Students Learn that Values, Ethics, and Caution are Essential to Navigating AI

The advance of Artificial Intelligence is truly taking the world by storm. Its use is now so wide-spread, and its development so exponential – and in some cases unpredictable – that many are starting to get quite alarmed. Young people need to be equipped with knowledge to understand how best to navigate the vast abilities of this new technological frontier and help society address it going forward.

It was great timing then that through funding from For The Honor of Truth Fund an endowment initiative supported by the Class of 1965 centered on the principle that truth is an intrinsic good and the continued search for truth is the lodestar of an exemplary life — upper school Humanities students were able to work with leading expert Sarah Newman, on the role, societal implications, and outcomes of Artificial Intelligence.

Newman is the director of art and education at metaLAB at Harvard, at the Berkman Klein Center for Internet & Society. Working at the intersection of research and art, her work engages with technology’s role in human experience, and interrelations between complex systems. In addition to her art practice, Newman is also co-founder of the Data Nutrition Project, which designs tools and practices for responsible AI development. She is also an educator and leads creative workshops to address interdisciplinary research problems.

Starting in December 2022, Humanities faculty member Kelly Joseph and  upper school students from her Media, Culture and Politics course began working with Newman on a workshop dedicated to exploring the “truth” in social media and AI. On March 28, MB’s Sinclair Room was turned into a small AI conference. To start off, she asked students to come up with their own opportunities and risks regarding Artificial Intelligence. Many people fear AI because of just how truly human it seems, and Newman agreed that it is important to be cautious about the potential dangers of the technology. Yet, at the same time, she urged students to remember that AI is programmed to appear real or sentient even when it is not.

She delved into the idea of “values” and emphasized that “our systems are only as ethical as the values we embed them with.” Also, technologies can have unpredictable consequences beyond what is intended when they are developed and implemented.

Senior Jake M. found this portion of the session particularly illuminating.

“I had never thought about how AI could impact the creation of false information. I also really liked the acknowledgment that people can have positive intentions with AI while still causing harm,” Jake shared.

Mid-way in, Newman asked students to brainstorm individually about how the technologies might impact the world. She then worked to group the students together based on their interests. Once in groups, they got to have fun building a physical representation of their AI concept connected to a query they developed as a group, such as: How might we trust AI when it is based on human behaviors and value, if these are subjective? How might AI influence the future of warfare?

They were all given a chance to show Newman their creation and she asked them questions in a lively back and forth.

I found it super interesting to learn from someone who plays such an integral role on the forefront of artificial intelligence research,” said junior Emma M. “It definitely opened my eyes more toward all of the ways AI is intertwined in our lives.”

Regarding the hands-on aspect of the collaboration, Jake added, “I really enjoyed how we used art to represent our feelings about AI and its growing impact on our lives. That created a really fun way to visualize our thoughts about AI and allowed us to be creative.”

Equipped with this knowledge, for the second phase of the project they were asked to develop an AI prompt (e.g.: How will AI impact education? How will it impact society or our wellbeing) along with a physical representation of this idea from their perspective. They then were to use an AI image generator or chat tool to get a “creative” response from the technology along the same idea. They set a date (April 20) to reconvene with Newman to present their projects.

In the backdrop of this academic project is the real-world debate about what to do about the speed of powerful AI developments. One influential group  that calls itself the Future of Life Institute described these developments as “Giant AI Experiments.” They released a high-profile letter calling to halt development of these power technologies for six months. This letter was posted just days before Newman’s first MB visit. She made sure to point out that it’s important to look at all sides in these types of debates. Some don’t agree with this development halt.

The debate aside, the specter of so many people in a state of alarm only reinforced the urgency for students to bolster their understanding of AI’s implications. It also caused them to consider for themselves how much of a presence they would like the technology to have in their lives going forward.

Newman used the presentations, this time taking place in the Y-Lab, as in-roads to ask students questions about their level of optimism, or pessimism, regarding AI. “This is a really interesting time to be going to high school or college because so much is changing,” she said to them. “How do you feel about these rapid changes at this time in your life?”

Junior Abby M. responded that she feels “it’s a little bit scary but also a little bit interesting. Unpredictability is a factor in so many ways. At colleges you don’t know how many people will use it, even if you choose not to.”

Newman concurred that there are “so many unknowns.”

One group directly addressed the anxiety of AI’s impact in the classroom with their research subject: “The fear and negative perceptions by teachers that AI will eventually take their jobs away from them.”

Using the AI tool ChatGPT they entered the prompt “Kids in a classroom following a computer overlord.” The resulting image: children in a classroom all with faces that were distorted, which seemed to convey uneasiness and uncertainty.

A student in this group shared their impression that “we have to be careful and precise if we want to integrate AI in the classroom, making sure the use of it doesn’t lead students to rely on AI and does not take away the need for human teachers.”

Another group explored the idea of “Protecting the mind from advancing AI” with their own visual representation including a Buddha surrounded by police tape. In this case, the group means “mind” as mental wellness. They used an AI image generator called DALL-E to get a response using the simple prompt: “Protecting The Mind.” DALL-E responded with a superhero image with a helmet.

They also did article research to explore different perspectives on how societies will be impacted by AI, uncovering scenarios where governments maintained control, but others that were more bleak: foreseeing AI as being weaponized with countries using it to target one another.

Again Newman asked, are you feeling more optimistic or pessimistic?

Senior Aiden B. said he is encouraged that more controls can be placed on AI especially regarding data collection, and that it can be used toward more positive ends, such as employing it as an analytics tool. “I feel that AI could be put to work in a much more useful way than just Chatbots, and that makes me more optimistic,” he added.

Another group noted that some AI developers themselves feel pessimistic about AI’s future impact on society, which only reinforces pessimism. Public awareness (or lack thereof) of AI’s presence in their lives already was noted as a disconcerting factor in the overall outlook.

Why not ask AI how it thinks it will impact the world? Great idea, one group did just that!

Utilizing the tool ChatGPT, the tool handed back a full essay on the topic, which students used in their presentation: “As an AI language model, I can say that predicting the future of AI is a challenging task, precisely because AI is advancing rapidly and evolving in unpredictable ways. It is true that AI has already transformed the way we live and work in ways that would have been difficult to imagine just a few decades ago.”

These students called for more exploration and consensus on how to steer the use of the technology. “As a society, we should be discussing what we are trying to achieve with AI. We need to be aware of the rate that it is progressing, and we should be discussing when to tell it it’s gone too far. Going forward we hope that people don’t put all of their trust in AI since its capabilities and security are so unknown. We should consider the downfalls while we are still able to control the problem.”

Thanks to the Class of 1965 for providing this powerful and important learning opportunity.

About For the Honor of Truth Fund

For The Honor of Truth Fund is an endowment initiative supported by the Class of 1965 centered on the principle that truth is an intrinsic good and the continued search for truth is the lodestar of an exemplary life. This fund provides important learning opportunities for students, including exploration of quickly emerging technologies such as artificial intelligence.