Discovered Pangram via this post. I haven’t run into blog posts where an LLM has written most of it — lucky that I don’t get to follow/stumble upon that part of the internet. But I have stumbled upon too many website copies being authored by LLMs and it makes me furious all the time.
⭐ the senior engineer death spiral
Must-read for any engineer.
Introducing System One Models & Jev
A new day, a new thing in the AI world.
Jev means structured output is interesting again
But as Nelson Elhage famously said, fast software doesn’t just mean we can do the same tasks faster, it means we can do entirely new kinds of tasks. What kinds of new programs can we write by injecting 100ms worth of dirt-cheap intelligence at various decision points?
My usual read :)
Build momentum! — almost always the right way.
Fascinating — never thought this would be done like this in game dev.
I have been recently experimenting with using sticky notes for managing my TODOs. It has been working out okay some days and I go off track on some days. But I have to tell you this: it is so satisfying to crush the sticky note for a done task.
AI Has No Wisdom and Neither Will You
Okay, finally somebody is making a prediction that I haven’t heard anyone tell.
Where there is a will, there is a way!
A good CXO would make you feel that they trust you and steer you towards doing your best work at hard times. The trust part isn’t always obvious. For example, when a CXO says “I don’t want the details”, you might feel rudeness, but what it could mean might be different. Imposter syndrome could be at play. I believe it takes both the humans to operate with optimism in tough times to achieve great results!
I have been opting more and more for the slow internet.
Burnout by the Kindness of Agents
Mental overload is real!
⭐ ✅ How Will You Measure Your Life?
(read 122 pages)
I haven’t experienced something like this. The book belongs to a genre that I can’t explain. This is definitely one of the best books I have read this year. And it goes straight to my “suggest-to-anyone-as-a-must-read-book” list.
Digital Minimalism
(read 64 pages)
I already have a good hang of some parts of my digital life. But want to revise my way of using social media a bit.
Show Your Work!
(read 36 pages)
Grandma’s Bag of Stories
(read 5 pages)
Bedtime stories!
Letters from a Stoic
(read 3 pages)
Where you arrive does not matter so much as what sort of person you are when you arrive there. — Letter XXVIII
The Daily Stoic
(read 2 pages)
Tao Te Ching
(read a few random pages)
Those who are good I treat as good. Those who are not good I also treat as good. In so doing I gain in goodness. Those who are of good faith I have faith in. Those who are lacking in good faith I also have faith in. In so doing I gain in good faith.
Maybe We Shouldn’t Be Reviewing All This Code
I have been thinking about the same problem in the background for a while now, i.e. AI agents are dumping a huge amount of code and how do humans catch up?
Must read if you are interested in “Business”.
Ain’t it fun, living in the real world?
Can We Stop With the Uptime Percentages?
I always translate uptime percent to be “how much time this was down for the past 30 days?” — it takes a while. The author suggests to directly give that information instead of the cryptic percentage (for general audience).
The irrational effectiveness of the Pi harness
Pi again!
Every Customer Gets the Same Software. That’s Ending.
Interesting point to think about in the current timeline.
The oldest architecture in computing
Nice! Didn’t expect what the answer was.
Got me thinking — what advice do I have?
The Safest Job from AI may be Writing
I am convinced!
Rich, fully attributed context timeout errors in Go
Good to read a “Go” post after a long while.
Inspiring! It actually gave me an idea that might allow myself to write more.
You have to beat the models at something
Wow, this is the second post that this week that relays that “writing” is one of the things that is more human in the age of AI.
The first ever essay that I got to share with my sister!
I read this one regularly but missed adding it in some of my previous reading logs.
⭐ ✅ Keep Going
(read 200 pages)
Fantastic book! Must-read for people who are interested in doing art.
⭐ How Will You Measure Your Life?
(read 82 pages)
Definitely altering who I am!
Make It Stick
(read 41 pages)
Fahrenheit 451
(read 7 pages)
Introduction to Information Retrieval
(read 6 pages)
Tools of Titans
(read 4 pages, plus a few more at random)
Here is what I read from 24th August to 6th September 2026.
AI models have gotten so good that it is becoming trivial to produce a lot of code in a day. Almost unreasonable amount of code that it would take even days to read through and get into them. I have been facing similar situation recently. Dealing with them is going to get normal. Let us figure out how to navigate this!
Why I am shutting down my bookstore in New Delhi
I have become interested in bookstores recently. Tried visiting some new ones. Trying to get a sense for how it would feel to run a bookstore. The one that was closed in the blog post was a 30 years old bookstore. If something that had been running for such a long time could come to close, why?
Good intro!
Rebuilding Linear’s delta sync read path
Interesting. This has made a few more interesting questions to delve on. Going to explore a bit. What are inverted indexes? What is the purpose of GIN?
Same sentiment as post that I shared at the top. Huge amounts of code are getting written with AI these days. It is getting out of hand. I suspect that there is place for tools like this that help humans to have their understanding in sync with what AI produces. I am looking forward to more new ideas and tools like this. It is upto us to figure this out now!
AI-pilling our company: lessons learned
Good takeaways. This is how most software teams will look like — they will face all the points mentioned in this post and have some version of the agent powering the team.
The Bottlenecks for AI-Driven System Design: why principal engineers are not (yet) obsolete
What is a hill-climbing exercise? I notice this term being used when discussing about performance.
Essentialism
(read 203 pages)
Finished this book. It gave me a sense of calm and the courage to ask “what is essential here?” whenever I like to.
Zen Pencils
(read 9 pages)
Make gifts for people - John Green
Advice for Beginners - Ira Glass
Make it Stick
(read 5 pages)
Fahrenheit 451
(read 32 pages)
This is going to be the first fiction for me this year. I want to get more into fictional works and enjoy them as much as I enjoy non-fictions. Planning to alternate fiction and non-fiction - let us see how it goes!
Build something to disagree with
Loved it!
I have been personally having trouble in asking AI to come up with a design doc because it usually dumps a whole lot of 8 to 12 pages for a simple feature work. I arrived at a point that this post suggests to do first after experimentation: first thing to do is make AI ask you a lot of questions. The blog author tells about a skill for it: grill-me (tried it, pretty awesome!).
If you are working at / building a startup, this might be a good read. My guess is that building a culture like this is probably what we might want for a startup and people in those startups to succeed. I don’t know if I am making sense, but I like blog posts like this which give a new or contrarian mental model that I might want to consider. TIL YC makes startups demo every week. The post is from 2014, so not sure if this is true anymore.
Nice to associate a term for this. I personally am into a personal reset, 3 weeks in. My main personal reset component is to read blogs/books before sleep. More strictly until I find myself in a sleepy state.
Made sense before AI and makes sense even more with AI.
Reminded me of the times I used to work in offices and the random chats that occur with colleagues in the physical space. I do miss those chats in remote setup sometimes.
I am going to use this acronym both with humans and AI to let it rewrite the response in a manner that will help in unslopping. But yeah, never send slop to friends. Always handcraft.
Post from the creator of Obsidian (one of my favorite pieces of software).
I couldn’t fully internalize it.
Loved all the points. My favorite is point (1). I was thinking of getting into music. Point (10) was a surprise and I consider it my signboard to pick up some music lessons.
Identify the risks and understand what you are dealing with.
If you are working in startups or planning to go work for one, then this might be the good read for you.
Essentialism
(read 29 pages)
Reflections on my relationship with social media
My usage of social media is already low. I can go on days without opening Instagram. My Twitter/X usage has been spiking up a bit these days, so I looked around to see if there is something out there to help me. It is a solid post and I read it to gain more momentum to be more productive with how I use my time. There was this one point about “environments” that resonated with me.
If AI Writes All the Code, What Do the Programmers Do?
I liked the “Do I Enjoy This?” section. One particular quote:
My job isn’t to write code, my job is to fix problems and to ship new features, and “write code” was just the way I got that done before.
You don’t have to be smart if you can think clearly
The difference between a “smart” engineer and a “strong” engineer is how they react to problems that aren’t solved instantly.
A short essay about what is a product.
If nobody uses it or cares, is it still a product? What makes a product is its usage. If nobody is using the thing you built, you don’t have a product, you have a toy. Have fun playing with it.
River binaries will get 10 kB lighter
Good to read something Go related after a long time.
I have been following Fatih’s journey for a while. I don’t think he will discover this blog, but if he did somehow: “congrats, buddy! I am watching this space closely and I look forward to your updates.”
Touring The Consensus, six months in
I lurk around on the Software Internals discord and know about the Consensus project that Phil Eaton was working on. I haven’t had a chance to check it closely myself and this post came up as a good reminder of what the project has evolved into. I am exploring the site as I write this.
AI is removing the middle class of software engineering
Must read for the AI era.
You need to contribute beyond what everyone already gets by giving an agent a prompt
I personally think every engineer will hit a phase like the one mentioned in the blog post due to the usage of AI at work. The blog post reinstates the engineering energy you need to tackle such situation.
Startups are not easy. All startups are businesses, but not all businesses are startups. This one will help you identify a startup.
We don’t want the human trapped in the agent loop. The human is the loop, and we tag the agent in occasionally, thoughtfully.
Must read for engineers trying to navigate the current AI landscape. Discovered this post via a podcast (Next Token, Ep. 3).
Write something worth becoming known for
Wholesome post for aspiring bloggers.
Notes from the AI Coding Transition
I too felt lost in this AI coding transition phase. I am glad that I started reading blogs right now. A blog post like this is exactly what I wanted to read. This post is making me think how I am handling this transition and how can I better handle it.
The Inner Game of Tennis
(read 131 pages)
Completed reading this book with this batch of papers. This is about a lot of things. It will make you a human being.
Tao Te Ching
(read a few random pages)
“The way is wide open”