😔 I've Been Failing Recently
and what I am doing about it
Good morning friends! ☕️
Last week this newsletter had 400 subscribers. But guess what? We have now crossed the 500 mark! ❤️
Recent growth has been amazing, so thanks to every one of you who reads this newsletter. I enjoy writing every edition and I am glad 500 people somewhat vibe with it!
This week I want to talk about failures and setbacks.
To be honest, this week wasn't my best at work and I was making many mistakes with my code, analysis, and general results.
As a Data Scientist, I like to think I'm good with numbers and at conducting accurate tests and experiments for my machine-learning models. Sadly, I didn’t demonstrate either of those two things in the days just past.
I dwelled on these errors for a good day or two but soon realized that this was not productive. Making mistakes is just part of the learning process. As Coleman Hawkins’ famous quote goes:
If you don't make mistakes, you aren't really trying.
It might be a cliche and a tad cringe but it does ring true. Don’t be afraid of making mistakes and make sure you learn from them.
One example that drives this home is Albert Einstein’s self-proclaimed ‘greatest blunder.’ Yep, arguably the greatest mind to ever live also made a mistake. For those interested, he assumed the universe is static and added a cosmological constant to achieve this in his equations of general relativity. Rather, the universe is expanding, and Einstein's assumption and proposed solution were incorrect.
So, even if you are an expert, you are not exempt from making errors. And if you do, it’s most likely a sign that you are pushing yourself.
When this does happen, I recommend carrying out the following steps:
Acknowledge the mistake and allow yourself to be annoyed for a specific period.
Let anyone know who is affected by the error and what it means.
Set a plan to rectify the mistake or reduce its impact if possible.
Write some notes or retrospectives on why it happened and what you should do next time to limit its chance of happening again.
This is my approach, but of course, everyone is different. The main thing is to not beat yourself up too much about it and make sure you learn from it!
Thanks for reading Dishing The Data! Subscribe for free to receive new posts and support my work.
What’s Been Cooking 🥘
Some tasty stories this week:
OpenAI DevDay - One of the biggest tech events this year was OpenAI’s DevDay last Monday. Some of their key announcements were GPT-4 Turbo, custom GPT, and a GPT store.
Amazon’s Olympus AI Model - Amazon is investing millions to build a new AI model to rival OpenAI. The project is codenamed ‘Olympus’ and is speculated to have 2 trillion parameters, double that of OpenAI’s GPT-4.
Samsung’s LLM Gauss - Samsung has stepped into the AI fight with their GenAI model named Gauss. The model has code, text, and image capabilities.
Weekly Favourites ❤️
🎬 YouTube - Watching Neural Networks Learn. Last week I said I had never watched a better video about neural networks. Well, this one is a close second and by the same creator. Neural networks are just giant universal approximation functions. However, so are the Taylor and Fourier series’, that predate neural networks by at least a couple of hundred years. So, why do we use neural networks? Well, this video answers the question and more! It also fits one to the Mandelbrot set!
📚 Book - Stolen Focus by Johann Hari. Been reading this book lately and it’s truly eye-opening. Being in the tech industry myself, I understand the power of algorithms, but the techniques implemented by big tech, particularly on social media platforms are scary. As a society, our attention is slowly collapsing. For example, the average office worker in the US can only focus on one task for three minutes. I plan to do a longer post about this book!
📝 Blog - Eryk Lewinson. If you enjoy time series and forecasting, like me, then Eryk’s blog is for you! Eryk used to work in qualitative finance and has now transitioned to a Data Scientist. He also wrote the ‘Python for Finance Cookbook’ which is a textbook about applying statistics and econometrics using Python in the finance industry.
(PS: Some links are affiliate links that I get a kickback from with no extra cost to you 😎)