31 Jan ChatGPT3 Will Change College Education! Are We Ready For It?
ChatGPT3! Wow! This new AI that was launched in November of 2022 has created a binary response. You can either marvel at its output or get really scared. And, if you haven’t heard of it, then either you have truly attained some form of moksha because technology advancement doesn’t bother you, or you have been on a month long Vipassana meditation. ChatGPT3 has taken the world by storm. In my conversations with students, I often ask them how they engage with it. One of my students did his entire Hindi assignment using a combination of Google Translate and ChatGPT3. But this is just the beginning — we are yet to see the most innovative uses and workarounds that will emerge from students all over the world.
With the onslaught of technology, we have been redefining what it means to be human. Once we designed machines that were able to outdo human calculations on the chessboard, we looked to the next challenge. We moved to intuitive games like “Go” which require more than the brute force of calculations to human-like intuitive thinking. When the neural network outsmarted the world champion in the game of Go, we moved to the world of art, including writing and design. We thought these were the areas technology could not touch. How wrong were we!
What we see now is the world coming under attack. The advancement in computing power, cloud storage, and availability of Recurrent Neural Networks (RNN) proved to be a catalyst for the advancement of Generative AI. The new age Transformers got trained on three key components Positional Encoding, Attention and Self-Attention and attained current level of sentence structuring that is unbelievable.
It is an engineering marvel that we have neural network transformers that can translate, write prose and poetry and draw abstract paintings. For this article, I have used DALL-E of OpenAI to produce three images giving references of great painters from the last 150 years. GPT3 — trained on 45 terabytes of data — and other AIs will only get better with time. But the question to ask is: what’s so special about Generative AI, like GPT3, and how will it impact education?
There is a lot about language acquisition that we do not know. Scientists are still trying to understand how human brains acquire and understand language. In early stages of childhood, children use random words and string them together to communicate. Grammar and sentence structure develop later on. Acquisition of language is a fascinating concept, so when machines are able to reach almost human levels of language production in a matter of seconds, it is something to marvel at.
It has already raised alarm bells at educational institutions and everybody is scampering to limit its usage or co-opt it in the studies. Canadian novelist Stephen Marche announced in the Atlantic that the college essay is dead! Technology columnist Kevin Roose in his article in the NYT asked educational institutions to co-opt GPT3 and other emerging technologies in the classrooms for a better learning experience.
In the coming days, we will hear voices of concern as well as sober suggestions of how there is no possible way of winding down from technological progress. As humans, we need to learn to live and thrive with it.
Will It Change The College Admissions Process?
The US universities, especially the Ivies and top-tiers, have used essays as a key differentiator to select students for the undergraduate, graduate and MBA programs. Most universities offer a compulsory writing program that students need to undertake to learn writing in general and academic writing in particular.
From our conversations with the stakeholders and our research, it is clear that a few new formats would get introduced very soon for screening students. We might see a new set of rules as early as this year for the admissions cycle of Fall 2024.
What potential new elements can be? In a twisted way, Covid pandemic created a robust framework for online testing. Institutions may introduce on-camera writing on a given set of prompts. Students would be required to compose their thoughts in front of cameras and write their responses. Also, we might see the introduction of asynchronous interviews as a testing mechanism. In all likelihood, it might prove to be beneficial for students as it would require the ability to take writing seriously and work on it over a period of time. It will also limit the scope of getting essays polished from external sources.
There is a feverish pace to identify writing done by bots. Google has been working on it to ensure that bots don’t beat its indexing algorithm. Using AI to detect AI is the other interesting approach by Edward Tian, a CS and Journalism student from Princeton University, by creating GPTZero. OpenAI has also stepped forward and revealed its work in watermarking text generated by chatbots. What we are sure of is that this space will see significant work where technology is going to unleash major tools and education institutions will be playing a catching up game.
What Changes Can Come In College Education?
For the longest time, teaching has been limited to providing information to students at a basic level. It has survived for centuries. Students are herded into a classroom and the teacher, a better informed subject expert, shares the information. This will change and in a decade, it will become an antiquated style of teaching. In some sense, history has to come to our rescue. A socratic style of teaching needs to come back to the classrooms where the teacher plays the role of an enabler to make students self-sufficient in making informed decisions.
A Move From What And Where To How And Why?
There is no way but for the education focus to change from information sharing to analytical analysis in view of this changing landscape. Teachers will be called upon to stop seeing their role as information provider about the historical events like industrial or French revolution but move the classroom discussions to analysis, learnings and prognosis.
Moving Up The Bloom’s Taxonomy Ladder
Despite the wide acceptance of Bloom’s taxonomy for knowledge sharing, the classroom education has remained confined to the first two levels – remembering and understanding. We need to urgently move to the next four levels. We are in a space where an institute’s success depends on its ability to quickly get its faculty to move from the first two levels to Application to Analysis to Evaluation to Creation.
After all the technology progress, for best classroom experience we need to go back to the world’s most famous gadfly, Socrate. To provide the best teaching experience, teachers need to learn the Socratic method of cooperative probing where students can learn meta-reasoning (reasoning about reasoning) and meta-learning (learning about learning).
No change is easy and all seismic changes leave behind a lot of debris. Instead of a gradual weather change, they represent a tornado. It is the job of educational institutions to quickly adapt and possibly evolve as fast as possible to the new paradigm. Most of us don’t like changes, but to survive and grow, we can draw lessons from zen philosophy and be like bamboo. And, the most important lesson that I want to draw from bamboo is to bend and be flexible. We must have deep roots in knowledge and know what it is to be a human. That’s what has got us here and it will be our biggest moat against any upheaval.