Coding Agents Have Crossed a Chasm  ## Metadata - Author: [[Changelog Master Feed]] - Full Title: Coding Agents Have Crossed a Chasm - Category: #podcasts - URL: https://share.snipd.com/episode/e02f3ffe-feb1-4edf-b5f2-538bb2f38468 ## Document Notes ## Highlights - **AI Coding Agents Transform Workflow** - David Singleton and Jerod Santo share how autonomous AI coding agents have become indispensable in their software development. - They no longer manually review code but iterate prompts to get the desired automated output. Transcript: Jerod Santo Coding agents have crossed a chasm. My recent experience aligns with David Singleton's, who says, quote, somewhere in the last few months, something fundamental shifted for me with autonomous AI coding agents. They've gone from a, hey, this is pretty neat, curiosity to something I genuinely can't imagine working without. Not in a hand-wavy, hype-cycle way, but in a very concrete, this is changing how I ship software way. End quote. I explained to Adam on Friends episode 98 that Claude Code feels like it's delivering on the promises that the AI hypesters have been making ever since ChatGPT broke the internet back In November of 22. I'm letting it code up many scripts and tools that didn't cross my personal ROI threshold previously. And David says the same quote, for personal tools, I've completely shifted my approach. I don't even look at the code anymore. I describe what I want to cloud code, test the result, make some minor tweaks with the AI. And if it's not good enough, I start over with a slightly different initial prompt. ([Time 0:00:48](https://share.snipd.com/snip/b88b4e53-a975-4829-ae85-f85779898cd5)) - **Avoid The Software Engineering Squeeze** - Software engineers must expand skills beyond coding to include product management and design. - Moving up the value chain is essential to remain relevant amid industry changes. Transcript: Jerod Santo The software engineering squeeze. Software engineers have had it easy for many years, but the times certainly are a-changing. Anton Zaydis is optimistic. Quote, all those companies started by vibe coders all around you, many will succeed and will need great engineers to scale up. Some engineers understand this and use the chance to scale up. To succeed, you'll probably need all the skills of an engineer, some of a PM, and even a bit of design taste. It's not just about shipping code anymore. But if you work as a code monkey, getting detailed tickets and just shipping them, you've done this to yourself. You won't be needed pretty soon. End quote. Say it with me, for the umpteenth time, we have to move up the value chain. Just keep moving on up to avoid the squeeze. ([Time 0:01:54](https://share.snipd.com/snip/d97a6d82-220e-4e39-9d98-65d75ec927cf)) - **Refining Kubernetes Defaults Matter** - Kubernetes suffers from too little opinionation, causing frequent misconfiguration and ecosystem churn. - Matt Duggan proposes ditching YAML and defaulting to IPv6 to enhance usability in a Kubernetes 2.0. Transcript: Jerod Santo What would a Kubernetes 2.0 look like? Matt Duggan has been working with Kubernetes for 10 plus years and has a lot of good things to say about it, but the journey hasn't been without problems. Quote, some common trends have emerged where mistakes or misconfiguration arise from where Kubernetes is not opinionated enough. Even 10 years on, we're still seeing a lot of churn inside of ecosystem and people stepping on well-documented landmines. End quote. So knowing what he knows now, Matt asks and answers the question, what could we do differently to make this great tool even more applicable to more people and problems? Here's the top level of his thoughts. One, ditch YAML for HCL. Two, allow Etsy de-swap out. Three, a native package manager. And four, IPv6 by default. Click through for the skinny on each improvement. ([Time 0:02:40](https://share.snipd.com/snip/40ed7757-b969-4af5-9387-53ec8895c8d7)) - **Coding Agents Are Commoditized** - Coding agents have become commoditized as their core functionality uses simple read/write loops with strong models. - Market competition focuses on token sales rather than unique agent differentiation due to minimal moats. Transcript: Jerod Santo Coding agents are already commoditized. Sean Gadecki does a nice job elucidating another thought I've had while talking to various devs about coding agents and seeing so many orgs shipping their own lately. These things have no moat. Quote, it turns out that put the model in a loop with a read file and write file tool is good enough to do basically anything you want. I don't know for sure that the closed source options operate like this, but it's an educated guess. In other words, the agent hackers in 2023 were correct. And the only reason they couldn't build clog code then was that they were too early to get to use the really good models. End quote. Sean says it's a tough market right now to sell a better AI coding agent in because open source solutions like OpenAI's Codex are already very good. But that's not even the point. The reason OpenAI wants you using Codex, and Anthropic wants you using Cloud Code, and Google wants you using Gemini CLI, and Sourcegraph wants you using AMP is not to sell you a better Coding agent. It's to sell you their tokens instead of you buying somebody else's tokens. The challenge with that strategy is that it has a very small moat. ([Time 0:03:42](https://share.snipd.com/snip/2b96c2ca-c839-45e5-9899-66f9831a840f)) - **Write To Clarify Thoughts** - Writing is challenging but necessary to reveal hidden flaws and biases in your thinking. - Use writing as a tool to critically examine and refine your ideas and understanding. Transcript: Jerod Santo Write to escape your default setting. Writing is hard, painful even. In the past, I've likened publishing an essay to birthing a child. They're both laborious journeys mired in contractions, heavy breathing, and occasional screams. My wife, who birthed six healthy children on our behalf, finds that analogy lacking. She deserves to. I'd come up with a better one, but writing is hard. I guess my point is, sometimes we need motivation to do hard things. If you need some reasons to write, this one's for you. Quote, at its best, writing can expose the ugly, uncomfortable, or unrealistic parts of your thoughts. It can pluck out parasitic ideas burrowed so deeply that they imperceptibly steer your feelings and beliefs. Sometimes this uprooting will reveal that the lustrous potential of a new idea is a mirage or that your understanding of someone's motive was incomplete. Maybe your own projected BS reflected back to you. ([Time 0:04:50](https://share.snipd.com/snip/46376302-b8e3-413b-bf4d-501644b79e1a))