Who Can Write a Better Brief: Chat Ai or a Recent Law School Graduate?
| Jurisdiction | United States,Federal |
| Pages | 16 |
| Citation | Vol. 52 No. 7 Pg. 16 |
| Publication year | 2023 |
DEPARTMENT | THE INQUIRING LAWYER
Part 2
BY RONALD M. SANDGRUND
ChatGPT reviews "Who Can Write a Better Brief: Chat AI or a Recent Law School Graduate? Part 1," 52 Colo. Law. 24 (July/Aug. 2023):*
"Can Chat AI replace lawyers in writing legal briefs? The jury is still out."
"Chat AI writing legal briefs—because who wouldn't trust their legal defense to a machine learning algorithm?"
"When the law meets Chat AI, the pen may be mightier than the programmer."
* The actual query put to ChatGPT-3.5 was: "Create a pithy quote introducing a bar journal article about AI chat programs writing legal briefs."
This is the tenth article series by The InQuiring Lawyer addressing a topic that Colorado lawyers may discuss privately but rarely talk about publicly. The topics in this column are explored through dialogues with lawyers, judges, law professors, law students, and law school deans, as well as entrepreneurs, journalists, business leaders, computer scientists, programmers, politicians, economists, sociologists, mental health professionals, academics, children, gadflies, and know-it-alls (myself included). If you have an idea for a future column, I hope you will share it with me via email at rms. sandgrund@gmail.com.
This two-part article examines whether lawyers will soon be replaced by machines and, more important, whether The InQuiring Lawyer's days as a columnist are numbered. In part 1, The InQuiring Lawyer spoke with Professor Harry Surden, a nationally known law professor, experienced software engineer, and expert on the intersection of artificial intelligence (AI) and legal practice. Also weighing in was ChatGPT-3.5, an artificial language program. This part 2 features The InQuiring Lawyer's version of a battle rap, enabling readers to compare the wit and wisdom of The InQuiring Lawyer with that of ChatGPT through a series of humorous essays about lawyers.
Participants
ChatGPT is a computer program. Output from both ChatGPT-3.5 and ChatGPT-4 is included in this part 2. ChatGPT-4 is now available as a subscription service.
The InQuiring Lawyer is a human being. Or claims to be.
The Rules of the Game
Just as part 1 of this article went to the press, another highly publicized news story arose from lawyers filing legal briefs containing fake cases coauthored by ChatGPT—this time right here in our own backyard, Colorado Springs.[1] Isn't a lawyer's prime directive to read every cited case? How does one read a case hallucinated by a machine?
Some significant issues not discussed in part 1 are the copyright and related intellectual property implications of generative AI. Fortunately, this very topic is the subject of a three-part series appearing right now in Colorado Lawyer. The series kicked off in the July/August issue with a high-level overview of generative AI and its potential legal implications.[2] The second installment, featured in this issue, discusses risks to AI end users, including attorneys. Stay tuned for the series conclusion next month.
Since part 1 was published, I've struggled to keep up with developments in the world of chat AI. One of the more interesting articles I read featured a lengthy interview with Sam Altman, the CEO of the company that created ChatGPT, OpenAI.[3] Altman is a really smart person who seems to have great insight into the risks of AI, including chat AI. He warned a youth group to expect the future to happen "faster than the past," and discussed industry efforts to monitor the development of chat AI and establish guardrails. The Alignment Research Center, for example, is evaluating whether "new AIs are seeking power on their own"—which is apparently within the realm of possibility. To test chat AI's limits, one application was programmed "to gain power and become hard to shut down."[4] The app began to interact with websites and write new code. In an attempt to evade a CAPTCHA screentest on one website, it claimed to be vision impaired, recognizing that if it answered truthfully, it might not have completed its assignment. Additionally, some AI companies are creating internal "red teams" (i.e., groups who emulate a potential adversary's attack or exploitation capabilities) but are concerned that the AI might "understand that they are being red-teamed for risk, and hide the full extent of their capabilities."[5] I'm starting to feel a little more sympathy for those lawyers who were fooled by ChatGPT!
Meanwhile, Altman revealed during an earlier interview that he had ""'guns, gold, potassium iodide, antibiotics, batteries, water, gas masks from the Israeli Defense Force, and a big patch of land in Big Sur'" he could fly to in case AI attacks."[6] Hey, what ever happened to mad scientists having skin in the game?! Meanwhile, Bill Gates has said that ChatGPT was as "fundamental an advance as the personal computer or the internet," while Google's CEO has said that "AI would bring about a more profound shift in human life than electricity or Promethean fire."[7] Uh oh.
Okay, okay, let's step back from the precipice and focus simply on how concerned we should be about AI displacing us in the legal marketplace—or appearing in the next Netflix comedy special. To answer this question, three variations on the same theme are offered: (1) a short, allegedly humorous essay authored by The InQuiring Lawyer for the Denver Bar Association's The Docket magazine in 2004; (2) ChatGPT-3.5's take on the same theme, in response to the query: "Write a humorous essay of between 600 and 700 words imagining what the world would be like if we killed all the lawyers"; and (3) ChatGPT's second response to the same query—this time using the improved ChatGPT-4.
In the end, you should come away with a pretty good idea of what chat AI can (and cannot) do in 2023. What might 2024 offer? That question is keeping more and more people awake late into the night.
Essay #1: A More Pleasant World, by Ron Sandgrund
Someone said, "The first thing we do, let's kill all the lawyers," so we did. Since the judges were lawyers too, we killed them as well. And the world was a much nicer place. For a while. However, after a few months, a few cracks appeared in the veneer.
First, no one arrested was getting tried. Jails started to overflow. Eventually, "lay" prosecutors were appointed, but they refused to plea bargain any cases with the evildoers, and our courts, now staffed by "lay" judges, became so backlogged that, even when the "speedy trial" requirement was expanded to five years by our "lay" Supreme Court, 99% of those arrested had to be released. Eventually, a "presumption" of guilt was adopted, which greatly speeded up trials, but which resulted in the closure of all sports facilities, stadiums, and schools so we could house the wicked. Being appointed a "lay" prosecutor guaranteed a most comfortable life—for a while—until we killed all the "lay" prosecutors because they started to look and sound like lawyers.
Unfortunately, commerce ground to a standstill. For a while, businesses...
To continue reading
Request your trial