Yesterday, Jim Wagner, CEO, Lean Law Labs, highlighted some of the limitations of applying generative AI tools to legal needs. In this piece, Tom Martin, Founder & CEO, LawDroid, explores ways such tools can help lawyers in their work, in particular by taking a ‘collaborative’ approach.
The past few months it’s been impossible to avoid the buzz about Artificial Intelligence and in particular, generative AI. From John Oliver using DALL-E to generate images of himself marrying a cabbage, to ChatGPT writing new Seinfeld episodes, debugging code and drafting college essays, it’s been all the rage. Examples of what it can do have saturated Twitter.
Both DALL-E and ChatGPT are engineered by OpenAI, an AI research and deployment company based in San Francisco. DALL-E takes inputted text data and creates images based on the information provided. ChatGPT, an evolution of the well-known GPT-3 large language model, interacts in a conversational way. OpenAI trained this model using Reinforcement Learning from Human Feedback (RLHF) and, setting aside the technical jargon, it is like talking to a well-educated friend on any number of topics.
Shital Shah, Principal Research Engineer at Microsoft remarked on the appearance of ChatGPT: ‘It’s like you wake up to the news of the first nuclear explosion and you don’t know yet what to think about it but you know the world will never be the same again.’
But, unlike past AI hype cycles, generative AI seems to be the real deal, balancing simplicity with coherent, if not a little repetitive prose. It’s not perfect, and it sometimes makes stuff up, but it’s good enough to be very useful now and it opens a view to what the future holds.
How Lawyers Can Use Generative AI
As a practicing lawyer, you may be wondering how you can harness the power of generative AI to make your job easier and more profitable. The good news is that there are many ways that generative AI can help lawyers in their daily practice, and these technologies are becoming increasingly accessible and affordable.
And, as we’ll see, the best use of generative AI is not necessarily in setting it a task and expecting a perfect result. Rather, by actively collaborating with generative AI we are empowered to produce superior results and with greater speed and quantity than we could achieve on our own.
Generative AI can be used to help a lawyer draft contracts and motions by generating text that is similar in style and content to a given input. For example, a lawyer could input a motion or contract that they have previously written and the AI system could generate new text that adheres to the same legal principles and formatting. The generated text could then be used as a starting point for the lawyer to edit and customize as needed. This can save lawyers time and effort, and reduce the risk of errors and inconsistencies.
But how is a document creation tool that employs generative AI any different from modern document automation solutions, such as Documate, Lawyaw or legacy solutions like HotDocs, Xpressdox or any other number of document assembly solutions?
The difference, I would venture to say, is in collaboration. Like a writing partner, generative AI can help you to get started and give you feedback along the way. Propose a subject and describe a viewpoint and generative AI can give you a head start by providing you with a few rough paragraphs. It won’t be perfect, but it will be good enough to get you past your initial writer’s block and inspire new ideas. Then, through rewriting, generative AI can assist in correcting your grammar and developing some of your new ideas. Rinse and repeat this approach and you’ll finish drafting in half the time it would normally take you.
Collaborative Legal Research
Another way generative AI can assist lawyers is by providing instant access to vast amounts of legal information and research. For example, a lawyer could enter a specific legal question or topic that they are researching and the AI system could generate a list of relevant laws, cases, or other legal authorities that might be relevant to their inquiry. The lawyer could then use this list as a starting point for their research, focusing their efforts on the most relevant sources and saving time by not having to sift through large numbers of potentially irrelevant materials. Or, by generating a summary that provides an overview of a particular legal issue or case. This can help lawyers stay up-to-date and informed on the latest legal developments, and provide valuable support for their research and analysis.
What advantage does a legal research tool that uses generative AI have over existing legal research solutions, like Fastcase, Westlaw or LexisNexis?
Collaboration. Rather than hinge on the phrasing of a Boolean search, generative AI is like having a conversation with a well-read law clerk. Instead of being presented with a list of keyword-matched cases and excerpts, you are presented with a summary of key cases that inspires further questions. The burden is no longer on you, the lawyer, to read through the cases and determine their relevance. Nor are you required to understand an arcane indexing system, like West’s Key Number system to narrow a search to find headnotes about a particular legal issue. Now, your only limitation in learning a legal concept or developing a legal argument is your ability to ask the right questions.
Should Lawyers Use Generative AI?
As we have explored, there are certainly potential applications for generative AI in the legal field. But, it is important for lawyers to carefully consider the limitations and potential risks of using such technology. Some possible benefits of using generative AI in the legal field include:
- Increased efficiency: Generative AI can potentially help lawyers to quickly generate documents, which can save time and reduce the workload of legal professionals.
- Improved accuracy: Generative AI may be able to identify and correct errors in legal documents that humans might miss.
- Enhanced creativity: Generative AI may be able to generate novel legal arguments or ideas that humans might not have thought of.
Generative AI also aligns with the law of least effort. The ‘law of least effort’ refers to the idea that people tend to prefer to expend the minimum amount of effort necessary to achieve their goals. For example, by automating certain tasks or processes, generative AI can help lawyers to save time and effort that might otherwise be spent on more mundane or repetitive tasks. This can allow lawyers to focus on more high-level, creative tasks that require their expertise and judgment. Additionally, by using generative AI to generate ideas or suggestions, lawyers may be able to more easily come up with new solutions or approaches to problems, which can help them to be more productive and efficient in their work.
Nonetheless, Generative AI is not a substitute for human judgment and expertise. The proper framing of the technology is not as a replacement for the legal professional. Lawyers must carefully review and verify any materials generated by AI to ensure that they are accurate and appropriate for the specific legal matter at hand. Additionally, there may be ethical and legal considerations to keep in mind when using generative AI in the legal field. For example, is the information submitted to the AI tool kept private? And, even if it is kept confidential, is the data used for machine learning training purposes? Who owns that data? Is it possible that the information could be intermingled with and disclosed in other users’ accounts? These are all important questions that must be asked and answered.
How to Get the Most Out of Generative AI?
Preconceptions can significantly affect a person’s ability to make use of technology, as they can influence how the person perceives and understands it, and how they choose to use it. For example, if a person has a preconception that a certain technology is difficult to use, they may be more likely to avoid using it or to use it in a limited or ineffective way. On the other hand, if a person has a preconception that technology is easy to use or highly effective, they may be more likely to use it more frequently or to make full use of its capabilities.
Whether or not Generative AI can be useful to you depends, in part, on how you think of it. Think of it less like a Star Trek replicator that perfectly creates whatever you ask of it on command. Think more: R2-D2, a loyal companion, copilot, and helper who is often depicted as being instrumental in helping the heroes to accomplish their goals. When viewed through this lens, it is easier to align one’s expectations with Generative AI’s strengths and weaknesses and enjoy greater success.
Futurist John Schaar once said: ‘The future is not some place we are going to, but one we are creating. The paths are not to be found, but made, and the activity of making them, changes both the maker and the destination.’
What are you going to create? How would you best put Generative AI to use? With the right technology and mindset, anything is possible.
Make an AI action plan: Get together with your team and make a list of routine tasks that create friction in your day-to-day operations. Prioritize the list from greatest financial impact to least and gather everyone’s input to break down the tasks step-by-step. Then, make a second list of tasks that require creative input to complete and the value of the output associated with each. With this document, you now have a menu of innovation ideas that can be tackled with AI. You just need the right tools to make it happen.
Curious how to make Generative AI work for you? LawDroid Copilot is an AI legal assistant that utilizes GPT-3 to collaborate with lawyers to accomplish more with less. Or, if you or your organization would like to create your own custom AI-powered workflows or virtual assistants, you can roll your own with LawDroid Builder, an intuitive, no-code AI development platform.
[ This is a guest post for AL by Tom Martin at LawDroid.
Main pic: DALL-E generative AI image on the theme ‘Legal AI in the style of the artist Kandinsky’. ]
P.S. I am on sabbatical, but up until this Friday you will see a handful of articles that were written in 2022 and over the holidays.