LANSING, Mich. (Michigan News Source) – As many professionals weigh in on the appropriate uses and limits of Artificial Intelligence (AI), the CEO of the company who developed ChatGPT testified before Congress advocating for certain protections on Tuesday. 

Senator Richard Blumenthal (D-Connecticut) Chair of the committee opened by stating it would be the first in a series of hearings which would seek to “write the rules of AI.” 

MORE NEWS: Elissa Slotkin Secures Victory in Tight Michigan Senate Race

“Our goal is to demystify and hold accountable those new technologies and avoid the mistakes of the past,” U.S. Senator Blumenthal said. 

He then proceeded to demonstrate some of the capabilities of ChatGPT by asking it to deliver introductory remarks as if it were him to the floor based on previous floor speeches.  

“I am sure that we’ll look back in a decade and view CHATGPT and GPT 4 like we do the first cell phone, those big clunky things we used to carry around,” U.S. Senator Blumenthal added, “But we recognize that we are on the verge really of a new era.” 

OpenAI CEO Sam Altman, the company which created ChatGPT, was one of several who testified before the Senate Judiciary Subcommittee on Privacy Technology & the Law on oversight regarding oversight of AI. 

“We believe that the benefits of the tools we have deployed so far vastly outweigh the risks,” Altman said. “But ensuring their safety is vital to our work. And we make significant efforts to ensure that safety is built into our systems at all levels.”

He also mentioned the development of GPT-4, which arrived in March, praising it as “more likely to respond helpfully and truthfully” all while refusing “harmful requests” than any other model of similar capability. 

MORE NEWS: Michigan Has More Teachers and Fewer Students During Teacher Shortage

“However, we think that regulatory intervention by governments will be critical to mitigate the risks of increasingly powerful models,” Altman continued, “For example, the US government might consider a combination of licensing and testing requirements for the development and release of AI models above a threshold of capabilities.”

ChatGPT has been hailed for its ability to perform a myriad of tasks that could aid students in the classrooms and beyond, and for its earlier version’s ability to perform on national tests. 

The penultimate version of the technology, GPT-3.5, scored in the 40th percentile on the Law School Admission Test (LSAT), 25th percentile on the Graduate Record Examination (GRE) Quantitative, and 10th percentile on the Uniform Bar Exam.  However, the newest version has outperformed its predecessor accomplishing a score in the 88th percentile on the LSAT, 80th percentile on the GRE, and 90th percentile on the Bar Exam according to statistics from OpenAI.  

Studies have also shown that cheating through use of the AI is also on the rise in K-12 schools and in higher education.  In the beginning of the new year, Michigan State University released a recommendation for faculty to use when considering the technology’s potential effects in the classroom written by Dr. Jeremy Van Hof, Director of Learning Technology and Development at the Eli Broad College of Business at MSU originally in an email to faculty. 

“I have been using ChatGPT daily, and it has so far written me custom Python code, told me how to prepare a stuffed Cornish game hen (the recipe was delicious), written a sonnet about roses, and provided a 1000-word essay comparing the philosophical positions of Kant and Descartes with a 5-source annotated bibliography,” Dr. Van Hof wrote, “While its poetry is a bit pedestrian and the cited sources are sometimes questionable, I can attest that the tool is powerful and the content it produces could easily be passed off as the work of an undergraduate.” 

While recognizing that the technology is free and open to students, Dr. Van Hof encouraged professors to adopt language in their syllabus to discourage misuse of the AI. 

“The policy of this class is that you must be the creator of all work you submit for a grade. The use of others’ work, or the use of intelligent agents, chatbots, or a.i. engines to create your work is a violation of this policy and will be addressed as per MSU and Broad College codes of conduct.”

He further demonstrated the capabilities of the AI by including sample syllabus language written by ChatGPT.