ChatGPT can only get better from here

Academics are finding that the bot is capable of passing professional exams associated with MBA, the Bar, and the United States Medical Licensing tests. A massive reworking of the educational system seems inevitable

By Chidanand Rajghatta

  • Follow us on
  • google-news
  • whatsapp
  • telegram

Top Stories


Published: Wed 25 Jan 2023, 8:50 PM

Last updated: Wed 25 Jan 2023, 9:41 PM

The fear of an artificial intelligence (AI) takeover, where machines become more intelligent than humans and potentially pose a threat to humanity, is a topic of much debate in the field of technology. Some experts believe that the likelihood of such a scenario is low. Others warn that it is a real possibility that should be taken seriously.

On one hand, the development of AI is still in its early stages, and current technology is far from being able to match or surpass human intelligence. Additionally, the development of AI is heavily influenced by human decision-making, and strict regulations and guidelines are in place to ensure that AI systems are developed and used responsibly.

On the other hand, some experts argue that as AI continues to advance, the risk of machines becoming more intelligent than humans increases. It's important to note that AI is not a monolithic field, there are different types of AI and each one of them has its own implications, for example, in the case of AGI (Artificial General Intelligence) or superintelligent AI, it is thought that the AI systems could be capable of recursive self-improvement and would be able to design and build machines that could be more intelligent than themselves.

Given the complexity of the topic and the uncertainty of the future, it's important to continue monitoring the developments in AI and to have a healthy skepticism about the potential risks and benefits of the technology. At the same time, it's also important to understand that the potential of AI can bring immense benefits to society, such as improving healthcare, education, and environmental protection.

Just so you know, the above paragraphs were written by the AI bot ChatGPT, short for 'Chat Generative Pre-trained Transformer,' in response to my question, "Are fears about artificial intelligence takeover overblown?" Released two months ago by OpenAI, an American artificial intelligence (AI) research lab bankrolled among others by Elon Musk, Amazon, and the Indian IT major Infosys, ChatGPT has become a favourite plaything of tech nerds, who are teasing answers from it on a range of questions ranging from "Does God exist?" ("no scientific evidence either is a matter of personal belief") to asking it to predict its own future.

Most answers to such questions are generic pabulum of the "on the one hand...on the other hand" kind, and how you perceive it depends on the level of sophistication and nuance you expect from it. Some critics have found it has a "devastating sense of humour" and can write brilliant jokes; others found its punning under-par but its ability to contrive rhymes excellent. Science geeks insist that it makes errors in calculation, and poets and philosophers are underwhelmed by its lack of nuance. Assessments vary widely depending on expectation. It is not intuitive – yet.

But these are early days, and no one disputes that it can only get better from here on. At least one US lawmaker, California Congressman Ted Lieu, one of just three members of Congress with a computer science degree, says ChatGpt "freaks" him out. He has called for a dedicated government agency, on the lines of FDA for drug oversight, to regulate AI, a move that ChatGPTs seems to have anticipated and reconciled to going by its own response.

The implications of this advance are already unfolding in the field of education. Students, from elementary school kids to college graduates, are tapping into ChatGPT to write essay assignments, which it does effortlessly and efficiently. One professor was so wowed by an essay on the morality of burqa bans turned in by a student, who admitted to have mined it from ChatGPT, that she now has her class do assignments in school while figuring out how to incorporate the new beast into the curriculum. Academics are finding that the bot is capable of passing professional exams associated with MBA, the Bar, and the United States Medical Licensing tests. A massive reworking of the educational system seems inevitable.

But that could just be a finger in the dyke to prevent humankind from being overwhelmed by AI. The arrival of automation technologies decimated millions of blue-collar employment across the world, leading to greater investment in higher education in the belief that white-collar jobs are relatively immune from technological advancement. Now that even initial iterations of AI bots are dazzling users with their white-collar virtuosity, fears are rising about what they will do to a range of jobs -- from writing and editing to accounting and lawyering and teaching and computer programming.

It may not take long either. While anticipated technologies like nuclear fusion and human cloning are still years if not decades in their practical and commercial application, the timeline for AI advancement is alarmingly short. Sure, humankind adapted to new technologies – remember switchboard operators, bank tellers, and travel agents? – but that makeover took years. Considering ChatGPT, the latest version of which was released only last November, is causing convulsions in the real world every day since its release, the world could be a different place within months.

- The writer is a senior journalist based in Washington

More news from Opinion