Engaging with colleagues and friends about generative AI (GenAI), I have been struck by how the conversations fall into three distinct categories.

The hallmarks of the first category are apprehension and scepticism when discussing the potential of tools like Open AI’s ChatGPT. Such discussions typically surround using the technology in legal or professional work. These conversations may likely continue, fuelled by more recent articles testing various GenAI tools and its accuracy and rates of hallucinations (see, for example, Problematic Stanford GenAI Study Takes Aim at Thomson Reuters + LexisNexis, Stanford HAI Tests Westlaw, But The GenAI Results Look Worse! and The Allens AI Australian law benchmark).

The second kind of conversation is hallmarked with feelings of curiosity. These discussions are with those who have yet to engage with the technology and seek advice on where to start.

Finally, the third category is conversations hallmarked by excitement and eagerness to explore more possibilities. The discernible difference in these conversations is that they are with people who have already played around with various GenAI tools but initially not for work. They used GenAI for fun or creative purposes. For example, they generate images and song lyrics or write parodies of popular songs. A personal favourite is the various videos of users experimenting with ChatGPT talking to itself. The introduction of Open AI’s GPT-4o may encourage similar experiments.

Reflecting on all these conversations set me on a path to understanding generative AI's role in creative thinking and problem-solving more broadly.

Many argue that embracing innovative technologies is essential for staying ahead of the curve in the changing world of legal practice. As practitioners navigate complex challenges, leveraging generative AI can unlock new possibilities, and one of the biggest opportunities is to augment human creativity.

By engaging with GenAI, lawyers have access to a broad information synthesis tool, which can also generate fresh ideas and explore unconventional solutions to problems. As Russ Somers (CMO, Quantified.ai) emphasises, "Generative AI can lead to a revolution in corporate innovation, resulting in automated workflows, increased productivity, and supercharged brainstorming sessions."

The Mindset

From the outset, it is crucial to approach the integration of GenAI into legal practice with a strategic mindset. Research out of Stanford University’s d.School suggests that simply providing teams with access to AI does not guarantee enhanced problem-solving capabilities.

Stanford University’s Jeremy Utley and Kian Gohar sought to examine the impact on how teams perform while using GenAI to assist them during brainstorming sessions. The hypothesis was that AI would improve a team’s problem-solving capability. The researchers described an experiment they ran whereby two groups engaged in a brainstorming session to solve a problem. Unaware of each other, the two groups brainstormed in two distinct ways.

The first group, the control, brainstormed the traditional way using Post-it notes. The second group used AI to assist in idea generation. The research found that those AI-assisted teams were not generating more ideas than the control group. In fact, they generated substantially fewer ideas. In addition, the teams using AI converged on average-quality solutions. The researchers commented that “[AI-assisted] teams were totally quiet like they basically had a resting AI face because they were staring into their computer with ChatGPT and trying to come up with solutions on their own. They weren't really talking to each other, even though they were right there.”

Adopting a collaborative and iterative approach is essential to truly harnessing the power of GenAI. The Stanford researchers discovered during the experiment that “…most folks interfaced with ChatGPT like a Google search query, with a single query and a single result.” They concluded that “to close the gap between what's possible and what's reality, we need to change how we think, work, and collaborate when using this technology.” Rather than treating AI as a mere search engine, professionals should engage in dynamic conversations with it, leveraging its ability to provide prompts, challenge assumptions, and explore alternative perspectives.

The Power of Play

When embarking on the journey of incorporating GenAI into practice, it is crucial to approach it with a playful mindset. As Somers emphasises in his discussion on using AI, “We learn fastest and best when we're playing.” By starting with personal interests, hobbies, or passions, professionals can familiarise themselves with the capabilities of generative AI in a low-pressure environment. This playful exploration lays the foundation for more serious applications in the workplace.

To act as an example, recently, I wanted to learn more about coding in Python. I decided to ask ChatGPT-4o the following:

Such an approach can apply to any area of interest. Another example prompt is “I’m planning a trip to Japan. Can you help me create a detailed itinerary for a week?”

When approaching the incorporation of GenAI into work, Russ offers a simple framework of levels. The first level is what everyone thinks they should do with GenAI, “take the things I don’t want to do… and outsource them to GPT. We know how to complete these tasks, so we can easily verify and check the results produced”. The second level is “getting GenAI to do the things we don’t know how to do. For example, I don’t know how to create fantastic social media posts, so I outsource that to GenAI.”

Leveraging AI for Ideation

Generative AI tools can be invaluable partners in the ideation process. Practitioners can generate a wide range of potential solutions by prompting the AI with specific challenges or problems. However, it is essential to go beyond treating the AI as a mere search engine. Engaging in a back-and-forth dialogue, iterating on ideas, and using AI as a devil's advocate can lead to more innovative and well-rounded solutions.

For example, one approach suggested by Tojin Eapen, Daniel Finkenstadt, Josh Folk, and Lokesh Venkataswamy is to use GenAI to promote divergent thinking. The authors suggest generating ideas through a process of trisociation — connecting three distinct entities. Try their prompt in ChatGPT and see what it produces:

“You will play the role of an ideator. You will randomly generate ten common nouns. You will then randomly select any two of the ten nouns. You will then ask me for a third noun. You will generate a business idea by combining or associating the two nouns you identified with the noun I identified.”

Here are my results in GPT-4o:

Collaborative Innovation

While generative AI has the potential to supercharge individual brainstorming sessions, its true power lies in facilitating collaborative innovation within teams. Kian and Jeremy’s research highlights the importance of effective team dynamics. Teams that actively engage with each other, discuss and build upon AI-generated ideas, and leverage the technology for iterative problem-solving tend to produce higher-quality solutions than those who rely on AI as a passive tool.

Eapen, Finkenstadt, Folk, and Venkataswamy suggest using multiple GenAI tools to build upon ideas to facilitate collaboration. Collaborators could first use ChatGPT to generate details of an idea before turning to tools like Stable Diffusion to obtain visual designs to build and iterate. Other experiments have even demonstrated that when multiple GenAI chatbots collaborate on a problem, it can help offset some of their limitations.

Overcoming Cognitive Biases

As with any new technology, the introduction of generative AI comes with its own challenges. One significant hurdle is the cognitive bias that can hinder AI-assisted teams from reaching their full potential. Legal professionals must be aware of the tendency to settle for mediocre ideas generated by AI and actively work to overcome this bias. By fostering a culture of critical thinking, encouraging diverse perspectives, and iterating on AI-generated ideas, teams can unlock the true potential of generative AI for innovation.

Building upon the above idea, I asked ChatGPT to “evaluate the business idea along the dimensions of creativity such as novelty, feasibility, specificity, impact, and workability.”

Generative AI tools like ChatGPT have the potential to revolutionise creative thinking and innovation in legal practice. Legal professionals can unlock new levels of creativity and problem-solving by approaching these tools with a playful mindset, leveraging them for collaborative ideation, and actively overcoming cognitive biases. As the legal industry continues to evolve, those who embrace the power of generative AI and harness its potential for innovation will be well-positioned to thrive in the future of law.

About the Author

Mitchell Adams is a senior lecturer at Swinburne Law School and a CLI Distinguished Fellow (Emerging Technologies) with the Centre for Legal Innovation (CLI). His specialisation lies at the intersection of law and technology, as an expert in intellectual property law, legal tech and legal design. Mitchell heads up the Legal Technology and Design Clinics at Swinburne Law School, where he creates the conditions for legal innovation through student and industry collaboration. As a CLI Fellow, Mitchell is working on developing education opportunities in prompt engineering for GenAI and a framework for the classification of legal GenAI.