Simon Eskildsen, an infrastructure engineer, writes on databases, performance, napkin math, and learning, with posts on MySQL, TCP, and neural networks.
Commenters value his napkin math approach to performance and link his neural network from scratch tutorial and his Minimum Viable Airtable idea.
Glad to see this company get more funding. Last year I ported all of our companys excel and access files over to airtable. Currently tweaking with their API to work with a MS-SQL database using amazon Lambda and RDS at the moment.
Their API documentation is fantastic, some of the best I've seen as well. They even have some of the best case studies for workflow management as well, posted on their blog, for setting up a relational database management setup in different use cases.
As a user who's also part of their alpha-test program, its exciting to see all the changes they made this year. They added blocks which lets you add app-like integrations such as forms, kanban, maps integration, page designer (think of mailmerge for word) etc. I personally have only dabbled in about half the things offered here, but its crazy how many things you can do with airtable.
I also wrote a popular excel VBA script for airtable too, for bulk downloading and renaming images found here on their forum. https://community.airtable.com/t/bulk-image-downloader-and-renamer-using-excel-vba/4913/3.
Airtable has lots of potential use cases for startup apps as well, I highly suggest reading this article from a philosophical standpoint. http://sirupsen.com/minimum-viable-airtable/
Great comment! I was inspired by this article https://every.to/superorganizers/how-to-build-a-learning-machine-299655 about Simon Eskildsen, which led me to his blog https://sirupsen.com/, which then led me to create https://juliusrobert.site for myself. It's not a traditional blog per se but it checks some boxes and I have the same quote as you do here as a reason to do this https://simonwillison.net/2024/Dec/22/link-blog/#writing-about-things-i-ve-found
Great questions, I struggled with this part the most when I was learning it.
`.grad` is set by `autograd` when calling `backward()`
Probably the easiest way to understand this is to play a bit with `.grad` and `backward()` on their own, with the first code sample in the `autograd` section [1].
[1]: https://sirupsen.com/napkin/neural-net#automagically-computing-the-slope-of-a-function-with-autograd
Take bricestacey's advice, use RVM. I was in the same boat as you when getting started with Ruby - it seemed like I was always fighting with gems and other dependencies - "there has to be a better way!", I thought. And there was...
I read this article[1] and it changed my life.
[1]http://sirupsen.com/get-started-right-with-rvm/
I read this as napkin math[1] for Kube and thoroughly enjoyed. You can only find the important numbers relative to performance and scaling by trying to accomplish some kind of goal. Benchmarks are mostly bikeshedding.
[1]: https://sirupsen.com/napkin
There's a monthly newsletter dedicated to napkin math. He hasn't made one in a few months for some reason, but judging from his tweets he's working on a new one. You can also practice on the old ones of course. https://sirupsen.com/napkin/