The AI Revolution in Mediation: Hype or Reality
Artificial Intelligence (AI) has suddenly emerged as a disruptive force across numerous sectors, including the field of Alternative Dispute Resolution (ADR). The integration of AI into the mediation process holds the promise of enhanced efficiency, reduced costs, and improved access to justice. However, at this stage, it is virtually impossible to fully grasp the potential impact on mediation, both in its immediate and long-term effects.
AI has been around since the 1950s and we are all familiar with it in one form or another. But GPT type AI is a different technology compared to earlier AI. It will change our work in mediation in many different ways.
What is CHAT-GPT? Let’s see what answer the AI gives us. Here are answers from CHAT-GPT4o:
Q: Please explain in simple terms what GPT AI is
A: ChatGPT
GPT (Generative Pretrained Transformer) is an advanced AI model developed by OpenAI that can understand and generate human-like text. It works by learning patterns in language from a massive amount of text data, so it can respond to questions, write essays, translate languages, or even help with creative writing, based on the context it’s given. Essentially, it predicts what word or sentence comes next in a conversation or text, making it seem like it’s “talking” or “writing” in a very natural way.
It’s helpful for tasks like answering questions, generating ideas, or providing information, but it doesn’t have emotions or consciousness—it’s just a sophisticated tool for processing language.
Q: Can you think?
A: ChatGPT
I don’t actually “think” the way humans do. Instead, I analyze patterns in the data I’ve been trained on and generate responses based on those patterns. While I can understand and generate text, answer questions, and even provide suggestions, all of this is based on probabilities and data, not independent thought or consciousness. So, I don’t have personal experiences, beliefs, or the ability to reason like a human being. My goal is to assist by providing useful and relevant information.
Q: Will you be able to think in the future?
A: ChatGPT
As of now, AI like me isn’t designed to “think” in the human sense—meaning we don’t have consciousness, self-awareness, or the ability to experience emotions or reason independently. Whether AI will ever truly be able to think like humans is a topic of ongoing research and debate.
Many experts believe that while AI will continue to become more advanced in processing information and mimicking certain aspects of human thought, like problem-solving and decision-making, achieving actual consciousness or true human-like thinking might not be possible—or at least not anytime soon. There are complex philosophical and technical challenges involved in replicating human cognition, emotions, and awareness.
So, for the foreseeable future, AI will likely remain a highly advanced tool for processing information rather than something that can truly think or experience the world.
I have added the underlining as it may be concerning since it is not ruling out the idea of having the ability to think!
Generative AI will change the way we practice collaborative processes. Will AI also change the nature of our work? Most definitely. While we fear what will happen in the short term, we should be more concerned about the long-term impact of AI. Comparing AI to other technologies supports this concern. Amara’s Law states that we tend to overestimate the effect of a technology in the short run and underestimate its impact in the long run. This will likely remain true for the use of AI in mediation.[1] Will it also change the nature of human thinking over the long term?
In the short term, the adoption of AI in mediation has been accompanied by a combination of excitement and great expectations tempered by fear of the unknown. There seems to be a common belief in the mediation community that AI will transform the way disputes are resolved by making processes faster and more efficient. That is likely true, but this initial hype may lead to disappointment when the technology fails to deliver the anticipated results. However, if it follows Amera’s Law, AI will do far more in the long run than we expect.
Gartner’s Hype Cycle is a framework for helping us to understand the evolution and adoption of emerging technologies and can help us in providing some guidance to the likely path of AI adoption as opposed to the incredibly optimistic view of many.[2]
The cycle consists of five phases:- Technology Trigger: The initial breakthrough spurs interest.
- Peak of Inflated Expectations: Over-enthusiasm and unrealistic projections for immediate impact of the technology.
- Trough of Disillusionment: Realization of limitations, leading to decreased interest.
- Slope of Enlightenment: Gradual understanding of the technology’s benefits and more realistic expectations.
- Plateau of Productivity: Mainstream adoption and consistent application, with clear benefits.
Where are we at with AI in mediation? Is it transitioning from the Peak of Inflated Expectations to the Trough of Disillusionment? During the Peak phase, there was/is great optimism that AI has potential to immediately revolutionize mediation practice.
However, as we encounter unexpected challenges such as bias, hallucinations, lack of transparency (the black box), privacy, data protection, informed consent, and the necessity for human oversight, a good deal of the of the initial excitement is wearing off, much to the relief of many practitioners.
The Trough of Disillusionment is a critical phase where developers and practitioners learn to manage the technology’s limitations and work on addressing identified gaps. We are not there yet.
As AI technologies continue to mature and become more powerful, they will enter the Slope of Enlightenment, where the benefits to us will be better understood. More realistic and effective use in practice will emerge until AI reaches the Plateau of Productivity, where it will be seamlessly integrated into mediation and other ADR processes. That is a long way down the line from where we are today.
AI can already do many things faster than we can and it is especially good at detailed and often repetitive tasks that are part of our work. It can potentially eliminate some of the more tedious aspects of mediation practice. But it can go far beyond this, as a study by MIT in 2003 on AI and mid level professional writing tasks demonstrated,
The tasks in the study weren’t quite replicas of real work: They didn’t require precise factual accuracy or context about things like a company’s goals or a customer’s preferences. Still, a number of the study’s participants said the assignments were similar to things they’d written in their real jobs — and the benefits were substantial. Access to the assistive chatbot ChatGPT decreased the time it took workers to complete the tasks by 40 percent, and output quality, as measured by independent evaluators, rose by 18 percent.20%.[3]
We are seeing the emergence of specialized AI apps trained on industry/profession specific data. Development of this type of AI will be hindered by the lack of available data for training from mediation, it will be increasingly used by lawyers and arbitrators.
AI can be a supportive tool, augmenting human capabilities rather than replacing mediators. It can already produce faster summaries of mediation briefs than human mediators. In order to maximize its potential benefit, mediators must become much more competent in the use of technology. AI is not a massive search engine. It is a conversation of prompts. Much as the was pushback against Online Dispute Resolution (ODR), there is pushback against the use of AI in mediation. Covid ended that. It is unlikely AI will be propelled to the forefront of practice by a similar event. Still it will continue developing towards greater use and integration into what we do.
One of the most promising uses of AI in mediation is running simulations in mediation training. While mediation training is commonplace, the opportunity for skills practice is not. Chat-GBT can be parties, mediator or counsel in roleplays to help the least experienced mediators gain experience. It can also provide feedback to the trainee mediator including suggesting possible alternatives approaches
AI-assisted work can be better faster cheaper and more enjoyable, but mediation practice will always needs human expertise to interpret, apply, predict, analyze and communicate. A real concern for mediators is AI demonstrating abilities to mimic empathy.[4] Humans are not good listeners. AI can be a much more focused listener. With pattern recognition combined with the ability to learn, we may have a new tool that can improve human relationships and interactions. We see the use of AI in counselling; why not mediation? We see empathy as one of the human skills required by mediators. An AI cannot replace this, or can it?
It will also require ethical and regulatory frameworks ensuring responsible AI use, focusing on transparency, fairness, and accountability to maintain trust which is such a critical part of the mediation process.
While AI can definitely enhance efficiency and handle routine tasks, it currently lacks the nuanced understanding of human emotions, cultural contexts, and complex interpersonal dynamics that are often critical in mediation.
The increasing integration of AI in ADR has understandably raised concerns among ADR professionals about job security and the future of our roles. Many mediators worry that AI might eventually replace human practitioners entirely. That is unlikely.
The empathy, creativity, and ethical judgment that mediation professionals bring to the table remains, at least for now, irreplaceable. Instead of viewing AI as a threat, professionals in the field should consider it as a powerful tool that can augment their capabilities, allowing them to focus on the more complex, high-value aspects of dispute resolution that require human insight and expertise.
The impact of AI today is just the beginning of an ongoing transformative change in the practice and process of mediation. Success in mediation is challenging to define. Success in integrating AI into mediation will require a balanced approach that capitalizes on AI’s strengths while addressing its limitations and ethical considerations.
As we navigate this uncertain future, we must remain adaptable and continue refining our approach to AI integration into mediation.
_________
[1] Pohan Lin, Amara’s Law and Its Place in the Future of Tech, 09/06/2024
https://www.computer.org/publications/tech-news/trends/amaras-law-and-tech-future
[2] Gartner 2024 Hype Cycle for Emerging Technologies Highlights Developer Productivity, Total Experience, AI and Security Stamford, Conn., August 21, 2024, https://www.gartner.com/en/newsroom/press-releases/2024-08-21-gartner-2024-hype-cycle-for-emerging-technologies-highlights-developer-productivity-total-experience-ai-and-security
[3] Zach Winn – MIT News July 14, 2023 – Study finds ChatGPT boosts worker productivity for some writing tasks. A new report by MIT researchers highlights the potential of generative AI to help workers with certain writing assignments. Available at: https://news.mit.edu/2023/study-finds-chatgpt-boosts-worker-productivity-writing-0714
[4] Lisa Bannon, Can AI Do Empathy Even Better Than Humans? Companies Are Trying It, October 7, 2023 The Wall Street Journal. https://www.wsj.com/tech/ai/ai-empathy-business-applications-technology-fc41aea2.
John Nota, Artificial Empathy: A Human Construct Borrowed by AI Understanding AI-based empathy requires a closer look at ourselves. October 9, 2023. Psychology Today.
https://www.psychologytoday.com/ca/blog/the-digital-self/202310/artificial-empathy-a-human-construct-borrowed-by-ai
Start the discussion!