Wisdom From the Women Leading The AI industry, With Ericsson’s Elena Fersman
Every new technology brings a threat. While AI is special, since it’s self-evolving, we should still have conditions or boundaries set in stone for its development. These conditions are in the hands of humans. We need to take responsibility and dictate the “rules of the game” that should never be broken by these algorithms. These rules are no different from what we humans have to comply with — laws, regulations, ethical frameworks and social norms.
I had the pleasure of interviewing Elena Fersman, the Head of AI Research at Ericsson. She specializes in industrial transformation, artificial intelligence, digitalization and IoT. At Ericsson, Elena has had various positions ranging from product management to research leadership, where her interests in technologies impacting businesses and people coupled with modeling and analysis of autonomous knowledge-intensive cyber-physical systems have been met. Elena is also a docent and an adjunct professor in Cyber-Physical Systems specialized in Automation at the Royal Institute of Technology in Stockholm. She has co-authored 20+ scientific publications and 30+ patent families, participated in several PhD grading committees and served as reviewer at many international conferences. She received a PhD in computer science at Uppsala University in 2003 and worked as a postdoc at École Normale Supérieure de Cachan in 2004. Her previous speaking experience has included keynotes on AI technologies and smart cities, as well as panels on industry 4.0, IoT and embracing AI ethics.
Can you share with us the ‘backstory” of how you decided to pursue this career path?
Growing up, I always loved mathematics. I went to a high school in Saint Petersburg that specialized in math and physics and loved these subjects. When the time came to make a choice about the direction I wanted to take for higher education, I chose computer science since it felt exciting to see how the basic technologies of math and statistics can make a difference in the world of machines, and how the power of information technology can influence in the world. That passion resulted in a master of science degree from the Peter the Great St. Petersburg Polytechnic University in Russia, and later a PhD in computer science from Uppsala University in Sweden. After that, joining Ericsson Research was an easy choice for me since I wanted to see how things work in industry without losing touch with the academic world.
What lessons can others learn from your story?
I followed my passion for numbers from the start. That gave me a solid background in what I do now at Ericsson, and even though it’s impossible to keep track of every technological development, it’s good to have a solid base and then build on it when needed. Another lesson to learn, is to remember to enjoy the journey and don’t forget about the people you meet on the way. For example, now, 16 years after my PhD defense, my PhD supervisor and I are good friends.
Can you tell our readers about the most interesting projects you are working on now?
We are accelerating the application of AI in all parts of Ericsson; within Managed Services, Networks, Digital Services, IoT and New Businesses. It’s impressive to see the potential and values that can be created for our customers. For example, seeing how novel AI-based features in Radio Access Network can improve end-customer experience, such as efficient secondary carrier prediction. Or preventive maintenance of our equipment in operator networks around the world.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
My parents pretty much decided the beginning of my journey. They are both professors in technical fields and settling for anything less than a PhD was unthinkable for them. Looking back, I am grateful for that because it became a smooth journey for me. It’s like running — if your goal is to complete a marathon, you will most probably do it. If you target a half-marathon, it would be harder to go for another 21k after completing the first 21k.
Another special person in my life is my colleague, co-author, friend, husband and father of my three kids — Paul. We met at Uppsala University 20 years ago, and since then we’ve never stopped discussing research problems. This has always motivated me in finding creative solutions in order to impress him.
Finally, my team at Ericsson, of course! There are so many people to thank and I wish I could list them all. But, I would not have been where I am today without them. They challenge me to create, innovate and experiment. In AI research collaboration is central because the algorithms and architectures that are not put in a relevant context are useless.
What are the 5 things that most excite you about the AI industry? Why?
· Symbiosis of AI and humans. We’ve seen the potential of AI in many areas, and the key for the AI industry is to put humans at the center. I share Garry Kasparov, the chess grandmaster beaten by a computer’s views: “It is all about us finding the right combination; how we can actually apply our very human quality emotions and other things that are uniquely human into this collaboration”.
· Applications of AI for restoring or accelerating human functions. Today, there are already existing technologies for non-intrusively reading bioelectric intentional signals of a human being, amplifying these signals, actuating the robotic part of the solution attached to a human body, and closing the loop by sending the signals of performed movement back to the brain. This procedure restores the weak of unstable intentional brain signals and hence the body function. In other words, thanks to AI, it is now possible to cure patients who suffer for things like spinal injuries.
· Advances in processing of human-generated input such as video, image and speech. Today’s machine learning software can recognize human voices with 95 percent accuracy, which is nearly equivalent to a human’s ability. This leads to new disruptions. Already, now we are moving away from typing to voice recognition software. The next question would be if voice is the optimal media for us to communicate with each other — as quite often I still want to record my thoughts silently. Personally, I think the area of reading and interpreting brain signals is very exciting.
· The interplay of AI applications from different industries. For example, algorithms that are responsible for the optimization of public transportation in cities are heavily dependent on the capabilities of the vehicles and the infrastructure in the cities. Vehicles and city infrastructure would, in their turn, pose requirements on the underlying ICT infrastructure that will adjust as needed by the application to provide the necessary quality of service. In other words, there will be AI algorithms on different application levels, and the better they understand each other and collaborate, the smoother the service will be for the citizens.
· Broadening the AI mind. Recent advances in narrow AI are ground-breaking. Narrow AI is about a well-defined domain description, and well-defined objective function. Similar to human intelligence, the broader our minds are, the more difficult the decisions are. When I was 18, I thought I had answers to everything. Now, in my 40s I am not so sure anymore. Subjects that are broader and less defined in nature are trickier for algorithms to solve, as the state of the space can grow exponentially.
What are the 5 things that concern you about the AI industry? Why?
· Diversity and inclusion are equally important for pieces of data, information and knowledge, as they are for humans. Skewed datasets lead to unfair results and this is something that we need to prevent.
· Symbiosis — not a replacement of human capabilities. As in the example with using AI and Robotics for restoring human beings.
· Lack of transparency. Transparency and the ability to explain the insights and decisions taken by algorithms are of high importance and often difficult to achieve due to the combinatorial nature of algorithms.
· Disbalance in acceptance. Even though we live in a globalized world where technologies and innovations reach across borders, there are many situations where varied levels of acceptance of new technologies are found, both from a societal point of view and from a regulatory perspective. This creates a disbalance in maturity of technology development in different parts of the world.
· Humans should not forget to develop themselves. Just because we have great speech-to-text or auto-correction software readily available, doesn’t mean we should stop learning how to spell or doing mundane activities. We should aim at being a good partner for our AI algorithms.
As you know, there is an ongoing debate between prominent scientists, (personified as a debate between Elon Musk and Mark Zuckerberg,) about whether advanced AI has the future potential to pose a danger to humanity. What is your position about this?
Every new technology brings a threat. While AI is special, since it’s self-evolving, we should still have conditions or boundaries set in stone for its development. These conditions are in the hands of humans. We need to take responsibility and dictate the “rules of the game” that should never be broken by these algorithms. These rules are no different from what we humans have to comply with — laws, regulations, ethical frameworks and social norms.
What can be done to prevent such concerns from materializing? And what can be done to assure the public that there is nothing to be concerned about?
To prevent such concerns, the “rules to live by” that I brought up in my previous answer, should be implemented to AI systems in the language they understand. That is, we need to digitalize these rules, make sure that the frameworks and software architectures are flexible and allow for modifications, and that algorithms can guarantee that they adhere to these frameworks.
How have you used your success to bring goodness to the world? Can you share a story?
One example is what we do at Ericsson within operational safety for humans. Tasks that revolve around maintenance and/or repairing telecom sites including Telecommunication Tower Climbing in the field are difficult, time-consuming and risky. They can also be time-critical if the functioning of the telecom network is disturbed by a failure in the field.
The moment me and my research team heard others say that “these tasks are very difficult to automate” we decided to take on that challenge. Now we have AR/VR-powered assistance for the engineers, that shares recommendations on when to climb and when not to climb and which tools to bring with them. We also provide drones that perform the actual inspection of a site. Robotics for maintenance and repair is an area that we are actively looking into at Ericsson together with our academic partners.
As you know, there are not that many women in your industry. Can you share 3 things that you would you advise to other women in the AI space to thrive?
· Get your hands dirty. You need that for credibility, and to get a base that you will build on. We’re talking about T-shaped people — those who have in in-depth knowledge in one field (vertical bar of T) and a broad knowledge of an application domain (horizontal bar of T). Having a strong in-depth knowledge is important when you start adding horizontal bars to your I-shape.
· Be resilient to criticism. We’ve all received criticism in different ways. Don’t give up, and keep innovating with confidence. All feedback is good, so use it.
· Don’t forget about the 80–20 rule. The pace of development is increasing and you can’t wait until you’re done or at the 100% mark before starting to share your results with others. To gain speed, start building teams around that first 80% of work that you have achieved in 20% of time, and who knows, maybe you will find someone who would love to do the remaining 20% of your work efficiently.
Can you advise what is needed to engage more women into the AI industry?
AI is gradually being built in different areas — healthcare, art, geology, social sciences, etc. — this is a natural development as advancements in AI prove their benefits. Going back to the T-shaped persona I mentioned earlier, you can pick any area that inspires you and add that to the horizontal T-bar of applied AI. The combination will be powerful. And, as the industrial needs of marrying computer science with other domains grow, more channels will become available. For instance, this will be reflected in our educational system. Many programs have already been added that include data science and applied AI courses as part of an education.
What is your favorite “Life Lesson Quote”? Can you share a story of how that had relevance to your own life?
Einstein’s “Everything should be as simple as it can be, but not simpler” has always been my guiding quote. I’ve strived for simplification and automation throughout my life. In my early career I was working with model-based software development, that was correct by construction. Nowadays we are looking at system development and improvement in a data-driven automated fashion. And whenever I start with a new task I always ask myself how this can be done as simple as possible (but not simpler). This resonates very well with Ericsson’s Quest for Easy, which is the aim at developing technology that is easy to use, adapt and scale.
You are a person of great influence. If you could start a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. :-)
I think the movement is already there — the Sustainable Development Goals set by the United Nations General Assembly as a way of achieving a better and more sustainable future for all. It is a call to action by all countries and contains seventeen global goals for improvement of life quality worldwide. We need to keep those goals in the back of our heads with respect to anything we do.
For example, at Ericsson, we continuously work in partnership with global mobile operators, government ministries and UN agencies to lay the foundation of a technology and innovation groundwork that will help achieve the SDGs. Digitalization is an essential part of our lives, and the introduction of new technologies has the opportunity to enable positive change around the world — from boosting livelihoods, promoting financial inclusion and gender equality, and improving access to health, education, government services and more. We also believe that leveraging new technologies, such as AI and 5G, will be fundamental to reduce carbon emissions by up to 15% through solutions in energy, agriculture, buildings and services. Technology continues to transform the world we live in, bringing change where it’s most needed.
How can our readers follow you on social media?
· Twitter: @elena_fersman
Thank you for joining us!
