View profile

On Experimentation and Finding Your Niche

On Experimentation and Finding Your Niche
By Mukul's Newsletter • Issue #10 • View online

Hello everyone,
Since I last posted, there’s been a shake-up in the content I’ve been making. For one, this newsletter is now hosted on Substack! Hopefully the likes/comments are less friction than hitting reply on these emails, and it would help me a lot in figuring out what resonated with you.
In this newsletter I want to bring you along this journey (with YouTube and life more generally) that I’m on as well as update you on the channel. Think of this format as behind-the-scenes with hopefully some advice sprinkled in?
What content should I make?
I’ve been thinking a lot recently about the kind of content I want to put out, and at the centre of this comes two conflicting philosophies. I’d love to hear your thoughts on where you lay on this - click that <reply> button!
Competition is for Losers
First, there’s the whole Peter Thiel “Competition is for Losers” or zero-to-one philosophy. If you haven’t come across this, it argues that you should find a contrarian belief, one that isn’t popular yet but you think the world needs, and then monopolise that niche. For me, I think that trustworthy machine learning isn’t something the broader machine learning community is talking about, hence why I originally started my channel.
Competition raises your game
On the other hand, this whole “niching down” too early can artificially limit the content you put out, especially if you’re like me and are a generalist / still discovering your interests. Last year I was interested in programming language design, and this year it is trustworthy machine learning. With my full-time role at Facebook coming up, I anticipate I’ll be spending time learning more about design and software engineering, and later down the line potentially startups.
My counterpoint to “competition is for losers” is that the act of competing can force you to raise your game. Staying in the comfort zone and just making technical content feels limiting and I kind of need that personal growth to sustain motivation. In the context of this channel, this means if I make software engineering videos, then because that space is crowded I’ll have to bring my own twist to the videos and add significantly more value.
My latest video!
With my latest video, I’m covering the crowded topic of internship advice, but using my unfair advantages (my high performance rating) to provide more value than the generic “How I Landed a FAANG Internship” video, which in my eyes all devolve to LeetCoding.
To up the quality of the video, I’ve also got myself a lavalier mic to improve audio, and improved the editing to make it more engaging.
How I Completed My Facebook Internship In 4.5 Weeks
How I Completed My Facebook Internship In 4.5 Weeks
What does this mean for the channel going forward?
Going forward I’m going to experiment more and try to be guided by making videos where I have a specific life experience (e.g. performing at the top of my year at Cambridge) or where I feel I can explain the concepts significantly better (like PyTorch). There you are, there’s two videos that are coming out over the next couple of months 🤫.
My view is that I’ll eventually find that niche to monopolise, but for now I’m still in that learning and growth phase where I’m still figuring it out! Being a small channel gives me that lightness to fail and learn.
Life Update
The last few weeks were a mad dash to submit my Master’s dissertation at the University of Cambridge. I’m happy to say that’s all done!
I now have about a month before I graduate to unwind and spend more time making these newsletters and YouTube videos. I’m hoping that this means the quality of the YouTube videos really take a step up, as I’ll have more time to edit.
Channel Update
We’re nearing 500 subscribers after around 4 months. The numbers and subscribers will, of course, take care of themselves as I push myself to make better content, but I’m excited to see where this goes.
I’m also going to make some content about the Flower federated learning framework. The team working on this reached out to me after my federated learning video, so I’ll also be contributing to the framework this month.
Lastly, thank you for your support so far. I made a video updating you all about this change of direction, and the response has been overwhelmingly positive. I had a worry that maybe this change of direction wasn’t what people signed up for originally, so thank you for sticking around!
CHANNEL UPDATE: Branching out with Future Videos + Responding to Your Comments
CHANNEL UPDATE: Branching out with Future Videos + Responding to Your Comments
Hope you’re having a lovely week wherever in the world you are, and let me know if these musings resonated with you. Perhaps this is the format the newsletter should follow from now on?
Till next time,
Mukul
Did you enjoy this issue?
Mukul's Newsletter

I write about my software engineering journey at Cambridge and Facebook and share technical tutorials on machine learning, compilers and beyond.

If you don't want these updates anymore, please unsubscribe here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue