Students, staff grapple with AI in school

“Hello, World!”
These are the first words that any computing and code language displays to ensure the system is working as intended. However, some can view it as the system being born into the world, ready for whatever function users desire.
These days, using computers to complete basic tasks is just part of everyday life. However, people are just beginning to experience Artificial Intelligence with emerging services like ChatGPT that can generate natural language text based on a prompt provided by a user that resembles writing by a human. The concept of machines being able to perform intellectual processes that are characteristic of humans — such as the ability to reason, generalize, or write — is something that people are just beginning to understand.
AI was first explored by computer scientist Alan Turing in the 1940s. In his report, “Computing Machinery and Intelligence,” Turing asserted that the power of an AI system can be demonstrated with a series of questions. These questions could range from what moves the AI could play in a chess game, to trying to fool players into thinking that the computer is human in a party game where a human judge must find the sex of both the computer and the human player using information given by the two players.
The so-called “Turing Test” was fairly limited considering the technology of the time. However, as years progressed, the understanding of AI grew further as the technology capable of emulating AI advanced, as did techniques capable of teaching the AI to produce more complex work.
Today, ChatGPT can generate whole essays and AI art generators such as Stable Diffusion or Dall-E can create images that resemble images in various artistic styles based on text prompts provided by a user. The power of these AI systems has prompted some school districts, including Austin ISD, to ban ChatGPT altogether, presumably to prevent students from using them to cheat on assignments.

OnRamps Rhetoric teacher Alan Brooks said he thinks blocking the ChatGPT domain from the district network is an overreaction. “I think it’s a little too early. AISD has already blocked (ChatGPT) into the district level for teachers and students. I think that’s a little premature and a little bit of a knee-jerk reaction to the thing,” Brooks said.
Brooks said he is not as concerned with the use of ChatGPT for cheating as some others are. He said students using ChatGPT to do classwork is not all that different from other types of seeking help that some might consider cheating from siblings, parents or other students.
“Chat programs have been around for a long time, even fairly sophisticated ones, but it’s really not much different than having an older sibling as far as students are concerned or a parent who helps them with some writing assignment they have or even does the writing assignment for them and then the student turns it in,” he said. “And we currently don’t have any real good way to check for that kind of cheating as it is. So in that sense ChatGPT is not anything new. It’s just that it’s more widely available.”

Brooks also crafted an informational PowerPoint for teachers to demonstrate the power of AI generative tools and how to prevent plagiaristic use of those tools. Brooks discussed how these tools are not perfect as well and showcased the AI writing an essay wrong from complex MLA format guidelines. The PowerPoint demonstrated the imperfections of AI art generation as well.
Brooks then concluded the presentation on how teachers can use AI to improve their teaching. The suggestions varied from running essay prompts through the AI to seeing how it compares to students’ work, using the AI to write prompts, and using the AI to create writing examples for students to critique and evaluate.
The Eagle’s Eye surveyed students and teachers to gather the opinions about generative AI programs like ChatGPT.
Social skills teacher Kyle Abel said educators are going to have to adjust to the availability of this kind of technology just like they did in response to other tools that have emerged in recent years, such as programs like spelling and grammar checkers like Grammarly. However, he did express concern about students passing off whole essays written by AI as their own work. “If a student claims the work done by AI is their own, I think it is a new type of plagiarism,” Abel said.
The Eagle’s Eye interviewed a student who admitted to using an AI to write an essay that was turned in as his own work in an English class. The student, who asked to remain anonymous, earned a 100 for the assignment.
“None of us wanted to do it. So we just let the AI do it because it was a meaningless essay to write,” the senior said.
However, the student said he acknowledged that he agrees that turning essays completely written by an AI should be considered cheating and should face consequences if caught.
“If they get caught, yes. Technically, it is plagiarism, so it can be punishable if they get caught,” the senior said.
An anonymous student respondent to The Eagle’s Eye survey said he didn’t fully understand how AI text generators work, but he did believe it was cheating. “I can’t fully answer because I’m not 100% sure on how ChatGPT functions, but the integration of other peoples’ writing just to make your own sound better sounds like cheating to me.”
Akins survey respondents admitted to using AI to write essays and brainstorm ideas to help them with their homework. However, 82 percent of 34 respondents said that they believe it was either cheating or “somewhat cheating.”
Nonetheless, AI is still writing essays, love letters, what-if’s, and much more. The limit is only human imagination. The Eagle’s Eye reached out to Advanced English II teacher Alex Vasquez for a student written example of a critique essay to compare with an AI generated one. The results of what Copy.ai demonstrated, an AI generative writing tool, were identical to the naked eye. However, it had small errors along the way, demonstrating that AI generation tools have a long way to go in terms of funding and research before they’re able to perfectly mimic human behavior.
Big tech companies such as Microsoft and Google see the writing on the wall. They know the potential of these programs and how popular and useful AI tools will become. In fact, Microsoft is investing billions of dollars into ChatGPT and Open AI, and Google is coming up with a competitor to ChatGPT.
On Tuesday, Feb. 7, Microsoft announced that AI will be coming to its search engine Bing. As a result of this collaboration, Bing will give you results without going to each webpage created from a web search. Making it easier to ask questions in a natural language and get answers. By investing millions into OpenAI, the company behind ChatGPT, and becoming an early adopter of AI, Microsoft hopes to better compete with Google, which currently maintains about 84 percent of the market share for web search.
Google executives have recognized the threat that AI technology poses to their dominance and recently announced that it will soon be releasing its own AI called Bard, which is based off of Google’s conversational chat LaMDA technology. However, when Google made its first public demonstration of Bard, the AI provided inaccurate information about the James Webb Telescope, showing that AI technology is not infallible.
Generative AI tech is far from perfect right now. There are plenty of examples found online of ChatGPT making mistakes in the writing that it produces. The technology is designed to create what appears to be natural language text based on the data and machine learning the system does on a regular basis. However, ChatGPT and other similar generative AI systems are not capable of thinking or analyzing the accuracy of what it is writing.
Both ChatGPT and Bard prove that. Microsoft even has a disclaimer that from time to time, Bing may pull up some inaccurate information. So while these AI models are groundbreaking in terms of how search engines will work, they still have a long way to go.

Adan Sanchez and Diego Hernandez also contributed to this story.