance that takes place in our heads, inspired by masterful storytelling, are very relevant to business settings too. Zak’s experiments show that character-driven narratives peppered with carefully crafted emotional content result in a better understanding of the key points a speaker wishes to make. They even enable a linking of short-term memory to its long-term sibling, leading to a better recall of these points weeks later. A message or a presentation should be attached at the proverbial hip with an absorbing, human-scale story. Why should your customers or your team care about the project you are about to embark upon? How does it change the world or improve lives? What would be the emotions that people would feel when the project is complete? Oxytocin indeed is a potent aid that can be used with more precision and impact than most people realise and in terms of making an impact and facilitating change, an engaging story blows the standard PowerPoint presentation out of the water every time!
The utility of this approach is not limited only to outward communication to your customers, research from Zak’s team has also shown that stories are useful within organisations as well. People are considerably more motivated by their organisation's higher purpose, how it improves people’s lives rather than by the mechanics of how it sells goods and services. Higher purpose is effectively communicated through stories – for example, by describing the sorry situations of actual, named customers and how their problems were solved by the efforts of your company, department or team. If you can make your people empathise with the pain of the customer’s ordeal, they will also feel the sweet pleasure of its resolution. If some heroics went in to reducing suffering or struggle or producing joy, even better. These are the stories that lead to creation and release of attention-grabbing tension and fuel Oxytocin synthesis and unlock all its associated benefits.
Joseph John Campbell was an American mythologist who worked in comparative mythology and comparative religion and his works cover many facets of the human experience. His famous book The Hero with a Thousand Faces was published in 1949, in which he discusses his theories of the journey of the classic hero found in mythological tales all across the globe. Campbell’s work tells us that the recipe that leads to memorable stories that have stood the test of time, is relatively straightforward, a character struggles and eventually finds hidden abilities and uses these to persevere and eventually triumph in the face of great adversity. On the other hand, as we discussed earlier, Paul Zak’s team’s research shows that thanks to Oxytocin, be it business or life, if you want to motivate, persuade, engage or make a memorable impact, a story of human struggle and eventual victory, wins people’s hearts by first attracting their neurochemicals.
Thus far we have talked about how assumptions and familiarity, if not deliberately handled, breed inferior communication rather than helping. We also discussed how storytelling has a significant monetary upside and if the intent is to engage, persuade and facilitate change then there indeed is a recipe, which if exercised appropriately, can lead to uncannily impactful communication. But we have known this for a while and many people in the corporate world, show-business and politics, have employed this with great success, so where does AI fit into all this?
Part 2…
Part 2 of this article will be published in a few days. In part 2, I will cover where does AI fit into this rant about storytelling and why it’s much easier to build an AI system that can detect a cat in a picture than it is to determine what is impactful speech.
Principal Solutions Architect at Emirates
1 年Interesting rant Shajee. I believe the generative AI's evolution, will soon be primed for such an application of this technology. If not already around, there will soon be products where you should be able to choose a theme (like a movie plot or theme) and generate a compelling story, though delivering it with authenticity and the necessary body language is altogether a different challenge.?Yes generative AI is definitely a good assistive tech, to enhance the impact of a story... waiting eagerly for the part 2...?
Product Leadership | Digital Transformation & Technology | Chief Product Officer (CPO) | Design Thinking | Product Manager ? Marketplaces ? E-commerce ? SaaS-based Solutions
5 年A very well written article Shajee. I am glad I came across it. Completely agree with the effects of storytelling on humans. In the book 'Sapiens', yuval mentions that the only difference between us and chimpanzees is not genes, but a mysterious glue that unites human beings in forms of stories. Strangers cooperate effectively when binded by stories such as nations, money, rights - which actually doesnt exist outside of commonly held beliefs. Likewise in the book 'hit makers' the author has given numerous examples from instances ranging from star wars to mordern design, all popular trends in society mature from some finite stories which have stood the test of time. Really curious for part 2 - can AI crack what makes good stories that can span across societies and time. And how those stories effect individuals individually.
Human Resources
5 年good read! lot of Oxytocin & waiting for part 2
Management, Aviation Security, & Training Consultant
6 年Finally got around to reading this Shajee. I'm not entirely surprised by the tappers & listeners results. Is it a fair assumption that humming would be much more effective? Yes, a good story blows power point out of the water every time, I agree. I'd say not many power point advocates are comfortable telling stories though! I'm very skeptical of us employing more AI to help us communicate better. It may very well help to get the message across, but then doesn't it de-humanize us further if we don't have to think about how other humans perceive our message? Another reason for us to reduce human interaction, retreat to the safety of our tech, shut down our brains and not worry about other human beings. We won't need to worry about having communications skills because AI will take care of that for us? One day perhaps we'll stop thinking altogether...!
Head of Learning and Organizational Development at Shapoorji Pallonji International
6 年Interesting read Shajee. Talking of Chatbots like CoachBot and Lark, they are using elements of AI to enhance UX. Perhaps the little ones are learning now, and as they consume more and more data from a variety of sources they could potentially 'act' like a user.? Couple of thoughts before your Part 2 comes up. :) 1. It would be interesting to see how the CoachBot developers can create algorithms that would address the variety of 'Cognitive Biases' in humans, which impact their behaviour and decision making.? 2. Imagine its application in the judicial system, where jury member outcomes are predicted by a bot, based on patterns in interest, vulnerability, degree of influence from different people, personal relationships, biases in racism, belief etc.? I read that a few years ago some companies have tried this in the US, especially on risk assessment of a defendant. Remains to be seen how far have these progressed and how have these addressed stereotyping.