Finding the Right Balance between Emotional and Artificial Intelligence
An exclusive interview with Fritz Lebowsky
Can you tell us a little bit about yourself and what do you bring to the table?
My background, and my PhD, are in human color image quality perception, color image signal processing and dynamic control systems theory. My focus has been on hierarchical system approaches, I look for rules and patterns exhibiting self-similarity across different levels of functional abstraction to more easily master complexity in astonishingly simple ways. My work at STMicroelectronics as a system engineer resulted in more than 25 technical publications and over 20 patents in 5 different application domains.
What we have discovered in product development founded on microelectronics, color science, and applied mathematics can have a wonderfully positive impact. I have learned with great admiration what color perception of the human brain is teaching us. Over the past 2 years I have also learned about innovative approaches in management and leadership. Now I am enthusiastic about transferring astonishing findings about human perception and performance into data science and business development.
How many people already experience the incredible potential of emotional and social intelligence? Activating these most powerful inner human values lets us develop excellent human resilience. Better yet, when combined with modern data science and intelligent metrics, we will experience how AI, ethics and social responsibility work well together. With such a comprehensive mindset we can achieve a significantly positive outcome while mitigating human conflicts and stress.
What challenges do you see in using AI in business today?
Currently, AI is often biased by quantitative statistical and probabilistic reasoning. We get excited about what works, create a main stream solution, and extrapolate in linear fashion. But much too often such solutions miss the weak signals or ‘outliers’ which tell us that something doesn’t yet fit in the periphery or across an interface poorly taken into consideration. AI focuses on more easily detectable patterns but has difficulties in effectively picking up the noise of the outliers; As a result, unexpected small errors or disturbances begin to accumulate. There is often lack of relevant feedback to easily correct or improve the algorithms. We are still in need of efficient feedback loops that explore the ‘unexpected’ or ‘yet unknown’ and create full transparency for best possible human evaluation in complex system context.
For example, how often and how well are people’s reactions taken into account on modern Internet platforms? How reluctant are a significant number of product design managers about analyzing the ‘weak’ signals coming from frustrated costumers?
Moreover, functional verification — enabling database coherency after some software changes or software updates — is often lacking. For example, sometimes some French data bases all of a sudden have debited wrong bank accounts. We are infatuated with what works. But social responsibility for unexpected consequences seems outside our field of vision. Many system design managers responsible for system performance and ‘sign-off’ often ignore weak signals or outliers, resulting in critical performance flaws. Instead, one needs to create meaningful feedback from such events. In my opinion, people should gain a better understanding of ‘entropy’ to master complex ecosystems with humans in the loop. Consequently, we need to think more about well implementing dynamic systems that can prove their robustness and correctness at all levels of functional performance across the entire ecosystem. Here is where emotional intelligence can help multi-disciplinary teams or groups managing challenging system complexity.
How do you define emotional intelligence?
Let us consider a simple thought experiment: There are things that one can sense and there are things that one cannot sense. Still, someone else might well sense things that one cannot sense oneself. Therefore, with the help of others one becomes more resilient toward the unknown. Raising self-awareness about complementing each other opens a wonderful path to honing emotional and social intelligence. In addition, we could bootstrap new technology that assists us in exploring the unknown more efficiently and effectively.
We need to better understand how each and every situation is unique. Then we may want to adapt to each situation in most efficient and effective ways. Well orchestrated teamwork is much more powerful than brute force of individuals. We need to take more time to reflect on things we want to implement. We have to leave time for collective interpretation and reinterpretation. Open-minded questions generate a wonderful platform enabling most suitable answers and solutions. A proactive mindset is much more powerful and efficient than a reactive mindset. Let’s emphasize on training stakeholders on such forward looking mindsets. Consider putting empathy to work. Altogether, we can leverage social intelligence and collective intelligence to better cope with time pressure, short term profits, expert bias, and hidden flaws in economical models.
What are you proposing?
To activate each organization’s social and emotional intelligence
Based on my personal experience, the underlying challenge remains entirely human. It’s too simple to just blame technology. One should rather think more about human logical reasoning that barely considers social and collective intelligence. More specifically, I believe we face another crucial human paradox. Technology must remain secondary and shall ‘assist’ in doing things right by design, applying powerful functional verification of all the desired and undesired tasks possible in the overall ecosystems. People shall more easily learn how best to complement each other across disciplines instead of ‘destructively’ competing against each other without realizing how much mental and physical energy is being wasted collectively.
Instead of just focusing on the traditional solutions which we feel so comfortable with (while ignoring the unknown or masking out the unintended consequences of the underlying human paradoxes) I personally propose to walk the extra mile together, jointly figure out how best to embrace our most crucial human paradoxes, and elaborate simple examples leveraging social and collective intelligence in view of achieving buy-in toward sustainability for ALL — from all stakeholders. One should focus more on mitigation and mediation. Let us avoid just measuring against singular objectives (example: how much profit?, how much more profit than last year?). Instead, let us create a meaningful set of alternatives and multiple objectives (example: how many types of meaningful returns of investment (ROI) can we think of?).
I would like to provide an example question applicable to a management team: How happy do you think your subordinates are with product performance as well as human performance improvement in their assignment, work, or team context? How well are your subordinates’ points of view being considered? This will reveal a potential gap between managerial expectations and collaborator perceptions. Data analytics can play a powerful role in analyzing the gap. Once we have sufficient feedback, we can begin to plot out how this gap is closing over time. Carefully processing all kinds of emotional information is crucial (see Dr. Sarah Spradlin’s comment: Managing Emotional Information with all its options).
What would you like to contribute and how? EI and AI shouldn’t compete or contradict each other. Instead, we have to make sure that we create very meaningful synergies. We imagine to activate each individual's as well as a team’ available ‘brain power’ in an optimal way, the FLOW STATE. We would like to show how people can get comfortable with the full power of multi-dimensional models while learning how to apply the underlying pragmatic methods. As a result, projects illustrate agility while tasks are being accomplished efficiently and effectively. People can achieve tasks in a matter of days — instead of weeks — and find sufficient time to recreate when meaningful metrics have been put in place.
We work from the visible to explore the significance of the outliers and the yet invisible. Figure 1 illustrates the overall context in industrial design. We started from Embodiment Theory highlighting how mind and body interact. One can more successfully understand the dynamics of such simple interaction. Then we introduced a two dimensional space to add AI and EI ( as force vectors of disturbances). We have at the origin the flow state: it reflects what we feel and what we recognize as our optimal state of performance. Every disturbance is measured holistically in relative quantities (as eccentricity) in relationship to the flow state. One may consider disturbances as an event that triggers a learning process.
Feedback from the participants/stakeholders is crucial to well understand what’s needed to more easily return to the desired flow state. We can use time series data to measure latency and reactivity to disturbances. A third dimension represents complexity in which environmental or systemic concerns can be described. The associated flow state transforms into a Black Magic ‘flow state’ which can be mastered by a group of experts capable of exploring shared meaning. AI can complement direct observations and therefore contribute in a multitude of additional spaces that are extremely difficult to explore manually by humans only. We explore hypothetical classes and check for their relevance in overall system context. Most importantly, systematic feedback analysis is absolutely necessary to reveal the significance of the underlying hidden components (blind spots).
What do you see as the challenges in moving forward?
Here is what I am trying to improve. Most humans need reassurance to move forward. It’s about growing the number of people who want to learn taking meaningful steps that embrace change together with the yet unknown, not only from a technological perspective but also from a people perspective focusing on the development of thorough social responsibility while moving toward positive global impact of ‘doing things right by construction.’
How best could we jointly explore structural complexity rather than look just for proxies and quantitative analysis coping with bias? We already developed generic tools with powerful ‘integral error control’ feedback loops founded on nonlinear statistics. We can apply them to specific pain points. I would love to partner with sponsors willing to explore such promising new methodologies. Based on my past experiences we can achieve significant success in just 4 to 6 months by establishing powerful links between technology, human creativity and social intelligence. It is important to address all three ‘domains’ concurrently as they are fully interdependent.
With a dedicated set of questions I would like to express my personal challenge and curiosity in searching for future answers: How best can we transform human ideas and hypotheses from their verbal or textual representation to a functionally and economically meaningful dynamic system context that can be thoroughly verified not only by people but also by algorithms or machines? How best can ‘machines’ explore and reveal complementary components all the way to the other end of the spectrum that many people cannot easily see in challenging multi-disciplinary ecosystems? Whom do you know who is interested in jointly exploring such an endeavor while applying modern collective intelligence, technology, and creative design thinking in significantly pragmatic structural context?
That said, you may enjoy listening to a key example which my partner Dr. Sarah Spradlin recently provided: Transforming Rejection into Redirection which increases Resilience ).
Empathy bridges the Sensemaking between the Engineer and the Enduser
Your Questions and comments will be very much appreciated!
Fritz Lebowsky December 13, 2019 fritz.lebowsky@icloud.com
