“We are facing the biggest revolution in the history of mankind which will reevaluate everything we have known,” claims Maciej Stanusch, CEO of Stanusch Technologies SA in the interview conducted by Tomasz Jurczak
Tomasz Jurczak: What is your vision?
Maciej Stanusch*: For us, the most important thing has always been thinking beyond “here and now”. 13 years ago, when we started building Stanusch Technologies, a company creating virtual assistants, almost nobody heard of chatbots, and artificial intelligence (AI) was discussed mainly at universities. But we strongly believed that AI would become widely available. And that happened two years ago. But we are one step ahead and we are already thinking about what will happen in five years’ time. Of course, today we are focused on creating chatbots and voicebots, but our research and development department is already working on solutions which will start to reach wider audiences in a few years.
Now everyone is excited over machine learning (ML) and neural networks. This technology has been known for decades, but only now is it gaining in importance because only now do we have enough data and processing power to develop it. ML is perfect for pattern detection, e.g. image and speech recognition, cybercrime detection or predicting customer behavior. You can also say it allows to imitate how neurons work in the human brain. That can be surely called artificial intelligence.
But ML also has its limitations: it is not so impressive when it comes to understanding a natural language. We still have problems with that, because there is not enough good training data, so rule systems are still the most effective ones.
How do you cope with that?
We follow a combined approach. ML is great for speech recognition and analytics, but to understand a language, in commercial projects, we use a rule-based approach. We used that model to develop EVA, our most recent chatbot, which is partially based on the neural network and partially on the rules.
You have called EVA “the most human chatbot”. What makes her different from other chatbots?
EVA doesn’t have any specific content-related knowledge. You can’t talk with her about a bank offer, for example. She is a simulation of personality. Talking to her is like making small talk with a new friend, having a chitchat on your first date. Who are you? What do you do? What don’t you like? What do you think about this or that topic? EVA can also exchange views, comment on interlocutor’s statements, ask contextual questions. It is a research project. We observe how users behave, how long they talk to Eva, what topics they are discussing and how they do it. We want the conversation to be as natural as possible.
What can such a chatbot be used for?
There are many applications. A significant number of users of digital assistants, such as Amazon Alexa or Microsoft Cortana, include people over 60 who have no one to talk to. Our solution could be an ideal complement to such systems.
For the first time ever, humanity can overcome the three greatest plagues that have laid waste since the dawn of history: hunger, wars and diseases.
But let us also think about the entertainment market: video games have their own characters. Today, interaction with them is very limited, but just imagine we could talk to such a character! We believe that within a few years all the devices we will talk to, as well as voicebots and chatbots, will be fitted with a perfect imitation of human personality.
Aren’t you afraid that some “central controller” will block such solutions?
There may be such a temptation at a local level. It is true that some countries have enforced such regulations. But this only means that those countries will see a slower rate of development and that after some time they will fall behind their competitors. Let’s take a look at what’s happening now between China and the USA or Europe.
Because of the fact that Europe has implemented GDPR and that there are no similar regulations in China, China has gained the upper hand in developing its technology. With more data (which today is our most valuable asset), the technology in China will develop faster, whereas in our region, due to regulations, it will progress more slowly. And that means only one thing: over time, our standard of living will decline.
But personal data protection also raises the standard of living…
If people find some technological solutions favorable or beneficial, they will want to use them.
Do you see any threat from artificial intelligence or machine learning?
I am definitely not overly excited about machine learning, as I find it unpredictable. It is very dangerous if seen from the perspective of commercial solutions. If we want to maintain “power” over ML, we cannot abandon the rule-based approach. We know that ML optimizes algorithms and decisions which are sometimes at odds with ethics. There are known cases where systems using ML discriminated against certain groups of people or where artificial intelligence based on ML became homophobic, for example. Perhaps some legal regulations will be implemented in this area too. ML cannot learn everything on its own; we have to provide it with certain guidelines.
Are we facing a revolution?
Societies will change. And it will be a total change. We are facing the biggest revolution in the history of mankind which will reevaluate everything we have known. For the first time ever, humanity can overcome the three greatest plagues that have laid waste since the dawn of history: hunger, wars and diseases.
Maciej Stanusch
What is more, with advances in genetics we will live significantly longer and, who knows, maybe even forever! In my opinion the question is not “Is it possible?”, but “When will it happen?”.
And one more thing: we will have a lot of free time because most of our work will be done by robots.
Where do you see your company in this world?
Our solutions are created mainly for financial institutions and telecommunication companies. However, we are introducing services that do not require “entry” costs. Customers are charged only for outgoing and incoming calls and do not incur any license, implementation or maintenance fees. Such services can be used even by small entities that wish to automate e.g. debt collection, marketing research or after-sales service processes.
How does that work?
Your calls to the call center will not be answered by a real person but by a bot. A bot can also give you a call. All else will look like a standard conversation with a human. The call is switched to a human only if the bot is unable to solve a customer’s problem.
Various institutions estimate that the exponential development of technology will result in 50-percent unemployment in Poland in 15 years.
Today, systems are capable of solving up to 80 percent of callers’ problems. Good solutions can be even negotiated with.
What is the future of artificial intelligence in this area in five years?
Today, the Polish labor market belongs to the employees rather than to employers. As employees are hard to find, many employers decide to engage “digital employees”, which is the only way to overcome labor shortages. This is very good news for the AI industry. I am sure the sector will grow even more rapidly.
Are we lagging behind the West in developing such technologies?
The problem in Poland consists in the mentality of managers, especially local managers of international corporations. They are rather reluctant to take risks and, especially, to implement innovative solutions. “Right now my call center is working fine, so why would I introduce an innovation that may not work and set me up for failure?” they think. They are focused on being comfortable with the job they have. It is no coincidence that Poland is at the very bottom of all EU innovation rankings.
What will the ideal virtual assistant be like?
Indistinguishable from a human, fitted with personality simulation. You will be able to talk to it just like you are talking to me right now. It will know everything about me, it will have access to my data, it will be capable of answering on my behalf to all the questions you are asking today. And your virtual assistant will conduct this interview on your behalf.
And what will happen in 20, 50 and 100 years?
Various institutions estimate that the exponential development of technology will result in 50-percent unemployment in Poland in 15 years. The good news is that during that time GDP is expected to increase by about 75 percent to one trillion dollars. This will allow to reduce the working time to e.g. four hours a day and to suppress the unemployment; alternatively the government will pay a basic income to the citizens.
Technology will change companies, entire industries and the world. Today you’re holding a smartphone that has more computing power than all the supercomputers NASA had in 1969 when it sent a man to the Moon. I think that next year you will be able to buy a smartphone with a memory of one terabyte and it won’t surprise anyone. And that is just the beginning. According to Moore’s law, in the early 2030s an ordinary smartphone will have a memory of 128TB. If the contents of such memory were printed on A4 sheets of paper stacked on top of each other, a giant tower, 7,000 kilometers in height, would be constructed. Such a smartphone will be able to do almost every intellectual task men can do today.
Are we going to carry our digital doppelgangers in our pocket?
Yes, assistants capable of doing different tasks for us. And this is where the theory of rising unemployment comes from.
And what then?
We will modify our DNA to overcome genetic diseases and cancer. We will print our organs and organ transplantation will become a non-issue. There will be plants able to shine at night, which will solve the problem of street lighting, for example. We will grow artificial meat, so we can stop killing animals. We will also print houses and clothes. Genetically modified bacteria will clean the environment we have polluted.
The best part is that all the technologies I have just mentioned already exist. Each of them is being commercialized. We’re not in a lab anymore.
And then?
Then it will be time for singularity. But that is another story.
*Maciej Stanusch – an AI industry expert, founder and president of the management board of Stanusch Technologies SA, a company specializing in projects using natural language processing (NLP). ST SA has implemented over 140 chatbots and voicebots, e.g. for the banking industry.
Maciej Stanusch has been included twice on the list of the most creative entrepreneurs in Poland. In 2007 he received the Silban Association Award for an innovative business idea and in 2010 he obtained the 2010 Regional Innovation Leader Award. He was nominated twice for the Ernst & Young Entrepreneur of the Year award and for the 2014 Wprost Innovators award.
Przeczytaj polską wersję tego tekstu TUTAJ