Please Go Brr, on Token Mandates
CEO token mandates aren’t about measuring AI usage. They’re an organizational search strategy for discovering where reasoning models create leverage, who can deploy them, and what businesses should fund next.
Ryan Lopopolo’s blog posts
CEO token mandates aren’t about measuring AI usage. They’re an organizational search strategy for discovering where reasoning models create leverage, who can deploy them, and what businesses should fund next.
Code reds are expensive adaptive episodes. Their output should not just be a fixed metric; it should be a maintenance loop.
Loki’s Tailscale upgrades used to threaten the access path they depended on. Codex turned that job into a checked-in harness with a canary identity, a runbook, and two human approval points.
Verification was always the problem. AI makes that obvious because every real task hides hundreds of small decisions behind one question: what does it mean to do a good job?
Let technical non-engineers go for it with coding agents. These folks are technical enough to have success. You need data-science-quality code, not prod.
A one-line red-team prompt sent Codex through the full security workflow on `intaglio`: prove impact, land the fix, prep the release, and publish a RustSec advisory.
Reasoning models scale with tokens. The ceiling is utilization: how much of the software lifecycle agents are allowed to observe and operate.
Symphony is distributed spec-first: a coding agent can rebuild it in another language from `SPEC.md`. The durable artifact is the set of decisions required to reproduce the system.
Full Japanese support for an alpha app would lose a roadmap fight. With agents, bounded work with a clear desired state and verifiable result can run as a background task.
With agents, a lint that finds 600 violations can ship with fixes and exhaustive coverage in one PR. Human attention moves from implementation to choosing invariants and evaluating proof.
@frantic asked for RSS, so I harness engineered the blog build (again): Vite-native SSR and assets, MDX posts, typed React composition, and static output that still deploys to GitHub Pages.
After six years of building a Ruby implementation in Rust, it’s time to archive Artichoke. This is a look back at why it started, what we built, and why I’m letting it rest.
I’d just like to interject for a moment.
Coding agents fail on tools they can’t see. MCP exposes a model-readable catalog—names, schemas, and prompts—so agents can discover, understand, and safely call your tools on the first try.
Ruby’s Enumerable mixin set the gold standard for elegant, composable iteration nearly three decades ago. With ECMAScript 2025’s new Iterator type, JavaScript is finally catching up—bringing this powerful pattern closer to becoming a universal language feature. Iterator type continues its expansion across languages.
Service meshes are less about tech and more about enforcing org-wide defaults without slowing teams down.
What is the purpose of a blog? Is it to write? to build and deploy the blog?
My résumé, written in LaTeX, is excessively over-engineered and 100% the way I want it. 🤪
To build iteratively and shed complexity, do the simplest thing that could possibly work.
As a Staff+ engineer, I’ve honed my focus by distinguishing projects where failure is acceptable, creating space for my teams to learn and grow, versus where it is not, requiring more direct involvement. Using this technique to delegate has further scaled my impact.
It is not the case that all proposals are open to all feedback from all stakeholders at every point of their lifecycle.
Pair programming with ChatGPT on Artichoke Ruby, writing a lot of Rust.
Pernosco is a hosted omniscient debugger that makes debugging complicated bugs tractable.
Interface compatibility does not require implementing an interface or a trait. Use compile-time flags and source-compatible APIs to swap out implementations.
Help your stakeholders by doing the hard work of distilling why the information you're sharing is important.
🌵 CactusRef lets you build cyclic data structures using strong references and knows how to deallocate unreachable cycles. You can use CactusRef to implement a doubly linked list. The CactusRef API is compatible with std::rc.
The border between Baarle-Hertog 🇧🇪 and Baarle-Nassau 🇳🇱 is one of the most complicated international borders in the world.
What I did this sprint: Less raw code output, more broad impact.
To be an effective senior engineer, you need to convince others to help you execute.
To learn Rust, I implemented an audio game. There were some things I liked and some things I didn't. Some things were easy and some were hard. And there were plenty of libraries that were a joy to use.
I contributed to open source more than I ever have in 2018; but there is a lot I can do to improve.
Terraform misconfiguration of SSM PrivateLink endpoint completely brings down hyperbo.la.
Moving blue-green deployments into terraform-managed autoscaling groups makes rolling out new AMIs easier.
Your AWS infrastructure is designed by your org chart. Alignment comes from increased communication and accountability.
A successful partnership between Engineering and Finance creates better business outcomes.