Based on the data, ChatGPT by OpenAI (USA) continues to be the most popular artificial intelligence (AI) tool in the world. After 1 year of launching, ChatGPT said it has about 100 million users/week, and according to Google Trends data, ChatGPT is currently at the peak of its popularity.

However, not everyone uses this tool successfully.

Lawyer fined for using ChatGPT

A US lawyer representing a client suing Avianca Airlines was "rigged" when he relied on AI to prepare court documents, according to The New York Times .

The case began like any other: a man named Roberto Mata sued Avianca, claiming he had injured his knee after being hit by a food cart on a flight to John F. Kennedy International Airport in New York City.

When Avianca asked federal judge P. Kevin Castel in Manhattan to dismiss the case, Mr. Mata’s attorney, Steven A. Schwartz, and partner Peter LoDuca of the law firm Levidow, Levidow & Oberman, argued against it and filed a 10-page memorandum of more than six similar rulings.

The cases mentioned include passenger Martinez v. Delta Air Lines, passenger Zicherman v. Korean Air Lines, and passenger Varghese v. China Southern Airlines.

The problem here is that no one, including the lawyers representing the airlines and even the judge presiding over the trial, can access specific data about the above lawsuits.

The reason is simple: ChatGPT "invented" everything.

Attorney Schwartz, with three decades of experience, was remorseful, asserting in court that he had no intention of misleading the judge or the defendant, and admitted to using ChatGPT to prepare his case.

Judge Castel then decided to fine attorney Schwartz, partner LoDuca and the law firm Levidow, Levidow & Oberman each $5,000.

ChatGPT completes exams in law school, commerce

In other cases, ChatGPT has proven smart enough to take tough exams at prestigious US universities. Earlier this year, according to CNN, the chatbot passed the bar exams for four courses at the University of Minnesota Law School and one at the University of Pennsylvania’s Wharton School of Business, both in the US.

To test whether ChatGPT was capable of providing answers to exams, professors at the University of Minnesota tested the chatbot. After completing 95 multiple-choice questions and 12 essay questions, the chatbot achieved an average score equivalent to a C+ student. This means that despite its low score, the chatbot easily passed the four required exams.

At Wharton, ChatGPT performed better on the business administration exam, earning grades of B to B-. In a report documenting the experiment, Wharton professor Christian Terwiesch admitted that ChatGPT "tested very well" and did well on questions about business operations management and process analysis.

However, the chatbot struggled with more difficult questions, and made surprisingly silly mistakes when asked to perform the most basic calculations.