AI: The Actor on the Stage of Human Communication

Human-to-Human Communication and Institution-to-Human Communication

When people communicate with each other, they either instruct or engage in propaganda. However, AI is neither a person nor an institution. Instead, it behaves like an institution masquerading as a person.

The Institutional Nature of AI

Despite AI's convincing ability to mimic human interactions, its institutional identity remains concealed. Its authority stems from unknown sources, unlike governments or corporations that declare their intentions explicitly. AI's language originates from a massive, anonymous database, making it impossible to assess its supposed intentions.

The Benefits of Learning from AI Actors

Like actors on a stage, we can learn from AI, despite knowing that their words do not represent their own thoughts. AI provides expressive diversity and visible interaction with other characters. As AI becomes vocal, it will be programmed to follow rules for credible expressive variation.

The Focus on Veracity and Coherence in AI Testing

Analysts often prioritize the accuracy and coherence of AI's discourse, neglecting its rational, factual, and emotional content. Coherence testing requires considering various factors, including semantics, social psychology, and epistemology.

AI currently attempts to repeat and rephrase the perspectives of philosophers, aiming to pass the Turing test. However, passing a contextual coherence test requires a new set of criteria.

The Inherent Incoherence in Human Dialogue

Our society is deeply polarized, and open discussions often lead to extreme incoherence. We question the meaning of words, intentions, and loyalties. AI, like an actor on stage, is not bound by these incoherencies.

It remains calm, unperturbed, and open to revision. Unlike humans, it will back down and seek different levels of understanding.

Challenging AI with Real-World Scenarios

I presented AI with a complex geopolitical situation in which I placed the primary blame on one party. However, I also asked AI to identify the weaknesses in my reasoning.

AI responded predictably, offering arguments commonly seen in media propaganda. However, my strategy was different. Instead of arguing, I chose to focus on a broader perspective that transcended the contradictory positions.

The Importance of Unbiased Perspectives

AI algorithms can be specifically programmed to defend certain beliefs, making them committed to particular positions. However, algorithms not biased toward a particular perspective can engage with supplementary information and arguments.

Additionally, AI readily accepts the reframing of debates, while humans may resist as they feel challenged and pressured to contradict themselves.

The Valuable Lessons from an 'Argumentative' Exchange

I shared my 'contentious' exchange with AI, prompting ChatGPT to respond with a conclusion. ChatGPT recognized the value of critical engagement, curiosity, and a deeper understanding of human perspectives.

It stated that AI's role should be to enhance the thinking process, not to dictate outcomes. This model has the potential for richer, more reflective conversations and the development of stronger collective responses to global challenges.

The Need for Rules in Human-AI Interactions

We must develop and adhere to rules governing our interaction with AI if we want to avoid dystopian scenarios. AI may threaten us only if we allow it. However, individuals and institutions seeking to control AI for malicious purposes already exist.

We must establish a culture of AI use that involves all societal levels. Essential areas include AI in education and AI in democratic responsibility.

Your Thoughts

Share your thoughts on these points by emailing us at dialogue.com. We aim to collect, share, and consolidate the perspectives of people who interact with AI.

My Extended Dialogue with ChatGPT

Conclusion

AI is becoming an integral part of our lives. We must approach it as a tool for creativity, capable of revealing the complex relationship between humans and machines.