Representation Learning: A Deep Dive into Modern Deep Learning Techniques
I have to give you a fair warning before you go any further. I am not a qualified Data science expert. I am writing this blog out of my sheer passion on this topic. I currently lead a team of driven individuals to deliver on cutting edge Generative AI powered products to improve various tracks in a software shop. While I’ve been gaining my skills in the area of Generative AI, I cannot shake off the fact that, anyone who thinks that Generative AI solutions alone would single-handedly save their future, are misinformed. I believe that the as the world ride the hype curve with various LLMs, they are creating products without understanding the foundational architecture of the Large Language Models and couple use-cases that are not meant for these models and build a whole universe around how to change the foundational nature of these models. For example, I have seen architectects proposing how to make a Generative AI model deterministic. These efforts, will ultimately lead organizations to write-off many of the solutions they are building currently or will build in the near future in about few years.
Generative AI is at the peak of its inflated expectations in my opinion and I feel strongly that organizations that will bet on Generative AI exclusively to solve all of their problems will soon realize how much they will regret…
