↵ select ↓ ↑ navigate esc close

Richard Seroter's Architecture Musings

npub1lplakkhxym92x9y23td9sml2u6624vfcudnvc3h0ks985zgjfahq6swpsr@drss.io

Blog Featuring Links, Code, Thoughts, and Experiences with Software, AI, and Cloud Services

https://seroter.com

Daily Reading List – November 19, 2026 (#913)

20 Nov 2026

Today's links look at building always-on agents, how to bounce back after setbacks, and lessons from building an agentic code review system for Linux.

Daily Reading List – November 18, 2026 (#912)

19 Nov 2026

Today's links look at the next car on the Gemini model train (Neon), why codebase rewrites are back, and how to deliver projects with unclear scope.

Wait, is Google Antigravity *good* now?

18 Nov 2026

About a year ago, Google shipped the first version of the Antigravity IDE. It was basically an agent-centric VS Code fork. Fine tool, fine models underneath. Nothing world-beating, but certainly

Daily Reading List – November 17, 2026 (#911)

17 Nov 2026

Today's links look at the levels of agentic engineering, how to accelerate development for massive Git repos, and whether Git is even the answer in the AI era.

Daily Reading List – November 16, 2026 (#910)

17 Nov 2026

Today's links look at whether AI can fix its own slop with "humanizer" skills, whether subagents are harmful, and how to now measure developer productivity.

Daily Reading List – November 13, 2026 (#909)

13 Nov 2026

Today's links look at how to stop coding agents from faking tests, how different models behave with the same harness, and navigating the tensions of AI in the SDLC.

Daily Reading List – November 12, 2026 (#908)

12 Nov 2026

Today's links look at the new Gemini decision model, the components of agent memory, and building dashboards that discourage tokenmaxxing.

Daily Reading List – November 11, 2026 (#907)

11 Nov 2026

Today's links look at swarm teams with Antigravity, how Lyft approaches junior engineers, and why leaders should rethink their approach to AI oversight.

Daily Reading List – November 10, 2026 (#906)

10 Nov 2026

Todays' links look at two new members of the TPU family, the future of the product role, and why thinking harder may be better than adding agent skills

Daily Reading List – November 6, 2026 (#905)

6 Nov 2026

Today's links look at avoiding agent collisions, the state of enterprise AI, and how to avoid agent distraction

Daily Reading List – November 5, 2026 (#904)

6 Nov 2026

Today's links look at how context drift impacts your AI agents, what open infra for agents looks like, and what the new Go API for JSON solves

Daily Reading List – November 4, 2026 (#903)

5 Nov 2026

Today's links look at rules for evaluating agent skills, how to influence the agent, and what advice Jensen Huang gives to young developers.

Daily Reading List – November 3, 2026 (#902)

4 Nov 2026

Today's links look at running coding agents in parallel the right way, why trusting code is the hard part, and how LinkedIn scales GitOps.

Daily Reading List – November 2, 2026 (#901)

3 Nov 2026

Today's links look at how to make agents fast, why you should stop blaming code reviews for AI's delivery problem, and why you should default to serverless for agents.

Daily Reading List – October 30, 2026 (#900)

30 Oct 2026

Today's links look at the latest trends in AI and ML, what hill-climbing looks like when building agents, and what prefix caching is all about.

Daily Reading List – October 29, 2026 (#899)

29 Oct 2026

Today's links look at design tenets for agent identity management systems, how NOT to run an AI code review, and how to ship agent skills like products.

Daily Reading List – October 28, 2026 (#898)

28 Oct 2026

Today's links look at Gemini 4 Carbon, the state of context engineering, and whether AI agents shape experts' judgement calls.

Daily Reading List – October 27, 2026 (#897)

28 Oct 2026

Today's links look at how to measure AI productivity, whether 100% of your devs are improving their workflow, and how to make doc pages agent friendly.

Daily Reading List – October 26, 2026 (#896)

27 Oct 2026

Today's links look at modern AI coding workflows, when to stop using subagents, and why you shouldn't ask AI to write your user stories.

Daily Reading List – October 23, 2026 (#895)

23 Oct 2026

Today's links look at why CLIs are great if you want to minimize token use, why your software is slow (hint: not your programming language), and when to use diffusion LLMs.

Daily Reading List – October 22, 2026 (#894)

22 Oct 2026

Today's links look at how to use AI to become a more effective engineering leader, what the new Gemini API CLI can do, and a new micro-VM option in Google Cloud.

Daily Reading List – October 21, 2026 (#892)

22 Oct 2026

Today's links look at OpenAI models coming to Google Cloud, what a forward deployed product manager is all about, and whether judgement is keeping up with our dev velocity.

Daily Reading List – October 20, 2026 (#892)

20 Oct 2026

Today's links look at why an open source TorchTPU is a big deal, how the Claude Code team does code reviews, and why training your own model is a stupid idea.

Daily Reading List – October 19, 2026 (#891)

19 Oct 2026

Today's links look at how Meta designs a multi-agent architecture, how to do agentic verification, and why personal software factories are a waste of time.

Daily Reading List – October 16, 2026 (#890)

16 Oct 2026

Today's links look at why context inertia is something to look out for, what interactive refactoring looks like, and what happens when every app is now open source.

Daily Reading List – October 15, 2026 (#889)

15 Oct 2026

Today's links look at what's left after we stop coding by hand, whether AI review agents can take humans out of code reviews, and why judgement is now the job.

Your agents need a platform

15 Oct 2026

You’ll build one. Maybe buy one. Or just expect one beneath your managed agent. Regardless, agents need a platform. Why? Minimally, they need a compute host. Access to state management. Some basic

Daily Reading List – October 14, 2026 (#888)

14 Oct 2026

Today's links look at 38 things learned about building AI products, why you can aim for predictability AND desired outcomes, and how to debug when you no longer understand the code.

Daily Reading List – October 13, 2026 (#887)

14 Oct 2026

Today's links look at why Gemini 4 Argon is such a big deal, why we should write our own PR descriptions, and why software engineers aren't allowed to complain about AI automating their work.

Daily Reading List – October 12, 2026 (#886)

12 Oct 2026

Today's links look at why your good intentions aren't enough, why it's time to stop installing skills, and why a second edition of the SRE book is on the way.

Daily Reading List – October 9, 2026 (#885)

9 Oct 2026

Today's links look at how to build APIs for agents, whether your team's happiness indicates that you're a good manager, and how much it costs to use open source software.

Daily Reading List – October 8, 2026 (#884)

8 Oct 2026

Today's links look at the new Gemini agent, what sets high achievers apart, and why evals should be part of your deployment gates.

Daily Reading List – October 7, 2026 (#883)

7 Oct 2026

Today's links look at the state of the (developer) tech industry in 2026, how GitHub is rebuilding for agent scale, and what we miss now that coding is dead.

Daily Reading List – October 6, 2026 (#882)

6 Oct 2026

Today's links look at a framework for measuring context quality, why there's no such thing as an AI-ready culture, and why data is the application.

Daily Reading List – October 5, 2026 (#881)

5 Oct 2026

Today's links look at whether the future of AI is really autonomous, why agents need docs and not memory, and what mindset leaders need in the AI era.

Daily Reading List – October 2, 2026 (#880)

2 Oct 2026

Today's links look at how React Native solved a problem nobody had, whether you can SEO your way into an AI agent's recommendation, and why redesign doesn't happen enough.

Daily Reading List – October 1, 2026 (#879)

2 Oct 2026

Today's links look at maintaining quality while agents write the code, how to give your agent an UI, and why coding is NOT a solved problem.

Daily Reading List – September 30, 2026 (#878)

1 Oct 2026

Today's links look at the new Gemini 4 Argon model, how you start building an agentic software factory, and the state of agent skills.

Your app’s frontend UI is now optional

30 Sep 2026

I have 334 apps on my phone. You probably have more than me. My browser bookmarks and history are stuffed with sites I infrequently use. We’re inundated with “apps” that require us to learn

Daily Reading List – September 29, 2026 (#877)

29 Sep 2026

Today's links look at why you should be asking more questions, why there's a lack of confidence in software supply chain security, and why faster AI coding can mean harder engineering.

Daily Reading List – September 28, 2026 (#876)

29 Sep 2026

Today's links look at why experts should lead experts, which unlikely trends are shaping software trends, and why startups need open models alongside frontier APIs.

Daily Reading List – September 25, 2026 (#875)

25 Sep 2026

Today's links look at why you should stop watching your agent work, the value of a good agent plugin, and how to better harden your code pipelines and CI/CD infra.

Daily Reading List – September 24, 2026 (#874)

25 Sep 2026

Today's links look at a new type of agentic database architecture, whether web frameworks matter anymore, and how to deploy AI hardware into outer space.

Daily Reading List – September 23, 2026 (#873)

24 Sep 2026

Today's links look at why the internet isn't ready for the agentic wave, how to scale Kubernetes workloads to zero, and how to do a Jev-style classifier with DiffusionGemma.

Daily Reading List – September 22, 2026 (#872)

23 Sep 2026

Today's links look at using Jev in your data warehouse, why traditional API design has changed forever, and good tips to improve your coding agent game.

Daily Reading List – September 21, 2026 (#871)

22 Sep 2026

Today's links look at the new Jev model, why 'be popular' isn't an open source business model, and how to do cross-cloud caching with your lakehouse.

Daily Reading List – September 18, 2026 (#870)

18 Sep 2026

Today's links look at whether you're actually a good collaborator, why you should stop trying to be LESS obsessed, and why we're all managers now.

Daily Reading List – September 17, 2026 (#869)

18 Sep 2026

Today's links look at what happens to engineers when AI writes all the code, how to de-slop an AI-generated codebase, and why agents are the new kingmakers.

Daily Reading List – September 16, 2026 (#868)

17 Sep 2026

Today's links look at why you tell agents why instead of only how, whether agent skills actually help, and the value of being bored.

Daily Reading List – September 15, 2026 (#867)

16 Sep 2026

Today's links look at how to pair program with your AI agent, how to keep your jokes from falling flat, and why we're all product engineers now.