AI and HI: an unparalleled “dream team”

4th October 2023 Ms. Lucia Palomba
Article Authors

Ms. Lucia Palomba

Lucia Palomba has a background in economic studies at the University of Zurich. Moreover, she has also been able to deepen and develop her interest about the study of issues related to sustainability and sustainable investing.

In the spotlight today is Prof. Dr. Anne Scherer, Assistant Professor of Quantitative Marketing at the University of Zurich, who is a distinguished expert straddling the worlds of consumer psychology and technology with fervent AI exploration. 

Drawing from her exceptional background in these domains, Prof. Scherer will lead us on a journey to uncover her insights at the crossroads of cutting-edge Artificial Intelligence and its profound influence on human existence. Through her guidance, we will unveil new perspectives born from the intricate interplay and divergence of these interconnected dimensions.

1. What are the most recent developments in AI research, and how do you see AI influencing society going forward?

In recent years, one of the most significant advancements in AI has been the development of generative AI. This technology stands to reshape sectors like customer service, personalized education, and healthcare by tailoring communication and services to individuals. While its potential is immense, it's crucial to ensure that these innovations are accessible to everyone, preventing the inadvertent widening of societal disparities.

2. What motivated you to do research on AI and explore it in the form of a book, called "You and AI"? 

My motivation to delve into AI research and subsequently write "You and AI" stems from a long-standing fascination with emerging technologies. Prior to the AI boom, I had already been deeply involved in researching technological advancements. What stood out during my AI research was the distinct need to democratize AI knowledge. I felt that it was important to ensure that everyone, irrespective of their background, has a clear understanding of AI's implications and uses. This led to the birth of "You and AI" – a venture to promote AI literacy and ensure no one is left behind in this transformation.

3. AI represents a very polarized arena. On a side, it is supported by a massive wave of users all over the world, while on the other side, is still perceived by many as a limited field managed by algorithms-driven robots which face the challenges of understanding context and emotions in human language. What is your view about it?

The polarized perceptions of AI are indeed rooted in our varied understandings of intelligence. Often, we gauge intelligence through a uniquely human lens, deeming tasks simple for us as 'basic' and intricate calculations as 'advanced'. It's pivotal to recognize that artificial intelligence is very different from human intelligence. This divergence is not a flaw but an asset. AI excels where humans might falter and vice versa. Embracing this complementary nature allows us to harness AI's full potential, filling gaps in our capabilities and enhancing our endeavors.

4. In the book, you discuss the impact of AI in various industries, including healthcare and finance. Particularly, regarding the latter you personally conducted a relevant experiment on "self-learning AI", which can overcome the limit of a simple algorithm-driven machine. Could you elaborate on it and on some of the important points you made regarding its additional value, by emphasizing the primary stereotypes that people are most worried about? 

Certainly. The experiments we conducted in this area illuminated two critical insights. Firstly, people's trust wanes quickly if they believe a system, like an algorithm, is static and can't learn from mistakes. Conversely, when a system is labeled as "self-learning AI," trust and receptiveness increase, since there's an inherent assumption of adaptability. Secondly, terminology matters. Simply labeling a system as "AI" often leads people to believe it's inherently adaptive, while the term "algorithm" might imply a more rigid, unchanging mechanism. This underscores the importance of clear communication and terminology when discussing AI to manage expectations and foster trust.

5. In the future, do you think AI will become more "humanized" or will humans become more "dehumanized" because of constant AI use?

As technology, particularly AI, continues to advance, it will indeed offer more natural interaction mechanisms for us. Take, for instance, the strides made in conversational interfaces and natural language processing; they mimic human-like communication, creating smoother interfaces.

However, whether we choose to 'humanize' technology is often context-dependent. Our research indicates there are situations where anthropomorphic cues might be counterproductive, especially when we want to avoid eliciting social behaviors, like fears of social judgment.

It's undeniable that our interactions with technology will shape new norms and behavioral expectations. Some of these might influence our human-to-human interactions. Yet, our social behaviors, born out of centuries of evolution and culture, run deep. While AI will bring changes, I remain optimistic that the essence of our human interactions will remain robust, and we won't merely "turn into machines."

6. In the book you mention the term “deskilling”, which happens when people lose capabilities and training skills because they excessively use AI to be replaced in their tasks. In your view, how could this problem be efficiently solved or reduced?

The phenomenon of "deskilling" is a real concern when we overly depend on AI. To combat this, it's vital to approach AI as a supplemental tool, not a complete substitute. Take text optimization as an example: AI can aid in refining content, but the onus of fact-checking and ensuring context and accuracy rests with the human user. In essence, as we integrate AI into our workflows, we should use it to free up bandwidth for more critical thinking and validation tasks, ensuring we maintain and even enhance our skills in the process.

So in my view, a significant avenue to counteract deskilling is through continued education. As the landscape of our professions change due to AI, upskilling and reskilling become essential. Lifelong learning ensures we stay abreast of the latest developments and maintain a keen, critical perspective, irrespective of the tools at our disposal. Additionally, setting boundaries on AI's scope in certain tasks and periodically engaging in manual reviews or exercises can keep our skills sharp. It's about striking the right balance between leveraging AI's strengths and nurturing our human capabilities.

7. Could you please discuss what were the most unexpected findings you came across throughout your research process and the challenges you faced in explaining complex concepts to a broader audience? 

One of the profound challenges I encountered wasn't so much the intricacy of breaking down complex AI concepts, but rather the initial hurdle of capturing attention. Surprisingly, many are quick to dismiss a topic as too convoluted, often without giving it a genuine chance. This ties into one of our most unexpected research findings: many individuals overestimate their knowledge about AI (and many other topics). This self-perceived expertise, backed by pre-existing assumptions, tends to influence their behaviors more than new information.

What's more concerning is that a large number of people either skim through or entirely bypass reading and understanding vital information. This presents a considerable obstacle in our mission to enhance AI literacy across society's diverse strata. Addressing this requires not just simplifying complex ideas but also innovatively engaging and motivating people to invest their time and effort in learning and understanding AI's impact on their lives.

8. What are the most demanding challenges that AI may encounter in the future?

As AI becomes an integral part of our daily lives, several significant challenges stand out. At the forefront is the imperative for fairness and unbiased decision-making. When we delegate more and more critical decisions to AI – just think of college admissions or judgments about criminal justice – it's crucial that these algorithms operate devoid of biases. If left unchecked, such biases could amplify systemic disparities and lead to unfair outcomes. 

Furthermore, privacy concerning the data used for training AI systems is another pressing concern. As AI learns from vast amounts of data, ensuring that this data is both representative and respects individual privacy rights is crucial. Striking the right balance between AI's data appetite and the necessity to safeguard personal information will be a pivotal challenge as we move forward.

9. Could you please share the main takeaways for readers to reflect about AI's capabilities and limitations?

Sure! In essence, AI excels at identifying patterns within vast data sets and executing intricate calculations. This prowess has enabled AI to revolutionize areas like medical diagnosis and pioneering novel treatments. However, while AI is formidable in these aspects, it often lacks what we term as "common sense" and the creative capability to navigate uncharted scenarios – domains where humans truly excel. Moreover, while AI can analyze data and suggest solutions based on patterns, it doesn't possess ethical judgment. Decisions that demand moral deliberation, cultural nuance, or empathy inherently require human intervention. As we integrate AI more deeply into our lives and institutions, it's crucial to have humans in the loop for decisions that have ethical implications.

In essence, while AI brings its strengths to the table, human intelligence remains irreplaceable in many facets. Together, AI and humans form a synergistic partnership, combining the strengths of both to form an unparalleled "dream team".

As our insightful conversation with Prof. Scherer draws to a close, the scope of AI transcends mere technological progress, evolving into a canvas where innovative creativity intertwines with revolutionary ideas. Through her unique perspective, we catch a glimpse of AI's exceptional potential to reshape industries, enhance human experiences, and uncover unexplored territories of societal advancement. 

Professor Scherer's invaluable insights undoubtedly serve as a guiding beacon, steering us towards a future where the boundary between artificial and genuine blurs with exquisite subtlety. In this context, while AI excels in data analysis and suggesting pattern-based solutions, it remains devoid of ethical discernment. Decisions that necessitate moral reflection, cultural sensitivity, and empathy inherently require human engagement. As we progressively integrate AI into our lives and institutions, the presence of human input becomes indispensable for choices that carry ethical implications. 

In summary, while AI lends its strengths, human intellect's indispensability extends across numerous domains. Together, AI and humans can establish a harmonious partnership, fusing their finest attributes to craft an unparalleled "dream team."


Leave Your Comment