Since I got the idea of building Brian, he has been keeping me awake at night. Brian is an AI-powered digital assistant that performs tasks high-paced business professionals never liked to do. He is like Alexa communicating via email and not only answering questions, but also providing content, translating and converting documents, and more. In this article, I’m sharing why I personally find applied AI that exciting.
Bringing AI to the People
There are thousands of very smart people working on countless amazing digital solutions for almost everything. The availability of AI building blocks enables creative minds around the world to build unseen applications—if it is in the area of vision, voice, analytics or natural language processing. The majority of people do not know these solutions, don’t trust them or don’t know how to use them. Knowing and making the state-of-the-art (AI) tech available to people in a very simple way is extremely exciting. Before building Brian, I was thinking that it would be just amazing to have a digital companion performing all the tasks that I, as a management consultant, found to be annoying. Thanks to AI, Brian is already able to perform many of these tasks while using natural language. It is a great brain teaser to decide what technology to apply for which task and how to make the user experience as simple and cool as possible.
Creating Virtual Beings
It is perhaps quite inappropriate to express it this way, but when creating virtual beings, one feels little like a god. What should be the purpose of this being? What should be its talents and skills? What about its character expressed in the way it communicates? What would be the user perception of a male, female or neutral gender of the digital assistant? What should be its visual representation? What communication channels should it use? The number of possibilities is just enormous and makes you think a lot. My true north is always: How would I, as a user, would like to have it?
Exciting Interconnections and Novelty
There is probably also a scientific explanation for my excitement preventing me from peacefully resting at night. Designing and building Brian means connecting the creative and analytical hemisphere of my brain. It also requires to connect humans and systems to work together in a new way. The high level of novelty drives my dopamine levels up causing the feeling of happiness. I am an AI addict, I believe. Additionally, building Brian is about connecting the experience of people from backgrounds in consulting, AI, (traditional) programming, journalism, digital marketing, and GDPR—quite a large puzzle with some moving parts to put together.
Tons of Use Cases
After being in management consulting for more than 10 years while focusing on support functions optimization in large corporations, I see tons of use cases for digital assistants in consulting and beyond. Usually, people associate digital assistants with getting answers to their questions. But just like asking a bot to deliver a pizza with salami and mushrooms, you can ask a bot for any digital product (peer comparison, industry report, graphics, …) you can imagine. Or you can trigger the execution of a particular task such as translating complete documents or sanitizing your files before saving them in your KM system. Once I understood the technology in-depth, I started to see hundreds of reasonable use cases—use cases having the potential to significantly save time, money and bring employee satisfaction up as well.
I am amazed by the speed of digital assistants for two reasons. One is how fast you can build one; usually, you can have a first demo within minutes or hours. Extending the skills and making them robust, of course, can take much longer, but you have your proof quite fast. The second reason is the speed of answers—assistants respond within seconds. Brian usually needs E2E two to three minutes to translate a multi-page document from/to over 100 languages. When using alternative channels like Slack and MS Teams we will significantly accelerate. The really strong aspect is that you get your answers 24/7 and what you taught your assistant once, he won’t forget ever again.
With natural language, for the user, it is just like speaking to a real person helping you at work. No ‘menus’, ‘buttons’ … you just ask straight away what you want. You do not need to install anything on your device, you do not need to remember your login information. And there is no nervous IT department managing software updates or thinking about the integration with other systems. On the technical side, the assistant software architecture (at least in Brian’s case) is very robust and flexible at the same time – allowing us to add new skills in a relatively easy way.
Any Digital Communication Channel
Once the assistant is created, you can connect him to any digital channel you imagine—Slack, MS Teams, Facebook Messenger, WhatsApp, SMS or your website… anything is possible. You can also call the assistant via phone and access the digital services and data services via an analog device! And, of course, there is email—everyone has one, everyone uses one. Communicating with an assistant capable of understanding natural language (like Brian) via email is a very exciting experience. Of course, writing the first email to a digital person feels a little odd, but afterward, it becomes very natural. How to shape the user experience via various channels… this keeps me awake at night as well.
It Is a Challenge
I love challenges and building Brian is definitely one. One fights, of course, with many problems, but you also visualize the situation, when you master the challenge. I would say it is quite comparable to an Ironman Triathlon (and I do speak from experience). You need a long breath, it requires to master multiple disciplines and you have to work very hard to make it. The really tricky challenge is to manage the human expectations. Some users are very skeptical e.g., believing it is all fake and that a real person actually answers the questions and performs the tasks. But there are also some people believing in an all-knowing intelligence being able to answer every question. There is no general AI as of now and it will take some time until we get there. But the narrow AI applications will become broader and broader with an increasing level of learning autonomy. Still, I believe the bigger challenge will be rather humans than technology – Humans creating AI and humans using AI. We need to learn, how to work with digital beings.
Huge Potential Looking Forward
And finally, what keeps me awake at night is the huge potential of Brian and other digital assistants. Brian already amazes his users. Considering the advances in AI, I believe we are just at the beginning. Feel free to ask for a free trial and build your own opinion! Hopefully, he will not make you awake at night as well:)