Data36

Yesterday 9 a.m. to 4 p.m. I worked on my data science project with complete focus.

No distractions.

Deep work.

Just as one wishes to work when reaching for high efficiency. And indeed, I churned out ~300 lines of Python code to crack a very challenging NLP classification problem. The result was good: the model achieved about 90% accuracy. That's pretty good, actually! However, for this specific task, I needed closer to 99%.

So I sat down again, later the same day, around 11 p.m.

And I realized that there is a waaay better solution to the issue! I completely deleted my original script and replaced it with ~50 lines of Python -- which I wrote in roughly an hour.

Now, the model reaches the desired 99% accuracy!

9am-4pm: I worked efficiently.
11pm-midnight: I worked effectively.
There is a difference.

P.S. The funny part is that it felt painful to delete my 9-to-4 code, even though I knew the second solution would be far superior. I wish I had thought of the second solution at 9 a.m., finished it by 10 a.m., and taken the rest of the day off. I guess I need to remind myself that mistakes are just stepping stones. ¯_(ツ)_/¯

Disclaimer: This is a repost of my original story from November 2022. I wanted to share it here again because it was a huge aha moment for me.

My favorite explanation of false positives, false negatives, true positives, and true negatives.

(And since we are talking: who came up with the names "Type I error" and "Type II error" in statistics? Clearly, it wasn't a marketer.)

Throwback to last year, when I was fortunate enough to be invited to consult on this factory's short-term and long-term perspectives for their data science projects. IoT, machine learning, advanced automation—it was incredibly inspiring for me to see that data can go beyond the virtual realm.

MrBeast A/B tested his videos and found that closing his mouth in the thumbnails increased watch time. Surprising... because many YouTubers have previously found that open-mouth thumbnails bring more engagement. While MrBeast hasn't disclosed the exact methodology of the split test, given his epic number of views and the significant ad revenue at stake, it's safe to assume the test was conducted professionally and the results are statistically significant.

The main lessons here are:

1) You run experiments to challenge the status quo -- and win.
People often say you should adhere to this or that "best practice," which is fine initially. However, at some point, you should question these norms and occasionally do the exact opposite of what everyone recommends.

2) What works today may not work tomorrow. Business owners frequently overlook how quickly things can change. Even if you have a process, an SOP, or automation that is currently optimal for your business, you should revisit and experiment with it at least every one to two years.

3) You might think of MrBeast as just a creative genius. Which he is. But he's also highly analytical!
I highly recommend his podcast interview with Lex Fridman. Listening to it, you'll realize the extent to which he relies on data in his day-to-day operations. Building what is arguably the best YouTube channel requires not just creativity but also rationality, analytical thinking, and a data-driven approach.

✌

The months in alphabetical order:

April
August
December
February
January
July
June
March
May
November
October
September

There's a lesson here.....and I'm not the one to figure it out.

Last year, I bought a bicycle ($250 investment) and so I spent at least $500 less on gas or commuting ever since.

That’s a 200%+ ROI in ~1 year!

Better than most crypto yield farms.