This doesn’t happen with every new client, but it feels good to offer a client half of what they were paying their previous accounting firm and still have an incredible margin.

We made the right call in building an accounting firm instead of making a SaaS for accounting firms.

I started college in 2002. Back then, everyone was worried about software engineering jobs being outsourced. There was talk about there not being any jobs left when I graduated. Why pay a developer $100k a year when you can pay them $20k?

Stop saying nobody knows how to define AGI. Until the $ got huge, there was a consensus in the field. And (see below) the people who coined the term still agree - both on what it is and the fact that we haven’t achieved it yet.

Everything else is noise.

garymarcus.substack.com/p/the-five-s...

One of my pet peeves is people being misleading in order to generate new business. If you're a marketing firm with a name that makes people think you're a VC or a recruiter with a "Director of Engineering" title, I'm unlikely to trust your services.

I'm a big fan of monoliths with every module treated as if it could be spun out of as a microservice.

You end up with slightly slower performance (too slight to notice most of the time), but the code is much easier to maintain and portable.

Kind of wish I could put limit orders on items in my Amazon wishlist...

I wonder if that would make Amazon more money or less.

The classic fear of automation is rooted in the idea of being replaced. This is because we tend to default to zero-sum perspectives.

Automation makes things cheaper, which tends to increase demand for the thing that was automated. That results in more work for humans to support the automation.

I’ve read a ton of stories about how various businesses turned things around by decentralizing decision making. This clip about Barnes and Noble is just the latest...

www.youtube.com/watch?v=Z9kO...

I often see a lot of fear an anxiety around releasing software. There's a tendency to believe that more planning and testing is always better. Yet, there is a point where you hit diminishing returns on planning. At that point, it is time to launch.
beekeycheung.substack.com/p/its-time-t...

Things they don't tell you about being an entrepreneur:
It makes sense to start responding to some of the spa... errr "cold emails" you get.

The conversations I have with engineers who have had to help maintain the infra for the code they write go very differently than with the engineers who haven't.

If I write bad code and someone else has to wake up at 2am, I'm going to be super apolgetic and promise it won't happen again...

Greetings!

I'm Beekey. I'm a partner at an accounting firm. We focus heavily on automation, but as a way to help our human staff instead of replacing them. That also sums up what I think of AI.

I post a lot about software development, organizational management, and technology in general.