Chatbot GPT passes exam at US law school
Microsoft made a major investment in OpenAI’s chatGPT this week
Works with artificial intelligence to generate streams of text from simple clues
A chatbot powered by vast amounts of data from the internet has passed exams at US law schools. Microsoft has invested heavily in AI to generate text stars from simple clues.
He passed the U.S. bar exam after writing essays on subjects ranging from constitutions to taxes to torts. ChatGPT works so efficiently that scientists warn that it can lead to widespread fraud. Furthermore, professors seem concerned that this may interfere with traditional teaching methods in the classroom. Chat GPT
Jonathan Choi, a professor at his School of Law at the University of Minnesota, took the same test as his students to learn his ChatGPT. The test consisted of 95 multiple-choice questions and his 12 essay questions.
Surprisingly, the results were just as amazing as the average student. But in his White paper titled “ChatGPT Goes to Law School,” which was published Monday, Jonathan Choi and his co-authors reported that the bot earned his C+ overall.
So this was a good enough score to pass the exam. One boy ranked last in most subjects and was “bombed” in multiple-choice math-related questions.
According to the author, “ChatGPT showed a strong understanding of the basic legal rules in writing essays and had a solid structure and structure throughout.”
However, the bot “often struggled to spot problems when faced with open-ended prompts, a core skill in law school exams.”
Following recent issues with ChatGPT, authorities in New York and other jurisdictions have banned his use of ChatGPT in schools.
According to Choi, this can be a valuable teaching tool. On Twitter, he wrote: “However, we expect law students taking exams and working as lawyers to benefit from working with people and language models like ChatGPT.”
Additionally, he dismissed the notion of cheating, replying to another of his Twitter users who said he found papers where two of the three markers were written by bots. In another tweet, Choi wrote: