2026-07 - W4

ZTA: Zero Token Architecture - Kelsey Hightower | PlatformCon 2026 - YouTube

Great talk with the legendary Kelsey Hightower.

We’re going to go from burning tokens in a loop, to using tokens only once to create the loop, and never use inference again, until we change the loop. … So if you think context is amazing for agents, wait till you give it to humans

The Productivity-Experience Paradox - Annie Vella

Well written article, and worth thinking about, especially on the DevEx space.

For a long time, research has told us that developer experience and productivity are related - that they travel together. That’s the core premise behind the DevEx framework, and the reason leaders have been told to focus on improving developer experience: improve that, and productivity will follow. Unfortunately, this is exactly the coupling my data suggests may be breaking once AI enters the picture. … The correlation between change in flow state and change in productivity came out at 0.02, about as close to zero as it gets. And that’s the paradox: over time, with AI, developer experience and productivity appear to stop moving together.

Fragments: July 21

In recent weeks I’ve been noticing the stench of LLM-speak more and more. It’s not just the common tells, it’s a sense of LLM miasma that pervades the prose. I’ve noticed it’s increasingly eliciting a visceral reaction, after a couple of paragraphs I just want to dismiss the entire article out of hand.

I also “suffer” from this. Not sure why exactly. Is it some sort of natural biological response?

A while ago, I was thinking that it was reasonable for folks who aren’t as committed to writing as I am to use an AI to help polish their prose. Now I’m turning to encouraging writers to reject it. That pervasive LLM-voice is just so common now, my sense is that it discredits the writing even before the reader has a chance to try to understand what is being said. I don’t think it’s good enough to ask the LLM to write a first draft and then tweak it. I’m not sure writers can edit the LLM-ness out of prose once it’s in there. I even worry about asking an LLM to suggest improvements, I think it’s just too easy to accept an LLM’s suggestions, and in the process trigger your readers’ LLM-antibodies.

The Dark Factory Floods: What Happens When Agents Write Code No One Reads | Nash Borges

advanced-context-engineering-for-coding-agents/wsff.md at main · humanlayer/advanced-context-engineering-for-coding-agents · GitHub

The Four Core Areas of Responsibility for an Engineering Manager

Whatever shape the role takes, I find that the core responsibilities of an engineering manager can be distilled into four buckets, each with varying depth:

• People Leadership
• Technical Leadership
• Product Leadership
• Delivery Leadership

It’s getting harder to focus every day | Glyphack

Me too, me too… I can’t focus more than 5 minutes… Especially after sending prompts to the AI agent, I can’t help but check other things…

The Prompt-Wait-Evaluate Loop: How AI Kills Flow Without You Noticing | Sandor Dargo’s Blog

In the same vein of the previous blog post… We are trying to optimize our time after prompting by checking something else in the meantime, but we are sacrificing our focus/flow by doing so… And we don’t even notice it.

LLMs reward expertise

The usefulness of domain knowledge suggests that human expertise will continue to be useful even as models get stronger. For many tasks, the human is the bottleneck, not the model, because the difficult part is in communicating to the model exactly what kind of solution the human wants. The information is “in the model” already, but it takes a very smart human to pull it out.

Engineering management after the cost of code collapsed

The org you run was never designed but it rather accumulated. Every role, ritual, and layer exists because something used to be expensive: typing, routing, checking, remembering. The prices moved. The org chart did not. In the agentic limit, the chart stops recording who produces and starts recording who signs. Headcount stops measuring capacity and starts measuring how much accountability you can afford. The orgs that get there will look small, quiet, and mostly empty: a short list of names attached to a long list of decisions, and nothing else left to manage.

(Sounds LLM-written, but the LLM cliché highlighter did not flag any sentences…)

Why is programming so captivating, yet so agonizing? | elsewhere

First is the sheer joy of making things. Second is the pleasure of making things that are useful to others. Third is the fascination of fashioning complex puzzle-like objects. Fourth is the joy of always learning. Finally, there is the joy of working in a tractable medium. … The true challenge and mission is to find real solutions to real problems, within given time, with available resources.

Chat, we’ve made a terrible mistake… | sgnt.ai

Interesting. Could be like some audit thing.

I am now going back to all codebases I’ve worked on recently, claiming a “terrible mistake”, and seeing if Claude can find one…

Don’t Take the Black Pill - Andrew Kelley | SSW 2026 - YouTube

Really inspirational…

The Unreasonable Effectiveness of Constructive Data Modeling - Alexis King | SSW 2026 - YouTube

Interesting take, to use the positive space instead of the negative one. Will most certainly try it.

rtk Claude Code Token Savings: A Skill Trial Benchmark

It seems rtk does not save money after all…

From Prompt Engineering to Intent Engineering | Daniel Miessler

I’m also having the same feeling, i.e. putting less and less step by step instructions on my skills/prompts/commands as the models are now “smart” enough to handle the work by themselves.

One of the things you should be doing with your smartest model, like GPT-5.6 Sol or Fable, is basically reviewing all of your various prompts and scaffolding to see where you are violating Bitter Pill Engineering in this way, and switching over your HOW prompts to WHAT prompts.