0
0
Check out the Squall’s original AI story from January 2023!
The development of artificial intelligence continues unimpeded and has made an unprecedented amount of progress since our first story in the January issue
WRITTEN BY BROCK McINTYRE
Artificial Intelligence: the talk of classrooms, businesses, and the Internet alike. Since our January issue, the advancement of this controversial technology has gone on uninterrupted. With the limited release of GPT-4, AI technology has suddenly gained many new capabilities.
There are several new features that come with this long-awaited update, such as the ability to analyze more than just text. For example, GPT-4 has the ability to analyze images, videos, and even websites; the software now has full access to the Internet, unlike its previous models, though it is worth noting that it has also become a paid service. One of the major problems of the previous iteration of GPT (generative pre-trained transformer) was the fact that it did not have up-to-date information or direct access to the Internet. This has changed with GPT-4, but whether this is a good or bad thing remains to be seen.
With full access to the Internet, artificial intelligence possesses infinitely more power, potentially a dangerous addition, especially when it comes to unethical use in the classroom. One of the issues with GPT-3 was when it came to artificially completing assignments, it was unable to cite sources. When a user would ask it to give a source, it would provide a link, but that link would not lead anywhere. Now it can and does provide active source links. The exponential increase in capability with the release of GPT-4 is a double-edged sword. It has the power to be beneficial to students for learning opportunities, but could also certainly be used unethically to cheat, even more so than GPT-3.
ROGUE ROBOT: The moment of disaster where “Nothing, Forever” went wrong and the AI-generated Jerry Seinfeld began making jokes deemed as transphobic during his stand-up routine.
There are many possible outcomes for the aggressive rise in artificial intelligence. In the past, the technology has been used for great achievements, but it can also backfire. For example, with the Twitch TV show, “Nothing Forever,” covered in our last issue, a parody of Seinfeld that AI could keep running forever, it’s programming did backfire and the characters said some controversial remarks. The live stream running the show was shut down briefly after these controversial remarks; it has since been canceled.
The Squall went back to the teachers interviewed in our first issue to see if they had changed any of their policies on the use of artificial intelligence or changed any of their teaching methods. The use of this technology has been a hot button issue with teachers, and there is still much debate about whether or not it is useful or appropriate for use in the classroom.
English teacher Zach Lindke informed The Squall that he has not changed his methods of teaching as he already utilized in-class assignments as a way to limit outside influence on student work, though he is just as concerned about the faith that society might place in artificial intelligence.
“If I want to know what a student actually knows and can do, I have and will continue to have them verbally tell me or write with paper and pencil. By far, this is the most authentic way to know if a person knows something,” Lindke said, explaining his current method that naturally limits the use of AI technology.
He mentioned the effectiveness of AI detection software, namely through Turnitin, explaining that this had proven to be a reliable method of recognizing academic dishonesty, potentially even effective enough to largely solve the problem at the time being. This sentiment was shared by AP/IB coordinator Debora Marsh, but with the acknowledgement that the accuracy of such tools was always subject to change due to the speed of technological development.
TEST TAKER: This graph displays the results of various exams given to both AI models of GPT-3.5 and GPT-4, demonstrating significant improvement. Could the turing test be next?
Marsh illustrated how prevalent the issue has become within the school, specifically within the English department. Keeping up with and developing an ever-changing response to the advancement of AI is frequently a point of conversation within the department, and the potential that AI possesses has become omnipresent in teachers’ minds.
English teacher Patrick Stolkey for example, in response to what has changed in the last few months, expressed that there were “no changes yet besides doubting the originality of everything not written directly in front of me.”
Clearly, the inherent issue with the potential artificial intelligence has in education has taken a toll on how teachers will conduct classes going forward and will undoubtedly be a contentious topic for the foreseeable future.
“I hope I’m wrong, but I think technology is creating things just to see if they can without considering the implications,” Lindke said. “And once it’s out there (which it is already), it’ll be too late to put it back.”
The world has not changed much since the end of our last issue; we’re only three months down the road, but it already seems as though we’ve jumped years ahead in AI growth. Artificial intelligence still shows no signs of slowing down and continues to shock society with its progress every day. Today, there is even more potential for great good or great evil, for now we are standing even closer to the edge.
