A senior lawyer’s interaction with AI: Learning to ask good questions 

THE advent of generative artificial intelligence (AI) has revolutionised various sectors, and the legal profession is no exception. AI has the potential to significantly streamline legal work, reduce time-consuming manual tasks, and enhance the efficiency of legal processes. Rather than be intimidated by the technology, the better challenge for lawyers is how we can best harness AI to optimise legal work.

All lawyers, from in-house counsel and law firms to advocates and judges, need to embrace the use of AI even as it becomes an indispensable productivity tool across the legal profession. Goldman Sachs estimates that 44 per cent of current legal work tasks could be automated by AI, compared with an average of 25 per cent across all industries. This doesn’t mean that generative AI heralds the demise of the profession; rather, those who do not use AI will be displaced by those that do.

Put another way, those who do use AI may be doing legal work in a way not done today, which brings a transformative opportunity that allows them to forge ahead of the competition. This would be much like lawyers from the past who mastered the skill of word processing by using a desktop or laptop, instead of relying on someone else to type all their memos, opinions and agreements.

New world, new tech, new approaches

Embracing AI is becoming increasingly vital for lawyers as it continues to revolutionise the legal field. AI can significantly streamline the legal processes, improve efficiency and reduce costs. Training and tips on how lawyers can best approach and embrace AI are readily available: from understanding what it is and how it works to starting with small, manageable projects using AI for legal research or contract review, and choosing the right AI platform as a tool to augment the practice of law and the provision of legal services.

All these possible uses present interesting options. One is Harvey, a game-changing generative AI platform built and trained on OpenAI’s GPT-4, that is especially suited for lawyers. It uses natural language processing, machine learning and data analytics to automate and enhance various aspects of legal work. Presently, the likes of PwC and Allen & Overy have partnered with Harvey AI, and there is a reported waiting list of more than 15,000 law firms seeking to gain access to the competitive technology of Harvey.

Those who have seen Harvey in action have been impressed. In some respects, it’s a reminder of my early days in practice with an international law firm that owned and operated a Wang Laboratories word processor (which occupied a small conference room) that was capable of churning out agreements at speed and scale.

At a personal level, I recently spent a day with legal colleagues in my business learning about the technical prowess of ChatGPT, using simple type instructions to assign the task that I wanted the technology to accomplish, following which the technology generated a text-based result. The results were amazing.

I have yet to try Harvey, which I understand responds with multiple tools specifically for lawyers: Users can ask free-form questions that are legal or legal-adjacent, including research and summarisation; perform clause editing and strategic ideation; provide detailed legal research memorandum on any aspect of law; produce a detailed outline of a document to be drafted, including suggested content for each section. Additional “tricks” that Harvey is capable of include the ability to respond to complex, free-form questions about any uploaded document without any pre-training or labelling required.

This Harvey output is generated in seconds and in multiple languages. Impressive, to say the least.

Forget the answers, what’s the question?

The prowess of AI, while leaving one impressed, inevitably leads to the question of one’s job security.

While AI is quick and seemingly comprehensive in its response to queries, I am reminded by my technology colleagues that the answers and output I receive are only as good as my input – that is, my questions and instructions to the AI.

Although AI like Harvey and Chat GPT excel at understanding and responding to natural language – allowing one to interact with it as one would with another person – to ensure the best results, it is important to phrase one’s queries as clearly as possible. While these chatbots are capable of understanding context, the more specific that one is when framing the questions, the better the AI can assist with its response.

For example, if one were asking for information about a specific topic, providing as much background information as possible is useful. Likewise, being specific in one’s queries, keeping questions simple and straightforward, avoiding the use of jargon or complex phrases, and asking one question at a time help the AI. Launching multiple questions at once can confuse the AI and lead to inaccurate responses.

In some respects, it reminds me of my late grandmother chastising me to slow down when I was overly excited about a new-found topic of interest and launched into a series of rapid-fire questions to acquire knowledge without pausing to build up my knowledge and understanding in a clear, concerted manner. Often in inquiring AI, there is a need to rephrase one’s question. A different way of asking the same thing might yield better results.

In my limited experience with AI, I have found value in asking open-ended questions. Starting my questions with “how”, “why”, “what” and such has yielded me more valuable responses.

I have also been counselled by younger legal colleagues (who are more tech-savvy and enlightened) that patience and persistence with AI pays off. If at first one doesn’t get the desired answer, try again with a different phrasing or more specific details. After all, AI is constantly learning and improving.

Imperfections and progress

While AI is intelligent, it is not infallible. Growing familiarity with the technology will certainly build confidence in its use and ultimately greater reliance on the tool. It is like having an intelligent assistant who is smart, highly connected and always learning. However, cross-verifying important information that the AI provides is critical.

AI is known to have “hallucinations”, an example of which is DeepDream. The AI behind the program can “see” and create patterns and objects that are not there. This is a result of certain types of neural networks, such as convolutional neural networks, that are designed to identify and amplify patterns in the data they process.

While AI hallucination is an interesting phenomenon, it is important to note that it does not mean the AI is conscious or experiencing hallucinations the way humans do. It is simply a by-product of how these neural networks process and enhance patterns in data.

AI is certainly transforming the legal profession, offering a wealth of opportunities to streamline legal work, enhance efficiency and improve accuracy. By harnessing the momentum that is quickly building in this space, enlightened lawyers will do well to leverage the technology and ride it rather than resist its incorporation into the practice of law.

Perhaps the great encouragement I have in embracing AI is the analysis by experts that human lawyers still hold the upper hand in contextualising information, possessing emotional intelligence – all of which AI does not possess. How long humans will be able to hold on to this advantage is a matter of debate.

For now, what makes human lawyers valuable is the humanness, experience and judgment we possess. For that, it behoves us to challenge ourselves to rewire our thinking around how best to use AI as a tool to scale the intellectual summit of having higher-order critical thinking that makes us better lawyers than the technology. There is certainly much to reflect on as we prepare for the new year. For me, it begins with learning how to ask better questions.

The writer is group general counsel and chief sustainability officer of Jardine Cycle & Carriage, a member of the Jardine Matheson Group. He is also the CEO of mental health charity Jardine MINDSET and serves on the boards of the Singapore International Chamber of Commerce, Mind Forward Alliance Singapore and the Global Guiding Council of One Mind At Work. He is an adviser to the sustainability panel of SGX association SGListcos.