Marginally Interesting - Issue #1
Welcome to the first issue of "marginally interesting" - the newsletter. Let's see how it goes. I'll be covering machine learning, AI, software engineering and how all of that fits together. Why "marginally interesting"? That has just always been the name of my blog.
Everyone Is Still Terrible At Creating Software At Scale – Marginally Interesting by Mikio L. Braun — margint.blog I have a hunch that once people saw the economic potential of software, they started looking for ways to "scale it up" and we haven't stopped searching yet.
This post of mine went to the hacker news frontpage (yay!) and got a lot of attention and interesting comments. It seems to have struck a nerve. If you ever worked on an enterprise codebase, you can probably related to many of the observations in the article.
Three themes emerged in the comments: open source as scalable approach to writing software (scalable in terms of people), the need for better programming languages and tooling, and some thoughts on the software-as-a-city metaphor.
An ex-colleague sent me the following quote I didn't knew before: "We build our computer (systems) the way we build our cities: over time, without a plan, on top of ruins." - Ellen Ullman
(Took me a while to figure it out, click on the Twitter icon to go to the thread)
Technology Review had a great interview with Andrew Ng, previously at Google Brain and Chief Scientist at Baidu, who seems to have taken an interest in helping manufacturing companies to use ML. He has some pretty good insights on the need of high quality data, and how things are different when you only have a small amount of data. In the "small data regime" you can actually discuss individual data points to understand your problem data. He says he would always look for better data than a more complex model.
WeCrashed: The Rise and Fall of WeWork - Wondery - Feel The Story — wondery.com Hosted by Business Wars’ David Brown, WeCrashed tells a behind curtain story of hubris and excess, and explains how this tech “unicorn” crashed from a dream into a disaster.
I had seen that Hulu had a documentary on WeWork, but then someone on Twitter mentioned that this podcast mini-series is actually better. I think I first became aware of WeWork by reading Scott Galloway's post "WeWTF". He is part of the podcast series as well which I over all found very interesting.
That's it for Issue #1, hope you found some of the content interesting!