Recently, Shantanu Narayen, CEO of Adobe where I’m currently working at, officially announced stepping down from his role after 18 years. During the Employee Meeting following that news, he mentioned the reason behind this decision: he felt like it was the right time for him to reinvent himself. Of course, there are many practical deliberations behind such a massive-scale decision, but I believe he spoke from his heart that the more personal reason behind it all is this notion of reinvention. This left a strong impression on me, and I couldn’t help but think about it in the context of my career.

I started programming about 6 years ago while doing a non-technical undergraduate degree, in the beginning mostly coding in R for statistical modeling and mathematical simulations. Then I discovered machine learning, fell in love with it, did my first research project and wrote my first paper on deep learning 4.5 years ago. I dreamt about getting into the tech industry, despite my unconventional background, and somehow managed to get a job as a data scientist 4 years ago. During that time, I became so fascinated by the engineering side of things that I pivoted into ML engineering 3 years ago, which then guided my decision to do a master’s degree in computer science instead of data science. 1.5 years ago, I rediscovered my passion for research and applied science, decided to join a research lab, where I got to publish some papers on graph ML. Finally, less than 1 year ago, I achieved my dream of getting a full-time job as an MLE. Yet it feels like I’m just getting to the starting line, and there is still a plethora of adventures ahead of me.

For other people, 6 years is more than enough time to become an expert specializing in something, if they know exactly what they want to do from the start and take a linear route onwards. For me, the first few years of my 20s have been well-spent exploring many different things. Both intentionally and unintentionally, I followed a non-linear, breadth-first search approach that, in hindsight, allows me to not only figure out what I’m passionate about, but also what do not fascinate me. All the steps I took, optimal and sub-optimal alike, are what shaped me into the engineer and scientist I am today, and I am happy that I did try my best to spend my early 20s living with intention. Looking back now, I realized this journey was made up of many (oftentimes courageous) decisions to reinvent myself. That’s why Shantanu’s speech deeply touched me.

What I have to do now is figuring out what reinvention I want to dedicate the remaining years of my 20s to. The context in which this decision is going to be made is so vastly different from the previous two decades of my life, but luckily, I am also now an older and (I believe) wiser version of myself to make the most out of this precious opportunity.

Today, I came across this LinkedIn post by Professor Hui Zhang. The following section particularly resonates with me in every word:

Because AI is changing the value of many things we know how to measure.

It expands access. It accelerates execution. 
It lowers the cost of turning knowledge into output.

In many domains, expertise is no longer the only bottleneck.

What becomes more valuable are the things that are hardest to measure:
✦ Curiosity to find the right problem when the "how" is increasingly available 
✦ Courage to act without guaranteed outcomes 
✦ Judgment to decide what is worth building 
✦ Moral grounding to ensure powerful tools help more than they harm

These are not soft skills. They are the hard skills of the AI era.

The most critical thing, then, is to become more authentically human.

Here’s the post for your full context:

I don’t know whether the next few years will truly lead to a reinvention of myself or not - “life can only be understood backwards”. Until then, I will keep working hard and trust that my curiosity, courage, judgement and moral compass will lead me there.