Charming Chatbots “Hey, how’s it going? Can you believe this change in the weather?” I am told by non-spectrum individuals that these innocuous, content-free sentences are called “small talk.” Apparently, these little bits of silence-filler serve a useful purpose in human interactions. Greetings, wry observations, self-deprecating asides – this is what makes us like and trust other people. Who knew? Well, Justine Cassell, director of human-computer interaction at Carnegie Mellon University knows. She is a linguist who has dedicated her life to studying how humans interact and applying that to machine interfaces. Her latest effort, a charming bot named Sara is capable of dexterously making small talk, picking up on emotional cues from her interlocutor and responding to those cues in appropriate ways. As Apple, Google and Amazon continue to explore ways to make their bots more effective and engaging, Dr. Cassell’s work will become central to the chatbot revolution. Application to Marketing: Preliminary research seems to indicate that the chatbot Sara is more effective at getting users to click on a link than an ordinary chatbot. Charm really does matter in sales and marketing. The digital revolution in marketing has unfortunately been paired with cutbacks in customer support and customer service for many companies. There’s no reason these two trends should be connected (just look at Zappos.) Chatbots could offer customers the kind of high-touch interaction that many of them crave. But it’s still not a perfect solution. There will always be a subset of the population that stands in line at the bank to talk to a teller. (Hi, Dad!) But chatbot technology does allow for the possibility of true one-to-one marketing. Next Steps: Siri is getting better. Alexa is pretty amazing. Spend some time talking to a computer today. Read More All rise! University researchers have constructed a machine learning algorithm that is capable of predicting with 79% accuracy the outcome of human rights cases from the European Court of Human Rights. How perfectly horrifying. The researchers are careful to point out that this technology could not (presently) replace a human judge, but it does show interesting insights about how judges make their decisions. The judges appear to base their decisions primarily upon non-legal facts, rather than specific legal arguments. In legal theory, this is called being a realist rather than a formalist. Lovely. The problem is that the average person imagines algorithms are capable of providing objective results. In fact, most algorithms are ridiculously easy to manipulate by altering the inputs. An AI judge could be manipulated by any lawyer who understood enough to repeat certain phrases or formulations. Studies like this encourage irresponsible use of technology because they imply that technology can easily replace judges or doctors or insurance adjusters or circus clowns. Application to Marketing: Marketing thrives on a hype cycle. One of the ways marketers defend their budgets in large corporations is by convincing the CEO that some new technology or media or agency or CRM program will help to grow the business. That hype cycle has now turned to what is being called AI. (I prefer machine learning.) Machine learning is incredible technology. But for marketers it is not a material improvement over the previous generation of hard coded algorithms. Marketing has a data problem, not a computation problem. Our data is, to be blunt, kind of crappy. And when you feed crappy data into an advanced machine learning system, it produces crappy outputs. There are exceptions. Google’s data is pretty good. But using machine learning to segment your Marketo data is probably a waste of money. Next Steps: Don’t believe the hype. (But do use the hype strategically in budget meetings.) Read More So what is machine learning? Fabulous question, I’m glad you asked. Machine learning is the application of neural networks to deep learning, aided by reinforcement learning. If you remain unsatisfied by that deeply-unsatisfying answer, I would suggest you click on the Techcrunch article below: “WTF is Machine Learning?” Application to Marketing: Mostly, none. Although Google is using machine learning in some of their ad products, you probably don’t need machine learning for your limited data set. But your boss or client is probably going to ask if she can cut next year’s marketing budget, so you might as well click on the Techcrunch article below: “WTF is Machine Learning?” Next Steps: You know what you should do? Click on the article below from Techcrunch: “WTF is Machine Learning?” Read More Beyond the screen Nine years into the modern technology era, digital experiences remain stubbornly separate from the physical world. Smartphones have allowed us to experience the digital world beyond the desktop. But these wallet-sized interfaces allow, at best, an experience that parallels the physical world. Technologists have taken a digital-first approach to bridging this separation between the digital and the real. Alternative reality is an attempt to privilege digital experience in a layered approach to mixed digital/real experiences. It’s better than a backlit screen, but it’s still a fake. However, in the worlds of art and architecture, people have started to learn how to use technology to power real world experiences. Physical objects manifest the output of algorithms in a way that allows for a true mixture of physical and digital. In the link below, a sculpture built of Tyvek origami forms is powered by software to create beautiful motion reminiscent of a flock of swallows. It’s a pity that more technology companies don’t appreciate the power of physical experiences powered by digital logic. Application to Marketing: Combining the physical and digital world feels like magic. We expect digital experiences on a screen and we become desensitized to their technical wizardry because we are aware how the trick is done. When you manipulate the physical world with digital logic, you re-awaken the wonder and the awe of the observer. When AR hits the market, it will be celebrated for bringing the digital and the physical world closer together. Unfortunately, I think it has misplaced priorities. A marketer who wants to create wonder and delight would do better to use the power of software to manifest physical experiences. Next steps: Watch the attached video and try to envision how physical and digital realities can be mixed. Read More