Humans drive the agent, agent drives the tools. - Humans - Declare intent / purpose - Agents - Decide how to accomplish it, fan out to tools - Tools - Provide deterministic info / feedback Improve the system each cycle. hamy.xyz/blog/2026-06...

AI creates 100% of my code now. I haven't written or edited a line of code in months. The closest I get is writing my blog posts which I happen to do in the same editor (nvim btw).

For one-shot prompts Ruby, Python, and Go are 1.5x faster and cheaper than Rust (ai-coding-lang-bench). But real agentic work is iterative. That's where Rust's compiler loop (expressive types, ownership) shines. These guardrails scale correctness to large codebases, orgs, and teams of agents.

I built a simple calculator for how I'm thinking about AI token spend ROI. Example calc: - Your cost: $100k - AI cost: $50k - Speedup: 1.6x - Your value with AI = $100k x 1.6 = $160k - AI value = $160k - $100k = $60k - Net value = $60k - $50k = $10k hamy.xyz/blog/2026-07...

SWE-bench Multilingual tested agents across 9 languages and Rust came out on top for task completion. Rust isn't the fastest or cheapest for AI generation - Python, Go, and Ruby win that. But the extra time spent iterating against types/compiler/linter turns out to be a feature, not a bug.

TL;DR - In 2026.H1, I started feeling like a real dad, took parental leave, went to a programming retreat, shipped CloudSeed Rust + QuietCut, reached 130k views, got back to building for fun, and started an AI diet. hamy.xyz/blog/2026-h1...

I built a fullstack webapp boilerplate using this approach: CloudSeed Rust - A Fullstack Rust Boilerplate for Building Webapps in Minutes - hamy.xyz/blog/2026-02...

I think talking will become the primary means of driving computers simply because it's the path of least resistance. I think of my mom, brother, and wife who still poke type and how much easier it's going to be for them to talk vs type large text blobs.

Anecdotal results of agentic engineering: - Before: I shipped 1-2 PRs a day, averaging ~7 a week - After: I ship 2-3 PRs a day, averaging ~12 a week So my guesstimate is a 50-60% speed improvement over my previous workflows. hamy.xyz/blog/2026-06...

How I build vertical slices with agents: - Build slices of atomic commits e2e - Iterate to get desired behavior - Push bottom commits to stacked PRs - Regen stack based on feedback / improvements hamy.xyz/blog/2026-06...

How I Balance Speed, Quality, and Mental Overhead in Agentic Engineering - Drive strategy, systems, and vision - Let agents do the footwork quickly - Do 1-2 things at a time and improve the system each cycle hamy.xyz/blog/2026-06...

QuietCut - a CLI that removes silence from videos: - Most of my videos are single take, talking head - This cuts editing time down to a single CLI call - Runs local, no subs hamy.xyz/notes/quietcut

The future of work with autonomous agents is like gardening: - Design the architecture, set the scaffolding, plant the seeds (tasks) - Set the agents loose - Leave and let things grow - Come back and tend to the system - inspect, refine, and redirect

I built a game where you spend Elon Musk's money: labs.hamy.xyz/labs/spend-m...

Why Rust pairs well with AI agents: - Expressive types make invalid states unrepresentable - Ownership + lifetime rules prevent runtime bugs like stale data and unexpected mutations The compiler becomes the deterministic check on nondeterministic AI output.

Hypermedia apps let you skip frontend frameworks entirely: - Backend controls frontend - Partial page rerenders - No code duplication Rust + Datastar example - hamy.xyz/blog/2026-03...

I've been talking to my computer a lot more and I think it's going to start being the primary way we use computers. - talk faster than type - LLMs can translate text blobs to commands - Free models / harnesses already good enough

Playing around with more personal software - here a reading dashboard for my last ~14 years on GoodReads. Make your own: labs.hamy.xyz/labs/goodrea...

I migrated my markdown blog from C# to Rust: - Latency: 2-8x improvement - Memory: 4x less usage Benchmarking my Markdown Blog in Rust and C# - hamy.xyz/blog/2026-02...

I've written Python, F#, C#, Java, and TypeScript professionally. But with AI writing most of the code I stopped picking langs for my own expertise and started picking them for foundational attributes like types, pefromance, and tooling. That's how I ended up with Rust.

The performance difference between my C# and Rust markdown blog doesn't really matter - sub-ms vs low-ms is imperceptible to users. But the difference in memory usage does, it allows me to run more apps per instance. - Rust baseline: 34 MB - C# baseline: 166 MB

Why High-Level Rust Wins Over Other High-Level Languages for Agentic Engineering - Expressive types + compiler hints - Mainstream lang w long-term support - AI is proficient - Great performance story hamy.xyz/blog/2026-04...

I host my apps in a multi-server configuration for a few reasons: * Learn how to do it, force myself to use infrastructure as code * Improve the existing DB bottleneck * Unlock future scalability w multi-server How I host my apps - hamy.xyz/blog/2025-12...

Why Rust Wins in the Age of AI: - Rust provides a deterministic foundation to build system on - AI agents provide nondeterministic ability to change the system Together you can move fast on a stable foundation. hamy.xyz/blog/2026-04...

Another one. FWIW Claude Code is not currently operational for me.

Want a reactive webapp without a JS framework? Hypermedia allows backend to control the frontend. - Simpler than SPAs - More interactive than MPAs Getting Started with Datastar - Build a Rust + Axum Todo App - hamy.xyz/blog/2026-03...

Datastar is a lot like combining HTMX + Alpine together in one library. - Server interactivity with SSE - Client reactivity with signals Rust + Datastar - Getting Started with Datastar - Build a Rust + Axum Todo App - hamy.xyz/blog/2026-03...

In Rust, cheap and expensive clones look exactly the same: - cheap_arc.clone() - ~10ns - expensive_string.clone() - 100+ns They look identical in code so one sloppy refactor can silently 10x your clone costs. So I built LightClone to allow the type system to warn you about these.

What I built at Recurse Center: - Applied AI for Engineering - Web Systems with Rust - Networked Games Plus a LOT of writing.

Why I'm moving to a Linux-based, terminal-focused dev workflow - and what it looks like - Ghostty terminal, zellij multiplexer - Fedora distro, Sway wm - neovim + lazygit - Obsidian for notes - Agents, agents everywhere hamy.xyz/blog/2026-03...