Video manipulation has been around for ages. In the last couple of years, things got more realistic with the introduction of ‘Deepfake’. In simple terms, Deepfake is a technology that creates very convincing counterfeit videos and audios of a specific person. Most often, this person is a public figure. Unlike old methods of video manipulation, Deepfake uses a specific realm of AI called ‘Deep Learning’ and hence the name ‘Deep’ fake. The first wide spread appearance of Deepfake videos were in late 2017 when a Reddit user started uploading a video series. And currently, Youtube has hundreds of Deepfake videos. You will even find some dedicated Deepfake channels in youtube like Speaking of AI, DrFakenstein, Ctrl Shift Face, etc. …

A meeting room with glass walls, whiteboard, and half a dozen people. A serious topic is being discussed. Multiple baritones can be heard along with one soprano. Many times, the soprano voice is being interrupted while she is speaking. Her ideas are good. Her arguments are valid. Still, a few are acting as if they didn’t hear her talk at all. She is usually me. The only girl in the team.

Once, over a cup of hot-chocolate, I told my friend about this. She nodded with empathy. She told me that she has faced similar treatment at work several times. “It’s just sexism” she added. I found that hard to digest. I know many compassionate men who support and promote women. …

K-means clustering isn’t the best clustering algorithm out there. It isn’t highly accurate because K-means assumes all clusters to be even in size and spherical in shape. K-means also can’t handle complex geometry. It is specifically useful only for flat geometry clusters. And outliers, if present, can have a major influence on the output clusters. Yet, K-means has been popular for ages and is widely used in marketing and customer domains. This popularity has puzzled me for sometime. And I have now reached a point where I know why.

#1 High Accuracy Is Often Not a Business Requirement for Clustering

In most of the use-cases where clustering is used, the business requirement would be quite straight forward. It would be something like ‘we need four distinct customer segments’ or ‘we need to divide the delivery zones’ or ‘we need to group the images’ etc. In these cases, the business requires only guidance in form of grouped data so that they can divert business processes accordingly. K-means is good enough to segment the data into decent and useful clusters as long as the input data doesn’t have too many missing values. …

Remember the “Kodak moment”?

The ability to capture memories into a picture using a personal device was what Kodak gifted to the world. Unfortunately, Kodak decided to cling on to analog cameras far too long, and eventually lost the game.

Nokia too has a similar story.

In 1979, Nokia created the first cellular network in the world. By the 1990s, Nokia was the leader in the realm of mobile phones. When its competitors started focusing on data and the internet, Nokia didn’t care. The company instead continued to focus on its hardware. Nokia did have great hardware. …

The Sydney summer sun, as usual, was hesitant to leave the sky till 8 pm. For those living in this part of the world, the holiday season meant summer cheer. Our office Christmas party felt very different. Instead of the usual one-big-party, we had separate venues for each team this year due to government restrictions. The team co-ordinator booked a large venue in the beautiful Darling Harbour even though there were only 27 people attending it.

“So, what are you doing for the holidays? ” asked my boss during a conversation that evening.

“Moving to my new house” I replied with ecstasy. …

“I’m pretty sure it costs around 22 lakhs,” I said while talking to my friend about a Mini Cooper model.

“22 what ?” she asked.

This was back in 2014, a few weeks after I arrived in Sydney for the first time. Me, like most people coming from the Indian sub-continent, used ‘lakhs’ and ‘crores’ when talking about large denominations of money. My friend, a migrant from the US, had never heard of those terms. She argued that it is not even an English word cause it’s not used in America. I disagreed. I grew up seeing these terms in English newspapers, in school textbooks, in magazines and books. …

Whenever I read about minimalism, there is an image that comes to my mind. The image of a clean living room. That room has a blue ‘L’ shaped sofa accompanied by a light-colored rug and a round metal coffee table. The coffee table is clutter-free and only has a small decor tray on top. Few tile blocks away stand an elegant white TV stand coupled with small bookcases on either side. The TV and the stacked books add more charm to the room. That entire image feels so comfy that as I am shopping for my new home now, I am choosing furniture that looks similar to this mental picture. …

The pandemic has gifted me a fairly harmless addiction. It’s called binge-watching youtube videos.

From data science to dance classes, youtube hosts contend about every subject on the planet. Of all the niches I watch, ‘morning routine’ videos seem to be my favorite. Of course, youtube figured that out much before I did. Their recommendation system gifted unnecessary encouragement to watch more. I’ve watched so many ‘morning routine’ videos to date that I can predict every move of the vlogger when I see a new ‘morning routine’ video now.

It all starts with waking up extremely early.

I do believe that waking up early is good in many ways. It makes us healthier and more productive. However, my understanding of the word ‘early morning’ used to be after or at least during the sunrise. 54 % of the morning routine videos have vloggers waking up between 4:30–5:30 am with justifications that are hard to digest. It makes me wonder if they start sleeping extra early or simply avoids a good slumber. While the former might affect their social life, the latter harms health in the long run. …