Recently, our school, one of the best in Cambridge in Visakhapatnam, had a stimulating seminar from Prof.Rama MohanRao Kotagiri, being one of the Fellows of Australia, both of the Australian Academy of Science and the Australian Academy of Technological Sciences & Engineering. Having spent over three decades from 1980 up to 2020 at University of Melbourne, Prof Kotagiri provided us with a detailed look at the two revolutions of this century called AI and ML and in a nutshell told us much of their transformative potential.
The seminar started with Prof. Kotagiri, speaking about the exponential growth of data because of the internet and the web and many other automatic devices that generate data. He stressed that data eventually becomes information and then knowledge through complex processes and powerful computational models. A key highlight was the description of such path-breaking AI models as AlphaZero and AlphaGo, which demonstrate how AI can challenge human limitations, playing games like chess and Go against themselves and defeating world champions. These examples illustrated how AI capabilities are not only innovative but can surpass human expertise in specific domains.Our school, known for having one of the best curricula in the region, was particularly proud to host such an enriching session that emphasized the depth and scope of these advancements.
One of the most interesting points that Prof. Kotagiri raised was regarding AlphaFold 3, an AI system that changed the game in protein studies. He explained how AlphaFold 3 could predict protein structures with astonishing accuracy, helping scientists understand the interactions of molecules that were fundamental to life. It was a technological development that received a Nobel Prize in Chemistry in 2024 and shows the profound influence of AI on fields such as medicine and biology.
Another interesting topic covered was Neuralink, a company which is working to help paralyzed people communicate by allowing them to control devices with their brain activity. Prof. Kotagiri mentioned that the future of Neuralink would be to enhance memory, cognitive abilities, and even motor, sensory, and visual functions, while also curing neurological disorders. This is an example of the actual application of AI in the quality of life for many people.
The seminar also highlighted the influence of such LLMs as GPT- 4. Prof. Kotagiri mentioned that GPTs can translate and process over 160 languages and even create human-like text based on the input they receive. However, he also warned that though such models are impressive, they are not infallible and sometimes give incorrect answers. The takeaway was that even the most complex AI systems err, and it is natural to err-something even experts experience.
Prof. Kotagiri presented valuable insights on how Machine Learning functions. The basic steps include the gathering of training data, choosing the decision and loss functions, specifying the goals, and then training with SGD. The concept is that of finding a function f which connects an input x to an output y, as demonstrated with simple examples like linear regression.
While he was going deep into the neural networks, Prof. Kotagiri showed us the strength to learn complex features, act like highly non-linear predictors, and serve as a parallel network of logistic regression classifiers. The professor also explained the concept of backpropagation, which enables efficient computation of gradients, a core part of current machine learning.
In the concluding session of his seminar, Prof. Kotagiri was enlightening and inspiring. He stressed that while AI continues to grow rapidly, integrating domain knowledge with AI remains a challenge. Nonetheless, the potential for real-world problem- solving through AI is immense. To thrive in this evolving field, he emphasized the importance of mastering the basics: computing, mathematics, physics, chemistry, and languages. The more languages one can speak, the better one’s brain will function. Prof. Kotagiri's words left us understanding that the future of AI bodes boundless opportunities, and we must be well-prepared to harness them.