Introducing Baseweight

Cover Image for Introducing Baseweight
Joe Bowser
Joe Bowser

It's early November, and it's been about four months since I officially left Adobe. Over the past four months, I've managed to do the following:

  • Take the family to Formula E and Overland Expo PNW
  • Hike 200 miles of the Pacific Crest Trail
  • Take the family on a road trip to Glacier National Park in the US
  • Incorporate and start a company

The last thing is what I'm going to talk about. In August, I incorporated baseweight.ai thinking that I was going to figure out a way to write software independently somehow. While I have a long career, I've really only worked at two different companies, probably due to the fact that I know that I don't interview particularly well. I've only passed a LeetCode evaluation once, and it was at Amazon, and at the time I decided to not take the job since it paid worse than my job at Adobe (and that team at Amazon later got laid off, so I kinda dodged a bullet).

What I am actually good at, is giving presentations. During the Nitobi days, since Nitobi was a software consultancy that ran an Open Source project as a loss leader, we had to all act as sales and give presentations at developer conferences to gain clients. I went to Google IO back in 2009, went to the startup lightning talks that they used to have at Google IO and pitched PhoneGap, and wound up getting interest from Sony, which increased our profile, generated more leads and got us more clients, and eventually got us acquired by Adobe in the first place.

Basically, despite people saying that I should just shut up and code, the writing code isn't the thing that made me successful in the first place, it was actually speaking at conferences, running open source projects, and getting people together to do stuff, which is all more founder skills than they are an individual contributor. Over the past five years working at Adobe, while there were fun moments, it kind of sucked that what I was doing wasn't openly visible. The GitHub contribution indicator showed a five year gap where almost nothing was done except for moments where I was able to write code for a talk.

So, with that in mind, I decided to take a stab at doing this startup thing and I incorporated BaseWeight. It's still very early days, and while I've definitely committed a fair amount of Python code over the last couple of months prototyping random things, I'm not ready to make a release announcement as BaseWeight yet.

Now, as far as my blog, I'm probably going to just keep the blog here, and post everything here. I've taken down the out-of-date instructions on how to build TensorFlow Lite for Android since TensorflowLite is LiteRT, and people should really be using the new Google Play version with the C API so they can write bindings to whichever lower level native language that they're using for Android (C, C++, Rust). It's tempting to do the whole "BaseWeight Blog" thing, but since I'm currently the only person directly associated with BaseWeight as the sole founder, I don't see the point of posting in two places right now and BaseWeight will probably just post content for official releases, which will be coming soon.

I'm going to be posting random technical blog posts about various topics ranging from Electric Vehicles, Computer Vision, Machine Learning and of course Mobile development. While I do a bunch of Web Development, I find that whole space extremely saturated and the fact that you basically have to use some weird Typescript environment to do basically anything anymore makes it very unattractive. That being said, you don't spend twelve years at a company known for design without strong opnions on user experience and design.