Harnessing voters' emotions has always been a cornerstone of democratic politics, but AI is poised to elevate this practice to unprecedented heights. By revealing patterns in our collective emotional intelligence (EI), AI could transform our politics—for better or worse. Let’s explore how AI reshapes how parties gather and use emotional information.

From Traditional Methods to Modern Challenges

To understand AI's potential, we must first look at how political leaders have traditionally gathered and used emotional information. Not so long ago, political leaders relied on their caucus members to report on voters' views and sentiments. Emotions were critical in shaping beliefs, but individual reports were irregular and subjective.

The advent of telephones revolutionized the process. Pollsters quickly reached representative samples, providing immediate, reliable public opinion insights. Focus groups added depth, allowing detailed discussions of voter attitudes and feelings. Still, these methods are far from ideal. While they are faster and more rigorous than caucus reports, the data is often patchy and imprecise, especially on complex issues.

Consider polling on a controversial decision to build a mine. Data might reveal conditions for compromise between environmental activists, industry, and community members, but nuanced questions about balancing these interests are likely to yield unreliable answers. At this level of granularity, emotions are typically “mixed” and volatile, making data interpretation risky.

Polling is most reliable on binary questions with high emotional content like taxes or abortion. These issues provoke clear choices and predictable outcomes, making message crafting easier. After 50 years, strategists know polling is a blunt instrument best used for provoking specific emotional responses—negative (anger, frustration) or positive (joy, empathy).

From a political standpoint, this mechanistic approach tends to favour short-term strategies for quick wins on “hot button” issues rather than fostering long-term relationships around complex issues.

Emotional Intelligence and Contemporary Politics

Many psychologists believe the mechanistic view misrepresents our emotions. Far from being blind and impulsive, emotions are often purposeful and smart, interacting in ways that help us evaluate alternatives, set goals, plan, and make informed choices. Emotions, we could say, have their own logic. People with high Emotional Intelligence (EI) are self-aware, empathetic, and skilled in regulating emotions. EI helps us navigate situations and interact appropriately with others.

EI also operates at the collective level. Collective Emotional Intelligence (CEI) focuses on shared group emotions such as public anxiety, optimism, anger, or enthusiasm. Just as EI reflects individual emotions, CEI provides a comprehensive view of collective emotions and their impact on behavior.

This raises a critical political question: Can AI discover robust, nuanced, and reliable CEI patterns in the social environment? The answer is almost certainly yes. For AI, finding CEI patterns is no different from analyzing traffic flows, stock market transactions, or weather phenomena. However, success depends on lots of high-quality EI data—a challenge increasingly met through new AI tools.

New AI Tools for CEI

Recall our example of the mine. Now imagine a political leader who wants to balance the interests of environmental activists, industry, and community members. Polling and focus groups can provide a useful snapshot of stakeholder sentiments, but as we’ve seen, they are an unreliable guide to the underlying emotional connections and dynamics.

By contrast, AI tools like these are revolutionizing CEI analysis:

• Sentiment Analysis: Analyzes digital text to determine the emotional tone of messages (positive, negative, neutral) across media, social posts, and articles.
• Beyond Verbal: Analyzes vocal intonations to detect emotions, offering real-time insights during voter interactions like town hall meetings.
• Cogito: Provides real-time emotional feedback during phone interactions, helping campaign volunteers adapt strategies.
• Affectiva: Uses facial expressions and vocal tones to assess policy resonance with audiences, providing insights during focus groups or speeches.

These tools show how AI already aggregates and analyzes emotional data at a collective level. They model complex emotional patterns such as anxiety related to climate change, optimism regarding the economy, and frustration over perceived governmental inaction. Within a few years, they will be incomparably better at both data collection and analysis. As they evolve, AI will also become increasingly skilled at using the data to craft targeted messages, policies, and interventions, and to assist political leaders in navigating and responding to evolving debates in real time.

Ethical Considerations and Future Implications

While these developments are promising and exciting, they raise serious ethical concerns. Privacy issues, potential manipulation, and the need for transparency and responsible AI use are critical. Without strict ethical guidelines, there is a risk that AI could be used to exploit voter emotions, undermining trust in the democratic process.

Looking ahead, AI's potential to transform political discourse and public policy is immense. AI-driven CEI could greatly enhance public engagement in areas like healthcare reform, climate change, and AI regulation.

Conclusion

So, will democracies make this shift?

There is no guarantee. These are challenging times for democracies. Political debates often revolve around simplistic and opposing views on complex issues, leading to fragmentation and polarization. A key takeaway from our analysis is that traditional data-gathering tools like polling are partly to blame. They encourage a mechanistic and transactional view of our emotions and politics.

As new AI tools emerge, political leaders have a unique opportunity to leverage these innovations to navigate complex issues more effectively. Embracing an emotionally intelligent approach could transform political discourse and foster deeper connections with the public.

However, this depends on their view of human emotions. Those who see our emotional life as largely blind, impulsive, and mechanistic likely believe the transactional approach is the only viable one for democratic politics.

As for me, I wonder—and worry—about how someone with these views might want to use these powerful new AI tools to advance their political goals. In my view, if we want AI to positively transform our democratic processes, we must have responsible use – and that starts with strong, clear, ethical guidelines.

Don Lenihan PhD is an expert in public engagement with a long-standing interest in how digital technologies are transforming societies, governments, and governance. This column appears weekly.