A dialogue-based artificial-intelligence (AI) chatbot ChatGPT took the internet by storm. The bot can understand natural language and respond in natural language with impressive precision and creativity.
Despite its limitations, it is exciting and scary at the same time. So far, it is the closest demonstration of machines mimicking humans, even getting better than us. That is also the scary part.
While ChatGPT has its limitations, it will learn fast. That is the nature of machine learning (ML) — it gets exponentially better with time as its interactions with humans increase. Many industries are going to be disrupted with this technology. But the first and most obvious one will be education.
Homework essays and coding assessments are now irrelevant. ChatGPT and its successors will churn out personalised essays on any topic. This essay will pass any plagiarism checker because it is, indeed, ‘original’. Given the structured nature of computer languages, chatbots are excellent code generators, code explainers and debuggers. A lot of basic coding can now be easily done by AI.
As for education, teachers must reinvent themselves. They are no longer the experts in the classroom. Till now, it was a search engine like Google that challenged them. But a search engine can, at best, reply to factual questions giving references. Teachers beat search engines by flipping the classroom, using plagiarism checkers and giving open-ended assessments that need original thinking.
AI-based natural language chatbots are much advanced. They will become our very own personalised all-knowing teachers available to us all the time. The chatbot can crunch tremendous amounts of information available to it, draw inferences from them and create persuasive arguments in a personalised manner based on human inputs.
When a student can learn better by interacting with the chatbot, what does the teacher do? He or she will have to teach the purpose of learning, what is worth learning and how to learn. This means transforming our teachers into ‘meta-teachers’ for which they will need to be trained.
The nature of jobs will also change — to training machines and helping them get better. That is where the competitive advantage of companies will lie. This implies creating new datasets as inputs to machines and giving feedback to the machine on the output to help it improve. This will change the nature of employment and add a different type of job category — humans who assist machines to learn and get better than competitive machines.
Just as in every disruptive technology before, there will be denial, scepticism and, finally, acceptance. The ones who adapt early will lead. It is difficult to predict the quantum of jobs that will be generated or destroyed. But every job will definitely integrate machine assistance to get better.
The extreme case would be a role reversal — humans assisting machines.
Our education system should factor in this reality. AI and ML can no longer be a course for students of computer science and engineering alone. Every student needs to learn this field of science, because every profession will integrate it and society will be shaped by it. Any new application, including ChatGPT, may seem exciting for a period and then fade. But we cannot deny the underlying technology behind the application — AI. ChatGPT has just given us a glimpse of what is coming. We need to go back to the fundamentals: what makes us human, why do we learn, what is worth learning and how do we learn. With this shift in mindset, we will assimilate technological changes and use it for our collective good, rather than get overpowered by it.
