iiinova — where your digital self keeps evolving

Dragan Stepanović @dstepanovic.bsky.social · 31/05/2026

Last Week in Collapse open.substack.com/pub/lastweek...

000

Dragan Stepanović @dstepanovic.bsky.social · 31/05/2026

open.substack.com/pub/lastweek...

000

Dragan Stepanović @dstepanovic.bsky.social · 28/04/2026

I feel for all the on-call engineers at this company

140

Dragan Stepanović @dstepanovic.bsky.social · 24/04/2026

Even if the rate of generating customer value increases with more code generated per unit of time, but at a comparably lower rate, you're still increasingly bloating your codebase with useless inventory faster than the value you're able to extract and driving up carrying costs.

a comparison between two embedded circles, where in each case the outer and bigger circle is generated code and inner and smaller circle is generated value. the increase in outer circle is bigger than the increase in the inner circle

111

Dragan Stepanović @dstepanovic.bsky.social · 12/04/2026

Preparing a talk "Agentic coding - A Systems Perspective", and along the way realized that cognitive/comprehension debt didn't arrive with the advent of agentic coding. Most teams doing work in isolation (individually), already heavily experienced it.

242

Dragan Stepanović @dstepanovic.bsky.social · 11/04/2026

Let's not forget that LLMs got fed with an ever decreasing quality of work our industry has been producing over the last 15 years as a result of cheap money with ever decreasing central bank interest rates. And both of these reinforcing loops are accelerating with LLMs dogfooding.

071

Dragan Stepanović @dstepanovic.bsky.social · 30/03/2026

So, here's the thing. If this, someone would say, embarrassing level of availability came as a result of "90% of our code is written by agents, and we (mostly) don't review the code they generate"- fine.

120

Dragan Stepanović @dstepanovic.bsky.social · 13/02/2026

I can guarantee you that addressing an actual bottleneck in the system definitely shouldn't look like pushing elephant-sized inventory through the boa constrictor (coding → review → integration → deployment → etc.). And you can tell by the elephant rolling its eyes.

111

Dragan Stepanović @dstepanovic.bsky.social · 11/02/2026

Full quote from the Lean Enterprise book

010

Dragan Stepanović @dstepanovic.bsky.social · 11/02/2026

I feel this slide from Jez Humble's talks has become more relevant than ever. It's not about generating bigger batches faster, but thinner slices validated sooner.

110

Dragan Stepanović @dstepanovic.bsky.social · 24/01/2026

I was trying to find an image for the change in the dynamics that will likely happen as a result of plummeting of the cost of generating more code with AI, and "elephant travelling through a boa constrictor" nails it. Essentially, the elephant just got bigger.

000

Dragan Stepanović @dstepanovic.bsky.social · 16/01/2026

Oh, I have two more from Eli (Goldratt) on AI:

000

Dragan Stepanović @dstepanovic.bsky.social · 09/01/2026

As long as LLM pricing is based on how much/how often you're allowed to converse with the model, Spec-Driven Development and big(ger) batches will be a normal response to that. Scarcity drives batch transaction costs up, which drives the average batch size up to compensate for it.

120

Dragan Stepanović @dstepanovic.bsky.social · 26/12/2025

Wait time variance goes up as the size of PR increases. The bigger the PR, the more uncertain when it's going to be merged. What does that mean? If you want higher predictability, you need smaller batches. Period.

133

Dragan Stepanović @dstepanovic.bsky.social · 28/09/2025

Sorry, it was a response to this LinkedIn post, which provides more context.

010

Dragan Stepanović @dstepanovic.bsky.social · 09/07/2025

Oh, I have another one! Available for pre-order!

093

Dragan Stepanović @dstepanovic.bsky.social · 09/07/2025

Or... Or... the team has mechanisms and practices in place to keep the cost of change low, so they are able to change code and adapt design just-in-time (emergent design), rather than trying to predict it when they have the least information about the code (design up front crystal ball gazing).

2.Code Churn

This metric measures the code added, modified, or deleted over time. It indicates how
stable or volatile the codebase is and how much effort teams spend on maintaining or
changing the code.

A high code churn means the code is constantly changing, which may imply that the code is poorly designed, the requirements are unclear, or the developers are not following best practices

030

Dragan Stepanović @dstepanovic.bsky.social · 17/06/2025

The problem with the code that is not expressive, i.e. implicit, is not only readability per se. It's that it's missing important domain concepts communicated in the code.

120

Dragan Stepanović @dstepanovic.bsky.social · 17/06/2025

Just a regular reminder that predictability (and accuracy of estimates) goes down the drain as you overload the system with too much WIP? How much WIP is too much WIP?

171

Dragan Stepanović @dstepanovic.bsky.social · 20/05/2025

fortunebuchholtz.substack.com/p/high-mood-...

000

Dragan Stepanović @dstepanovic.bsky.social · 04/03/2025

Any time we say “it will be faster to do this one bigger step than the two small ones” we're missing an opportunity to change the system in a way that makes iterating cheaper. It is faster to do one big step, _precisely_ because iterating in that system is expensive. 1/2

31910

Dragan Stepanović @dstepanovic.bsky.social · 18/02/2025

@github.com here's a feature idea: Instead of “protecting” the main branch by preventing pushing straight to main, try inverse and prevent creating branches other than main. Right, I know it doesn't sell, but that's what high-performing teams actually do.

151

Dragan Stepanović @dstepanovic.bsky.social · 15/01/2025

Yes, I like that idea as well and it's my go-to method. Leaning towards finding boundaries more based on empiricism than prediction (speculation). At Ohio university they've let people walk on the grass to let the path patterns emerge and only after that paved those paths.

010

Dragan Stepanović @dstepanovic.bsky.social · 02/05/2024

One of the big lessons of XP that often flies over people's heads is, don't focus on being fast. Focus on going steady instead. Fast is a byproduct of steady. Steady means you're minimizing rework and surprises by taking small steps and continuously invalidating assumptions along the way.

100