ChatGPT: Unoriginal Yet Essential for Human Interaction
Consider Jorge, a teen who is caught with a big amount of marijuana by a school administration and faces expulsion if he is reported to his parole officer. If the administration does not report him, they are breaching the law; if they do, he will be assigned to one of the city's worst schools and will almost certainly recidivate.
This is a case study that we presented to a class of 60 Harvard Graduate School of Education students. We asked them to act as teachers or administrators at the school and devise a plan of action. We supplied them with ChatGPT's research analysis one hour into their talk.
"We must undertake a review of [the school's] existing policies and procedures relating to substance misuse, with the goal of ensuring they are consistent, transparent, and reflective of best practises... The school should be compassionate, but it should also be clear that drug addiction and related offences will not be condoned... This approach should be pursued while ensuring that the school is attentive to its pupils' individual needs, particularly those from low-income and working-class families."
Read Also: How ChatGPT Fooled Lawyers into Citing False Case Law: Unveiling the Controversy
Read Also: 100+ Top ChatGPT Prompts for Every Situation
Our graduate students initially outperformed this chatbot. They, too, were prone to regurgitating the same monotonous rhetoric about justice, equity, and education—discourse that appears appealing but lacks substance, failing to provide a clear method beyond broad ethical aims. According to one student, "we were just saying formulaic, buzzworthy stuff, instead of talking about anything new like we said we wanted to when class started."
The pupils were also noticeably surprised by how similar ChatGPT's ideas were to their own. They expressed their fear that these remedies sounded exactly like what a school would practise. Then they questioned themselves and their capacity to devise answers that were distinct from what others had been doing for so long. They described themselves as being locked in a "loop." One student attempted to de-escalate the situation by rejecting ChatGPT's input as "not really saying anything." "Did we really say anything?" asked another.
However, it wasn't until ChatGPT pointed out the students' lack of imagination that they were able to consider ideas that they, or any automatic language scrawler, would not have considered. They realised that the matter was totally focused on the administrators' point of view, and that their previous debate had left no opportunity for action that included instructors, students, and Jorge as well.
The kids began to challenge the logic and legitimacy of current structures that impact their choices and outcomes, such as schooling and juvenile justice, and began to propose new, more creative ways to Jorge's case. One student quipped that the teachers should all smoke pot with Jorge (in order to become targets for law enforcement rather than remain as innocent spectators). Another advocated for the abolition of schools. A third person mentioned grandparents who destroyed public property in the name of environmental justice. These concepts may appear illogical at first, but anything that disturbs old habits of thought is likely to appear illogical at first.
Students had not only addressed their immediate, conscience-clearing answers in the context of Jorge's case, but also contemplated prospective actions by the end of the conversation. Students came to realise that if sufficient collective power is developed, it is feasible to both respect and defy the law. For example, they may turn Jorge in while threatening to go on strike if he was expelled—acting neither as administrators nor as saviours. Rather than removing schools entirely, this one school should be closed.
Our students were unable to reach an agreement. Nonetheless, the discourse grew more nuanced, innovative, and deep once ChatGPT forced them to address their fear of being unoriginal. Students progressed from repeating other thinkers' words and frameworks (at one point, they even asked ChatGPT to analyse the case in the style of Paulo Freire's Pedagogy of the Oppressed) to confronting the implications of these frameworks and the limitations of imposing them on complex realities.
Read Also: 10+ Best New and Upcoming Google Bard Features
Read Also: What is Google PaLM 2 language model
This was not a simple process. For much of the debate, despair hung in the air, as it often does when AI compels us to confront our own predictability. We also do not disregard the legitimate, well-documented worries that many people have expressed about AI, particularly the fear of mass job loss as much of the everyday cognitive labour that we do can now be performed by technology.
However, we believe that AI language models such as ChatGPT can serve as catalysts in situations where predicted reactions have consistently failed us: climate change, race relations, income inequality, and other issues. They may boost our "productivity," not by offering better answers, but by confronting us with unoriginal and average-of-everything-on-the-internet responses, allowing us to move into the region of fresh possibilities that ChatGPT cannot forecast. ChatGPT, in a manner, serves as accidental satire, demonstrating how inadequate and bland our answers can be. AI will not solve these problems, but it can be used to change the way we think, work, and act by reflecting back to us what we will most likely say the first time and urging us to try again and again.
.png)
0 Comments