ChatGPT has ignited a level of enthusiasm in me for a technology like never before, and I'm eager to share my personal story of why.
Growing up in Surat, a town in Gujarat, within a predominantly Gujarati-speaking community, English was not a language that came naturally to me. My parents recognized the importance of English education and wanted the best for me, but circumstances forced them to enroll me in Gujarati medium schools from an early age. When I switched to St. Xavier's, a private school, in the 3rd grade, the sudden exposure to an English-only classroom was a major shock.I distinctly recall feeling anxious about going to school because the very first class of the day was always English. As expected, yet shockingly, I received a zero on my English midterms because while my classmates were already reading stories and singing songs in English, I was still struggling to learn the basics. It was a humbling experience, but my mother and I took it as a mountain to climb.
We decided to work hard on my English skills, and by the 7th grade, I was excelling in the language albeit just in classwork. I wrote multiple essays, even though only one was required for the exam, to prove that I was a legitimate English language student. However, the real test came in college, where I had to use English not just as a subject, but also as a medium of instruction. Until then, I had learned math, physics, chemistry, and biology in Gujarati, and my conversational skills were practically non-existent. I even enrolled in a class for "Spoken English," hoping for a soft landing in college, but that wasn't the case. I found myself struggling to communicate with scores of people who didn't speak Hindi, and I had to rely on a technical dictionary to translate Gujarati terms to English. I even blanked on simple words like "numerator." It was a difficult time, but I immersed myself in the language, reading fiction books, watching Hollywood movies, and practicing speaking whenever I could. By the time I graduated, I felt fairly comfortable with my English skills, although not completely confident.
As someone who had to work hard to learn English, I always appreciated technology's potential to bridge language barriers. However, it wasn't always smooth sailing. In the 1990s and 2000s, speech-to-text tools were notorious for their inaccuracy. Even if you spoke slowly and clearly, the software would often misinterpret your words, leading to hilarious and sometimes embarrassing consequences. Autocorrect wasn't much better either. It would often change words without your permission, leading to awkward misunderstandings. Even until recently, natural language understanding was still a challenge for technology. While Alexa devices became omnipresent in many households, their use cases were still relatively limited. Sure, you could ask Alexa to set a timer or play music, but it wasn't always reliable when it came to more complex tasks. It couldn't always understand context or nuance, and it was often frustrating to use. For example, some users have reported that when they asked Alexa to turn off the lights, it instead played a song by the band "The Lights". Others have experienced frustration when asking Alexa to order a specific item, only to have it misinterpret the request and order something completely different. Anyone who has used a smartphone has likely experienced an autocorrect mistake that changed the meaning of their message entirely. For example, if you're trying to write "I'll be there soon," but your phone autocorrects it to "I'll be there son," the meaning is completely different and could cause confusion for the recipient. Imagine you're trying to text your friend that you're going to be late for dinner because you're stuck in traffic. Instead of typing "I'm stuck in traffic", your phone's autocorrect changes it to "I'm stuck in tragic". This can cause confusion and make your message difficult to understand. In this case, the autocorrect feature failed to recognize the correct word based on the context.
Pursuit of Mastery
Despite the challenges I faced in learning English, I was determined to master the language. In college, I immersed myself in the language by reading both fiction and non-fiction books. I devoured classics like To Kill a Mockingbird and 1984, and I also read books on science, technology, philosophy and business. I found that reading was not only a great way to improve my vocabulary and grammar, but it also expanded my worldview and enriched my imagination. But reading alone wasn't enough. I also wanted to be able to write in English with confidence and clarity. So, I started writing screenplays and producing short films (, , ) with my friends. We would spend hours brainstorming ideas, writing scripts, and filming scenes. It was a lot of work, but it was also a lot of fun. Through this process, I learned how to tell stories in English and how to express myself creatively. And now, with this blog, I'm continuing on my journey to mastery over the English language. It's a chance for me to share my thoughts and experiences with others, and to practice my writing skills. I'm proud to say that this blog has a growing audience of 1500+ readers, and I hope to reach even more people in the future.
The Transformer architecture, which was first introduced in 2017, revolutionized the field of natural language processing by providing an efficient way to process sequential data with long-range dependencies, and paved the way for the development of large language models like GPT-3, which powers ChatGPT. The self-attention mechanism of the Transformer enables the model to focus on relevant parts of the input sequence, allowing it to learn complex patterns and relationships in language, and has led to breakthroughs in tasks such as language translation, question answering, and text generation. It represents a breakthrough in artificial intelligence and natural language processing that has the potential to transform the way we communicate, learn, and think. It's a testament to the power of human ingenuity and innovation, and it's changing the way we see ourselves and our place in the world. In a way, my coming of age has parallels with the breakthroughs in language models like ChatGPT. Just as I struggled with language learning and overcame the obstacles through perseverance and hard work, the developers of ChatGPT and other language models have had to overcome tremendous technical challenges and push the boundaries of what was once thought possible. And just as I continue to learn and grow in my mastery of the English language, language models like ChatGPT will continue to evolve and improve, opening up new possibilities for human communication and understanding.
So, as I reflect on my journey with language and my experience with ChatGPT, I'm filled with a sense of awe and wonder at the incredible potential of human innovation. And I cannot wait to witness the new frontiers that large language models will open up in the years ahead.
My personal favourite so far. Found the situation where you got a ) on the midterms so endearing. As always you continue to inspire in your ways.