As you grow older and have more years of experience than you care to include in your resume, you struggle to find new things to learn in your line of work. I spend a few minutes every morning scanning my LinkedIn feed to see what my peers are talking about, what is going on in groups that I belong to and so forth. Increasingly, I find that people trot out things as if they were completely new, unique and path-breaking when in fact none of that is true. Once they do that, the community blindly promotes their story amplifying that mythology until it becomes an "alternative fact".
Case in point is one of the posts I read today. The author purported to explain the spectacular success of using AI at Snapdeal citing an use case about extracting color palette from an image. The keyword stuffing in the piece was a bit out of control- deep learning architectures and what not.
Needless to say, this has been done for a long time and there is absolutely nothing new here. This is like a person who spends days and weeks editing pictures in MS Paint because they did not know Photoshop existed. One day they find out and proceed to write up their epiphany as something the world needs to know most urgently. Color palette extraction has been done for years by many and code can be found on GitHub.
Is this author who took it upon himself to educate us how Snapdeal recognizes a green sari to be green and identify the colors in a patterned one, expecting the readership to be in awe of the fact they were able to do a simple Google Search and find a working piece of code to do this job? I simply failed to understand why he took it upon himself to write this and further why it was deemed worthy of promotion and re-posting by the community.
All of this makes me want to return to my roots and start to code again. Like one of my managers used to say "Can't argue with working code"
Case in point is one of the posts I read today. The author purported to explain the spectacular success of using AI at Snapdeal citing an use case about extracting color palette from an image. The keyword stuffing in the piece was a bit out of control- deep learning architectures and what not.
Needless to say, this has been done for a long time and there is absolutely nothing new here. This is like a person who spends days and weeks editing pictures in MS Paint because they did not know Photoshop existed. One day they find out and proceed to write up their epiphany as something the world needs to know most urgently. Color palette extraction has been done for years by many and code can be found on GitHub.
Is this author who took it upon himself to educate us how Snapdeal recognizes a green sari to be green and identify the colors in a patterned one, expecting the readership to be in awe of the fact they were able to do a simple Google Search and find a working piece of code to do this job? I simply failed to understand why he took it upon himself to write this and further why it was deemed worthy of promotion and re-posting by the community.
All of this makes me want to return to my roots and start to code again. Like one of my managers used to say "Can't argue with working code"
Comments