Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Creating a shared reality between humans and AI to foster authentic connections

Fast Company Executive Board member Jason Snyder believes intersubjectivity is vital to building more empathetic AI models.

[Courtesy of Jason Snyder]

Profile photo of Fast Company Executive Board

Most companies don’t have a chief philosophy officer, but they should. New technologies, particularly AI-powered, are advancing at a dizzying pace. These developments promise improved efficiency, productivity, and customer experience but also introduce complex philosophical and ethical challenges. 

Jason Snyder, global chief technology officer at Momentum Worldwide, an experiential agency within Interpublic Group of Companies, believes that intersubjectivity is crucial as companies continue to explore AI’s vast applications and implications. The philosopher Edmund Husserl coined the term, which refers to two or more subjects creating a shared understanding of reality. Intersubjectivity is rooted in empathy, enabling people—or, in this case, people and machines—to experience and interpret the world in similar ways.

“This notion is critical, especially for marketers, because the key to unlocking the full potential of the audience-brand interaction is having technology that can evolve beyond output that feels generated to output that’s fostering true connections,” says Snyder. “Right now, a big trend is relationship building with AI tools, using natural language through chatbots and large language models to mimic the kind of meaningful exchanges you would see in person-to-person customer service. The success of generative AI hinges on this ability to create a connection where these two parties, machines, and people, can understand and influence each other.” 

EXPLORING INTERSUBJECTIVE CHALLENGES AND OPPORTUNITIES 

A significant challenge with generative AI is that it can fall into patterns that feel generic and don’t deeply resonate with people. 

“In advertising and marketing, it’s all about bringing archetypes to life because a brand is just a metaphor for a story,” says Snyder. “Then the creative part is figuring out how to tell that story. Brands need to invest in developing GenAI models with refined empathetic understanding because the tech has to learn how to read between the lines of human communication. In intercultural communication, one of the core principles is around high-context and low-context communication styles. High-context communication relies on body language or nonverbal cues, while low-context communication is more straightforward. When you speak to a machine, even though there’s this implied notion of intimacy, it doesn’t share that high-context and low-context communication without real effort on the part of the developers.”

Snyder explains that another challenge brands must manage with GenAI is maintaining authenticity in users’ and technology’s shared reality. As AI creates more media and content, brands must be committed to transparency, ensuring people understand how and where AI is being used and putting measures in place to safeguard against manipulations. To address these challenges, Snyder recommends a multifaceted approach. 

“At Momentum, we use an adaptable feedback loop to support the ever-evolving relationship between GenAI and human audiences,” he says. “Intersubjectivity demands diversity. To avoid bias and blind spots, voices from all kinds of different backgrounds have to be present in every stage of the GenAI strategy. We need to help AI models understand real-world customer engagement data so that they can be further refined and trained to become more empathetic.” 

Momentum built and deployed a customer-facing AI ambassador for a multinational manufacturing client over two years ago. Snyder and his team are constantly training and refining the AI model’s technological and human engagement capabilities. 

“This project came into being before these large language models being available at scale commercially,” says Snyder. “We want to help any client understand that GenAI is not a shortcut to empathy. To have real intersubjectivity, you must invest in sophisticated technology and gather nuanced human insights. And so our technology architecture reflects that dynamic.” 

STRATEGIES FOR FOSTERING TRUE INTERSUBJECTIVITY

To continue making GenAI more engaging, empathetic, and intelligent, leaders need to take a proactive, hands-on approach, explains Snyder. The feedback loop of training, review, and refinement should be constant. 

“GenAI output has to be reviewed by people who can add nuance and ensure authenticity, checking that genuine intersubjective elements are present in it,” says Snyder. “You have to look beyond the shiny wow factor of the technology itself and ask: Does this move us closer to genuine interaction and understanding?” 

Companies should also evaluate and update their training data on an ongoing basis, looking for areas for improvement. Are they relying on heterogeneous information sources and data sets? Are they working to uncover hidden biases in the inputs and outputs? 

“To understand the outputs, you have to understand where the inputs are coming from,” says Snyder. “That means you need diverse training data. My developers strive to incorporate massive data sets that reflect a wide range of perspectives and cultural references to avoid mirroring just one segment of society. Feedback refinement is also key. We have teams dedicated to identifying biases and inaccuracies in outputs, so through human evaluation and technical tools, we’re constantly working to improve responses and minimize potential harm.”

Navigating this complex new landscape requires vigilance, curiosity, and humility. Leaders who recognize their strengths and limitations and those of the technology they use will be better equipped to make informed decisions.

“AI is a force multiplier,” says Snyder. “It’s an amplifier, and it’s going to amplify whatever we put inside of it at scale. We’re quickly moving from academic or philosophical concepts to implementing these ideas. This is a collaboration between humans and AI. We should always treat generative AI as a tool with human oversight—that’s how we get to nuanced understanding and ethical decision-making. Our focus must move beyond language and image processing to better emotional intelligence.”