It is important for me to think these things through, even -or especially- when said during a casual small talk at the water cooler. Some day I was having coffee with a colleague and he said:
…Claudia, believe me! All mobile apps could be replaced by a Chatbot. Not because the technology is fancy, but because it is actually a better user experience.
If I would give such a generalized statement during a casual conversation, it is typically followed by this reaction of the other party: I bet you will just nod your head and put on a light smile to signal understanding or agreement. During the time frame of 300ms in which your senses collect, filter and interpret all the information around you just to keep you alive and your subconscious forms the utterance to answer (in order to keep the conversation going and appear like a human being), you won’t come up with the cognitive capacity to think this through. Not to say that anyone would be too stupid to think things through. I state that this needs time and you shouldn’t propagate your generalized statements lightly. After all, deep thinking costs us time and energy and it is not bad to trust your conversational partner on his or her expertise. Still – I have some thoughts on this: I worked in media psychological research for almost 6 years, but I immersed myself just for the past two years into Chatbots and Mobile Apps. For me, both of them are rather a cute couple than rivals. The technology of conversational interfaces is astonishing and a natural language input is the most intuitive way of human-computer interaction. The vast increase of messenger apps shows that computer mediated communication is accepted faster, when it offers the same elements like a face to face interaction. Although real and virtual conversations differ, we see similar patterns like turn-taking, nonverbal cues, emoticons or even the use of video chats which make it easier and more comfortable to communicate.
Plain text can be sufficient but is not the most important thing about communication. Social interaction is extremely important and language understanding is burned into our DNA because of the evolutionary advantages it gave this species. As a child, we pick up our mother tongue by trial and error, observation and daily training, because it is crucial to our survival. After all human language serves so many purposes, we use it to learn from each other by telling stories, we use it to communicate how we feel or to comfort others when they’re in sorrow. When we are little kids, we talk to our toys as if they could hear us and if we have to give a presentation it makes some of us nervous to speak in front of others. Lets think about it from the other perspective: A computer program is a set of rules for a certain task. These tasks can be as complex as artificial intelligence, which is ought to perform certain tasks which, until now, required human intelligence. To program a computer, you need a language that the computer understands such as Java or C ++. To use said program, you need an interface which communicates between humans and computers, e .g. mouse and keyboard. Also, the computer gives us signals that are translated in our language, such as error messages, notification tones and other style elements that communicate the ongoing processes. Therefore, one could argue that the human language is translated through the interface into computer language and vice versa. But every translation leads to errors or even loss of data. If the computer is not programmed to understand or signal something specific, it won’t. The interaction between humans and computers is yet not perfectly designed. It would be, if we both speak the same language.
The idea of talking machines is thought to be as old as 70 years and based on the philosophy of Alan Turings’ Imitation Game.
Turing announced this famous Game, also today famously known as the Turing Test, in the year 1950 in his paper ‘Computer Machinery and Intelligence’. Since then, digital telecommunication technology, e.g. messenger applications are a reoccurring subject in the research of computer-mediated communication. When creating a software which mimics a natural conversation using Chatbot Technology, the computermediated conversations between human agents should be analyzed and used a model. In order to make a Chatbot , we have to understand all different aspects of a natural language itself. A natural language has rules of syntax, grammar, vocabulary but they are not as strict that we could easily translate them into a language which is understood by a computer. Difficulties for the Chatbot development to make it seem intelligent are e .g. contextawareness, detection of sarcasm and irony and using subtle cues or detect emotions. Basically, all the things that humans are also not perfect to do and which leads to misunderstandings in communication. Because a natural language is as complex and has different varieties, dialects, infinite combinations and is changing over time, research in the field of NLP and NLU is found as a recent topic in a range of different domains since all those years. Coming back to the premise that Chatbots will save the world and our Mobile Device will just be a Companion to talk to without the known interface etc pp. In 2018, 52% of all worldwide website traffic was generated through mobile devices , which is 51% more than 10 years earlier. If we see that the natural language interaction comes easy and solves many usability problems, it seems legit to assume that the Mobile Trend will be overcome by the Chatbot Hype. That’s like saying we don’t need desktops because Mobile devices come in handier. For a lot of use cases yes, but if you require certain functionality or a program so complex it is not imaginable to use it on a phone than your hypothesis is busted. The same goes for Chatbots in mobile applications. Sure, there are many examples of how a Chatbot makes your life easier: Booking a concert ticket, writing an e-mail, letting your electricity provider know that your address details changed and so on. Automatically filling in your personal data or selecting search filters can be tedious and would be much easier handled if you just tell your phone what you want, like speaking to a real person. Service and support bots are common because tasks are easily standardized and customized to collect and use individual customer data. That way the experience is even more engaging than a static form, but still cheaper than a human agent. The advantages of a Chatbot for business and customers are clear. There were almost 2 million apps publicly available in 2018 and a lot of them cannot be replaced by a chatbot. Just take an app like candy crush, which has been downloaded over 2.73 billion times. The gamification is simple and fulfills many needs like escapism, hedonism, entertainment, excitement, mood management, pass time motives and choose something else it does for you (I’ve seen you playing it in U-Bahn, don’t lie to me). A chatbot couldn’t add value to the game itself. Which is the same for artistic and creative media usage. A 3d model can hardly be sculptured with a Mobile Phone and photo editing would be kinda weird using a Chatbot. Sometimes these creative processes need happy little accidents or human error and wittiness to end in something beautiful. If you don’t see the gaming industry or arts and culture as something which should be considered important for the technological strategies in our society, well let’s think of an example for our little Platon-fans and defenders of eudaimonia… hmmm… Spending a few moments I came to the conclusion that we wouldn’t have phones at all. But this would go too far and maybe the part-time philosophers in this audience of 5 who are still reading can help me out here with some crowd intelligence. I hope you could follow me so far and see that you wouldn’t choose to do the controlling in your company on a mobile phone and you wouldn’t like to look through your instagram feed by giving instructions to a Chatbot. Some Apps can be replaced by a Chatbot and some can’t, but it is important to see the chances of combination . A Chatbot can add so much if integrated well. Here are just the first three things which popped into my mind ten minutes into brainstorming…
Mobile Companions
To support existing feature between apps. Like Siri , who takes me with a simple customized shortcuts into the third level of my flight booking app or checks the next destination in my calendar and starts navigation. That doesn’t mean I don’t like to manually scroll through the overview of my calendar or to contribute photos as a google local guide. These apps offer more for me than just being little slaves for my daily tasks.
Virtual Guide
When first opening up the App of my car insurance company, a friendly assistant greeted me and guided me through the features of the app. It has something refreshing to not click through a tutorial but to be able to ask questions about the functionalities that matter to me the most. It doesn’t mean that you boost your app which has bad usability, but by default the human perception is limited. Someone explaining the funtionalities of a program to you, neither means you are stupid nor the usibility is bad. But the UX colleagues might argue with me on this point.
Support Chat
In any messenger, you can mimic a conversation with a Chatbot for many purposes. If there’s a fancy IoT device which you want to control by chatting with it, or if you want to get the most recent news on your favorite soccer team you could have a WhatsApp contact for this. You wouldn’t need two additional apps for these purposes, but it is still just the one functionality.
Conclusion
In conclusion Mobile Apps won’t vanish from the face of the earth just because Chatbots are getting smarter. Just because I can chat with my calculator doesn’t mean I would want to. Maybe I am too old for this and the next generation will say I am crazy, but that’s the exciting part when talking about innovations. Nobody knows what will be, so let’s see what the future brings.