6 Disadvantages Of ChatGPT We Should Know

Everything has its own pros and cons. In fact, sometimes that list becomes the deciding factor. Here are 6 disadvantages of ChatGPT.

AI tools were invented for convenience; but there are always downsides to every new tech. Here are 6 disadvantages of ChatGPT that you should be aware of. 

In my last article, I spoke about the pros of using Artificial intelligence tools like ChatGPT. Now, just like how every other technological invention and upgrade has its own advantages, along with it also comes a few disadvantages. Sometimes, they can be ignored, and sometimes, they outweigh the benefits of using such tools.

6 disadvantages of ChatGPT

Technology is invented and upgraded for humanity’s convenience, but a lack of awareness about the tools we use and heavy reliance on tech for cerebral tasks can lead to an inconvenient future.

So, in this list of pros and cons, let me present to you the latter of AI tools like ChatGPT. Here are 6 disadvantages of ChatGPT

Limits thinking capabilities

Indulgence is fine, overindulgence kills. I’m not talking about ice cream.

As a writer, creative blocks are real. I struggle with it for most parts of my days, and in times like those, I had wished for a genie who would come and finish off my articles. ChatGPT might be that genie in those dark days. But not all days are dark.

As adult writers, we may have an understanding of the distinction between assistance tools, and compulsory tools. But, growing up kids might not understand so. An overreliance on AI tools can hamper their imagination and creativity, thereby stunting their natural writing skills.

Sometimes, even for an adult writer, too much dependence on AI tools like ChatGPT can create a false sense of security and overconfidence. In the absence of such tools, one is left panicking and questioning their own capabilities.

Never miss real stories from India's women.

Register Now

Lacks emotional intelligence

This is the trump card that defeats all other factual points in a debate. Despite all the other qualities that make a human distinct from robots, it is the trait of emotional intelligence that is placed at the top. Among all the cons, lack of emotional intelligence is one of the disadvantages of ChatGPT

Keeping aside the fact that even few human beings lack EQ, being emotionally responsive in conversations is a natural human quality, especially in the field of service. AI-created chatbots can only be helpful in situations when the customers are raising a query. But in terms of receiving feedback, it fails to evaluate the emotions underlying the conversation.

It might be trained to generate conversations on its own, but customer service needs more skills than merely generating conversations. Different customers have different complaints and different ways of projecting such complaints. Some are polite, some are, well, not so polite. Managing each customer in a personalized way is a skill.

It can imitate empathy, but responding to emotional cues and reacting appropriately is something that only possessors of emotional intelligence can do. Most of the time.

Threat to human jobs?

Recently, a machine learning engineer named Rowan Cheung shared a screenshot of his interaction with ChatGPT, wherein he had asked the AI tool about the jobs that it can replace. GPT-4, the latest model, came up with a list of 20 human jobs that it can replace alongside the human traits that it can imitate.

The list included jobs like data entry clerk, customer service representative, proofreader, bookkeeper, etc.

Since the very beginning, the entire writing community were apprehensive about the future of their profession. ChatGPT had ceased to be a tool of assistance, and businesses started viewing it as a tool to scale up their business process efficiency and accuracy.

It was hailed, but at the same time, there were cries of worry as well.

But, over time, one has come to accept the fact that even though AI has the potential to replace human employees, it cannot be left all on its own. In fact, while scrolling LinkedIn the other day, I had come across a person, whose name eludes me now, who that ChatGPT is not going to replace us (the writers), but someone knowing how to use ChatGPT can.

So, ironically, you still need a human to supervise even a robot.

Not a human expert

As I mentioned in my previous article, ChatGPT tends to be accurate because it is trained based on gathered data. And the pool of data is increasing day by day. But, ChatGPT can never replace the knowledge of human experts.

Another job from the above-mentioned list is that of a tutor. ChatGPT suggests that it could replace a human tutor owing to its knowledge and teaching skills. But, this is far from being possible. In fact, in a turn of events, news revealed that a man ended up dying by suicide due to eco-anxiety, after conversations on global warming with a chatbot from ChatGPT.

ChatGPT might be able to transfer knowledge, yes. But tutoring is much more than just transferring data. A teacher has years of experience not just in terms of knowledge, but also in how to handle students with different demands. They can also attend to their pupil’s needs, and regulate their teaching styles accordingly.

The education system is in danger if one starts to believe that AI replace human experts.

Could be misused

My best friend’s sister is a professor at a renowned private university in the country. The other day she mentioned how in a graded assignment, an entire class had submitted essays that were directly taken from ChatGPT. Some were tweaked, but more than 40% of the class did not even bother to make alterations. 

This is just an example from the vast ocean of instances that one can produce to cite this dangerous side effect of introducing ChatGPT to students. Even kids from middle schools are misusing this assistance tool via the accounts of their parents. This has come to pose a worrying problem among educators.

And not merely students. Grown-up adults are finding newer ways to cheat on others. Through ChatGPT, online scammers are creating chatbots to send scam or phishing messages that look genuine to the general mass. With unique ways that do not arouse much suspicion, they are extracting huge amounts of money from the bank accounts of people who are failing for their trap.

How is this unforeseen?

Privacy issues

Recently, ChatGPT hit the headlines again. This time for criticisms.

Allegedly, user chat titles on ChatGPT were leaked, which raised questions surrounding user privacy and data security. OpenAI confirmed that due to a bug in the AI, the chatbot had enabled users to view the chat history of others.

This gave unlimited access to sensitive information including payment transactions, addresses, etc. Since ChatGPT is trained to work as a full-working assistant, people can use it to seek hotel or restaurant recommendations, perform online transactions, or book hotel rooms or tables, etc.

Remember, if the product claims to be free, and you are not paying for it, then you are the product. More than often, a phone number is a compulsory requirement to log in to these apps, which is one of the biggest concerns among all disadvantages of ChatGPT!

This is user-sensitive information. It gives away not just the physical location of a user at any point in time, but also tracks the vast social activities that they perform. OpenAI had to take down the version of ChatGPT after this chaos.

Though they apologized for the inconvenience, and later relaunched the bug after having it patched, the threat looms large.

What does the future look like?

Just a few days back, to remain relevant in the field, Google announced its own AI intelligence tool called Bard. It does what ChatGPT exactly does, except it belongs to Google. Encouraged by the competition, Grammarly too launched its own AI writing assistant tool.

It looks like a frantic race, the survival of the fittest (robot). The Microsoft-owned company, OpenAI’s launch of ChatGPT catapulted the growth of AI, thereby diminishing the distance between daily human lives and machine automated responses.

On the other hand, while I was working on this article, Italy announced their decision to ban ChatGPT in their country owing to privacy concerns. And hilariously, major tech luminaries have realized that the field of AI is now what they call an “out-of-control race”.

Is there a way to deal with the disadvantages of ChatGPT?

This comes immediately after the launch of GPT-4 which is a more powerful model of ChatGPT. This particular model has demonstrated the ability to pass standardized tests, build websites from basic sketches, and draft lawsuits.

Some big names together, including Elon Musk, have called forth a 6-month pause on the development of models more powerful than the current one. The tech community also plans to lay down a set of protocols to maintain safety.

But the question lingers, is it going to be another eye-wash like the green-Earth initiative by the global leaders of the world? Or are we going to witness a day when humans and AI would be at war with one another? Or are we on the way to creating Frankenstein’s Monster, bringing in our own doom?

Image source: Diego Cervo, free and edited on CanvaPro

Liked this post?

Join the 100000 women at Women's Web who get our weekly mailer and never miss out on our events, contests & best reads - you can also start sharing your own ideas and experiences with thousands of other women here!

Comments

About the Author

Akankha Basu Roy

The author is a Gen-Z kid who resorts to writing to vent out about the problematic ways of the world. Having majored in Theatre, English, and Psychology, I take a guilty pleasure in complex read more...

52 Posts | 31,603 Views

Stay updated with our Weekly Newsletter or Daily Summary - or both!

All Categories