Much ado about ChatGPT

By Artchil B. Fernandez

The fields of education, art, and letters are recently shaken with the appearance of ChatGPT. Alarm bells sounded particularly in the academia when some students were caught submitting requirements written by ChatGPT. Has a new era of cheating begun? With the advancement of software like ChatGPT, is a new age of AI (artificial intelligence) at hand? Is AI about to take over the world with the next level of AI application? These are the broader questions concerning ChatGPT and similar softwares. There is nothing to panic with the arrival of ChatGPT and similar AIs. AIs have been with humanity for a long time. Since the advent of computer age, artificial intelligence has been around and has been used by the people.

People’s reaction to ChatGPT however is understandable since it is a new development in the AI field, particularly in conversational bot or conversational artificial intelligence.  Like anything new, it generates attention and hype and breeds frantic reaction when people are confronted with something they hardly know.

ChatGPT is a chatbot developed by OpenAI, an American artificial intelligence (AI) research laboratory. It was launched in November 2022 and was an instant hit, generating 100 million users in just two months. ChatGPT is built on large language models (LLMs) and is a member of the generative pre-trained transformer (GPT) family.

An LLM is a deep learning algorithm that can handle large amount of data – recognize, summarize, translate, predict, and generate text and content from them. The datasets LLM can handle practically cover all that are available in the internet.

Mimicking human conversationalist is the core function of ChatGPT. The application however is adaptable and flexible. When given a prompt or a question, ChatGPT produces a response based on the given information and how does it fit into the vast amount of data at hand. It can write essays, compose music, answers test questions, make business offers or proposals, play games, and a host of other functions.

ChatGPT4 is the current vogue and is making headlines. A more powerful version, ChatGPT5 is about to be unleashed in the market. The 5.0 version aims to rival human intelligence.

Given the capabilities of ChatGPT, it is not surprising people are using them for various tasks. Students lazy to write papers for example simply turn to ChatGPT to do the work for them. This has created a storm of controversy in the education sector. Educators are grappling with a problem many of them unable to deal with or respond to properly.  Should teachers be worried? It depends on the kind of teacher they are.

Some teachers ban the use of ChatGPT and resort to punitive actions. These are teachers who had a hard time dealing with the development, hence resort to easy solution, force. There are those who chose to ride the wave, incorporating ChatGPT in the learning process, taming the technology instead of taking an adversarial stance.

Lazy and incompetent teachers can be easily fooled by ChatGPT but not the committed and adept educators. In paper or thesis requirement, teachers who simply and uncritically accept the final output can be tricked by their students but not those who put emphasis on the process of paper/thesis writing. Making the students undergo the process of writing, from conceptualization to the constant revisions until the final output is done prevents them from submitting a ChatGPT final product.

While ChatGPT produced text appears glossy, sleek, and smart, a close scrutiny can reveal its farce. Attentive teachers can see that AI generated text is boring, bland and mediocre. They lack substance and feeling and a closer examination will reveal they are merely a tapestry or a collage of ideas.

Like any man-made creation, AIs like ChatGPT are imperfect and prone to “hallucinations.” Hallucinations are fabricated outputs. ChatGPT was asked to produce a list of references on a certain topic. It did create the list, but upon checking the sources, some of them turned out to be fakes or non-existent for ChatGPT merely assembled the words from various sources. ChatGPT cannot distinguish right from wrong so it is predisposed to produce defective and erroneous outputs.

AIs like ChatGPT are not capable of higher level of thinking. That is why they “hallucinate.” Critical thinking and creativity are still the exclusive domains of humans.  Humans can create something out of the blue while AIs cannot since they are dependent on available datesets. Lacking the capacity for critical thinking and unable to define right and wrong, ChatGPT has no way of verifying if the datasets contain wrong information.  The standard adage “garbage in, garbage out” still applies to ChatGPT.

However, ChatGPT can outperform humans in lower level of thinking. Jobs that require low level of thinking are in peril. ChatGPT and similar AIs are likely to replace people in jobs that include customer service representatives, technical writers, translators and interpreters, copywriters, and data entry clerks.

On the whole, the hysterical reaction to ChatGPT is much ado about nothing.  Calculators, the most basic AI used by humans everyday have been around for decades and they have not replaced humans. They aid humans in computations, but the analysis and usage of the output remains with humans. Calculators while fast and can handle huge numbers deliver wrong output if feed with wrong input.

Instead of being spooked by ChatGPT, people should welcome this development for they can make life easier to everyone. Repetitive emails can be done by ChatGPT sparing people of mundane and routinary tasks, thus giving them time to do work requiring higher level of thinking. There is no need to panic, yet, with ChatGPT.