Introducing The Judgment of Z by Zac Shaffer

In the spirit of spring, as the season for growth, renewal, and rebirth, we are refreshing this column to review AI in the legal field. Judgment of Z will seek answers to common questions about AI products and the pros and cons of using each one in your practice.

Every month, I will rank an AI tool or platform on a scale of 1-5, based on a series of different criteria, including but not limited to accuracy, speed, cost, and usefulness. I will put these tools on trial so you don’t need to.

In this month’s edition, we revisit the AI platform: ChatGPT. It infamously made fools out of attorneys in the early days of its release. From creating fake cases to having full conversations with users and seemingly pulling information out of thin air, ChatGPT and the legal profession have a tumultuous history to say the least. But, like everything in life, it deserves a second chance.

ChatGPT uses a neural network architecture that processes natural language inputs to generate relevant, human-like responses. In layman’s terms, it is an electronic librarian with near unfettered access to the internet. It learns, grows, and adapts to every user. You can have conversational-like discussions in almost real time.

There are different subscription levels and services for ChatGPT, with options designed for personal and business use. The costs for each are very affordable. I personally have the “Plus” subscription and pay $20 each month. I incorporate it into numerous aspects of my life, but I mainly use it as a research assistant.

Cost-effectiveness is not even the best part of using ChatGPT. Its value comes from its abilities as an electronic assistant. As I said before, ChatGPT is like having a personal librarian who can quickly gather information. Many legal professionals, including myself, use a simple research approach: Search broadly through databases like Google or similar search engines and then use that research to transition into more specific research, usually with services like LexisNexis or Westlaw. What I have found to be more useful is replacing search engines with ChatGPT for broad research (or at least using both in tandem). The benefit of this method is that it uses ChatGPT’s ability to learn and understand. Compared to search engines that find results based on word references, ChatGPT’s neural network architecture reads the prompt’s sentence structure and searches based on the question being asked.

Another very useful aspect of ChatGPT is the ability to help brainstorm and organize thoughts. As the service uses conversational prompts, it can learn from you and develop additional proposed prompts and investigate further based on the results and flow of the conversation. This is a very useful tool, as you can either focus on specific thoughts or simply dump all thoughts at once into ChatGPT and get assistance organizing thoughts with feedback.

Now, like everything that seems too good to be true, ChatGPT is not perfect. You still need to verify that the results or answers you get are true and accurate. This means you cannot simply rely solely on the responses you get. But, there is not a single AI tool out there whose information you shouldn’t double-check. Just like brainstorming with a person, ChatGPT could misunderstand
the context, the phrasing, or even the sources it should be relying on. But ChatGPT shouldn’t be your “one-stop shop.” If you don’t know the answer to a question, you would ask someone and then look further into what information you were provided. The same is true for ChatGPT.

Overall, ChatGPT’s value exceeds the risk it presents. The risks presented by ChatGPT are no different than citing a case in a motion without verifying its accuracy and relevance. ChatGPT has presented its evidence, shown the value it can offer, and it is now decisive. ChatGPT deserves 4/5 points and I would absolutely recommend this service to others. And that is, the Judgment of Z.