Vertical tabs are now natively supported in Chrome \o/. You still have to enable it under chrome://flags/#vertical-tabs and then go to chrome://settings/appearance and toggle ‘Tab strip position’ to Side. I also think the URL bar and the extensions haven’t yet received a finalized position. I am not complaining though – this is solid progress and is still an improvement for those of us that live the vertical-tabs life.

The second feature is for those who like command line agents. chrome-devtools-mcp allows your command line agent to directly access your logged in session of chrome. This comes with the caveat: understand the security stance of your machine before entering the lethal trifecta. This is still a big unlock. I am testing out the boundaries and hope to have an update post on the features and the capabilities soon.

There’s an old saying about not meeting your heroes. In practice, leaders tend to confirm this over time. This is true across domains, and it’s rarely a single gaffe that does it. The interesting question is why the disappointment usually takes the same shape.

Disappointment does not always show up in the form of a bad conversation. Often there isn’t any conversation at all, at least not in the way people imagine one. As space disappears, interaction collapses into reaction. Responses come faster, positions are stated rather than tested, and dialogue gives way to declaration. At a certain distance, leadership becomes parasocial by default, taking the form of broadcast. There is nothing to push back on, only things to react to. By the time the gaffe happens, the system has already collapsed.

Ariadne Conill

@ariadne@micro.blog

Why leaders often disappoint us: ariadne.space

January 22, 2026, 21:54

This was a dense read. However, if you take the time to read and process, I guarantee you that you are going to learn something important for your career – be that of a worker or be that of a leader.

Harmonious integration is the most overlooked trait that good leaders can provide. The stability and safety of being able to be who you are and contribute to the shared cause is a delightful environment. This is something that I’ve learned time and time again. The garden metaphor is my go to for leadership. It rings true here as well.

Bertone is calling this a ‘neo-retro’ rebirth of the Autobianchi A112 Runabout by Bertone, a concept car for the 1969 Turin motor show. Styled by Marcello Gandini, it was inspired by contemporary – wait for it – speedboats. The 2026 version comes as either a permanently open barchetta or more practical targa, and is a proper production car you can drive on actual roads, albeit severely limited in numbers, with only 25 slated for production. None of them float, just to be clear.

With my.WordPress.net, WordPress runs entirely and persistently in your browser. There’s no sign-up, no hosting plan, and no domain decision standing between you and getting started. Built on WordPress Playground, my.WordPress.net takes the same technology that powers instant WordPress demos and turns it into something permanent and personal. This isn’t a temporary environment meant to be discarded. It’s a WordPress that stays with you.

This is very neat. This was the promise of the thesis behind progressive web apps. I am glad to see it finally realized. I am not going to be pollyannish and believe that this will somehow change the world – there are many more variables and inertia for the web to actually be a viable application platform.

However, combined with LLM coding, this is a SUPER neat way for developers to provide application services without needing cloud computing but leverage the web environment present in almost every computer in the world.

To me that’s super exciting. To me that’s an avenue not many people are playing at. To me, that sounds like opportunity.

Our interfaces have lost their sensesAll day, we poke, swipe, and scroll through flat, silent screens. But we’re more than just eyes and a pointer finger. We think with our hands, our ears, our bodies.The future of computing is being designed right now. Can we build something richer—something that moves with us, speaks our language, and molds to our bodies?

What a well crafted site to communicate a thesis. <3

The WhatsApp notification pings have finally gone silent. My phone sits on the desk, a quiet slab after an hour of writing three posts. Three blog posts in a single day is not a feat of willpower. It is the result of a system that is (hopefully) starting to compound.

My agents used to be strangers every time I opened a terminal. Now they are a persistent extension of my own skepticism. The hypothesis that a compounding personal context could change the way I work is building some additional reps. As a fan of the open source ethos, I want to share the scaffolding of the work that makes it possible.

The Scaffolding: Building the Persistent Memory

I start in the home directory. This is the root of the digital nervous system. I create a hidden folder to house everything that matters.

mkdir ~/.agents

Then I populate it with “skills.” These are the functional primitives that allow the agents to do more than just chat. I spent an afternoon browsing the awesome-llm-skills repository, picking the tools that fit my workflow. Based on existing skills, I developed my own blog-coauthoring, humanizer, and personal-context. Each one is a specialized capability I can invoke at will as well as setup the agents to default invoke. For example, I always set personal-context to be loaded first.

The real heart of the system is the personal directory. This is where I store the “earned frame” of my life. I create three documents: about-me.md, my-voice.md, and working-style.md. I seed these with the grit of my actual process. My skepticism, my preference for short sentences, and my mandatory branching workflow all live here.

mkdir -p ~/.agents/personal
touch ~/.agents/personal/about-me.md
touch ~/.agents/personal/my-voice.md
touch ~/.agents/personal/working-style.md

I also need to give the agents their marching orders. I create AGENTS.md and GEMINI.md inside ~/.agents. These files contain the “Memories” that ensure every session starts with perfect calibration. They mandate a “Compounding Protocol” that forces the system to learn from every interaction.

## Shared memories
- Always pull personal-context using the 'personal-context' skill
- Mandatory Compounding Protocol: At the end of every interaction, synthesize what you have learned and propose updates to the personal directory.

Once the repository is ready, I install the tools. I use Homebrew to get codex and gemini on my machine. But I do not want them using their default, fragmented configurations. I remove the standard AGENTS.md and GEMINI.md from their respective folders and replace them with symlinks to my central ~/.agents directory.

ln -s ~/.agents/AGENTS.md ~/.codex/AGENTS.md
ln -s ~/.agents/GEMINI.md ~/.gemini/GEMINI.md

Now, every time I open a terminal, my agents are looking at the same source of truth. I test the setup by asking a simple question: “What are your skills, and do you know who I am?” When they respond with my specific voice and a list of my curated tools, I know the bridge is built.

The Resolution: The Shared Sovereign Workshop

The hypothesis is proving true. Three posts in a single afternoon is positive empirical evidence I needed. My agents are can be magnifying glasses for my own intent. I am no longer fighting the token limit of my own memory or the tiredness that comes with mixing the drafting and the editing phases.

I am sharing this scaffolding because I want to see what others build when they stop repeating themselves. The open-source spirit is about more than just code. It is about the “earned frame” of our creative lives. I hope this guide helps you build a system that finally remembers who you are.

I am excited about this shared nervous system that stays awake even when I switch machines or switch context. The results aren’t perfect. It often treats exceptions as the norm. Tends to reuse some language across its narratives, which do require an intervention. It might be that this can never produce non stilted prose. Will this compounding context eventually lead to a hazy simulacrum of my own voice? I do not know yet. All I can do is keep working at it and sharing the results. It is a quiet power to know that my tools are as resilient as my ideas, even if the destination remains unproven.

I am sitting in a room finally cleared of five years of paper. There is a specific relief in seeing a clean desk, but it also highlights a hard truth: willpower is a trap. I cannot rely on my own discipline to keep things this way. I decided to outsource my memory to a hidden folder in my home directory. My agents used to be strangers every time I opened a new terminal, but now they remember the exact weight of my skepticism through a few strategic symlinks.

The thud of a five-year stack of documents hitting the recycling bin is a heavy sound. It is the sound of a mess finally addressed. As I sit here chewing gum in the quiet, I realize that my productivity has always been a precarious beauty, held together by the duct tape of temporary focus. I am forty-one, and I have finally admitted that I cannot outrun my own inconsistency. I need systems that compound. I need my tools to remember the “earned frame” of my previous work so I don’t have to rebuild the context from scratch every single morning.

The realization hit me while I was deep in a session with Gemini, rewriting the CSS for my website. The code was clean and the logic was tight, but then I hit a wall: out of tokens. I switched to Codex to finish the footer, but the context was gone. It was like waking up with amnesia in the middle of a conversation. Codex didn’t know about my preference for vanilla CSS or my creative philosophy. It is a redundant masterpiece of an LLM that is effectively a paperweight without my personal context.

The token budgets of these models are fragmented, forcing us to jump between them mid-thought. I found GEMINI trying to write its working memory into ~/.gemini/GEMINI.md, while my other agents were looking for a similar file in ~/.codex/AGENTs.md. It was a mess of fragmented context. My tools were behaving like strangers in the same house. I decided to stop the fragmentation by unifying everything into a single ~/.agents/ directory. I used symlinks to point both .codex and .gemini back to this central source of truth. Now, when I update a skill or a personal preference in one place, every agent feels the change. It is a shared memory that I can version control on GitHub and carry across my laptop and desktop.

This is where it gets meta. As I sit here chewing gum and drafting this very post, the system is watching and learning. I am using a blog-coauthoring skill that I refined from a doc-coauthoring template. Every time I correct a tone or pick a specific analogy, the agent is updating my personal/my-voice.md and working-style.md files. It is capturing the grit of my actual process.

We are building a feedback loop that compounds. Each session isn’t a fresh start; it is an iteration. By the time we finish this draft, the agent will have a sharper sense of my skepticism and my preference for short, punchy sentences. I am not just writing a blog post; I am training a cognitive auxiliary to understand the “earned frame” of my creative life. This setup gives me a sense of sovereignty over my tools. If I switch from my laptop to my desktop, or from Gemini to Codex, the memory is already there. The system doesn’t just work for me; it grows with me.

The Sovereign Workshop: A Compounding Algorithm

# The Sovereign Workshop: A Portable, Compounding Nervous System
Setup(New Machine):
  - Clone(Nervous_System): "git clone context-repo ~/.agents"
  - Link(Shared_Memory):
      - Create_Symlink: "~/.agents/AGENTS.md"  -> "~/.codex"
      - Create_Symlink: "~/.agents/GEMINI.md"  -> "~/.gemini"
  - Verify(Calibration): "Call 'personal-context' to wake the memory"
Session(Start):
  - Sync(Context): "git pull latest from ~/.agents"
  - Calibrate(Persona): "Activate personal-context skill"
  - Observe(Environment): "Scan current workspace for READMEs and existing drafts"
  - Align(Inference): "Connect current goals to the 'earned frame' of previous work"
Session(Execution):
  - Action(Creative_Work): "Apply Illustrative Clarity and Active Voice"
  - Feedback_Loop(Correction): "Update internal voice and working style models"
Session(End):
  - Synthesize(Compounding):
      - Analyze: "What did we learn about voice, identity, or process today?"
      - Propose: "Draft updates for about-me.md, my-voice.md, and working-style.md"
  - Persist(Longevity): "Commit and Push context updates to the Nervous System"
  - Coda(Skeptical): "End with an honest doubt. Avoid the 'AI happy ending'."

The real win is portability. By pushing ~/.agents to GitHub, I am no longer tethered to a single machine or a specific local disk. If my laptop dies or I decide to wipe my desktop to clear out the digital cobwebs, I don’t lose the years of context I have built. I clone the repo, run a few symlink commands, and my agents are back to knowing exactly how I like to work within minutes.

I have built a shared nervous system that stays awake even when I switch machines. I hope this is the end of the “fresh start” and the beginning of a truly compounding creative life. But I am skeptical. Systems have a way of decaying or becoming their own kind of weight. Or in the metaphor of a garden, how do I keep culling?

There is a deep dissonance here between what we perceive as “active” software and what actually constitutes a stable business. It is the difference between a cast-iron skillet and a smart appliance. One has a clear premise: it sears meat and lasts for generations because its “technology” is local and requires no external permission to function. The other is a promise of future convenience that only works as long as the manufacturer’s servers stay on and the venture funding keeps the lights on at the office.

The Perceived Decay

A sharp wince is an uncomfortable way to start a morning. It isn’t the coffee, which feels just right. It is the realization that a digital tool I have used for a decade is being drafted for an obituary. On Reddit, the consensus is clear: Alfred is “in trouble” because it hasn’t pivoted. It hasn’t announced a venture-backed “AI engagement point” or a flashy new UI. It seems like the community conflated iterative feature-creep with foundational stability.

This is not a tirade against LLMs. In the eerie quiet of my office, I just used an LLM to fix a broken footer on my website in minutes. The pace of deployment is intoxicating, but it masks a fragile truth. We now judge a tool’s pulse by the frequency of its visual updates rather than the structural longevity of its business. We have started thinking superficially about what makes software last, assuming that if a developer isn’t shouting into the void, they must be drowning in it. We have to wonder if “lasting” is even a goal we still value.

The Economics of Longevity

The ceramic mug in my hand is a local solution to a physics problem. It doesn’t require a firmware update to hold coffee to remain a vessel. It is functional, silent, and stable. This is the structural premise of Alfred. It is a self-funded, local-first tool with almost no dependency on a cloud instance. Its primary costs are human: a few developers ensuring it stays patched for the latest macOS update. Its economic incentives are aligned with its longevity because its survival isn’t tied to a growth-at-all-costs mandate. It costs very little to run freely and it charges for the powerpack, which seems like it could be profitable.

The architectural dissonance lies in the mismatch between a tool’s functional premise and its underlying economic incentives. Raycast, by contrast, is a venture-subsidized experiment. There’s no value judgement here about venture-subsidized experiemnts. Innovation requires money and I am glad we get to try different things. Raycast’s shift from a launcher to an “AI engagement point” isn’t just a feature update; it is a fundamental change in its supply chain. By pivoting to the cloud-heavy requirements of LLMs and expanding to Windows, my thesis is that they are building a “smart appliance” that requires constant external power. This power comes from both compute and venture capital and a new, modern way to engage its community. If the cashflow or the cloud infrastructure or the community engagement disappears, the tool follows. We are witnessing a divergence between local-first solvency and the precarious beauty of a venture-backed pivot.

I must admit that the community engagement is a marketing mechanic that has both pros and cons. The pros are a very engaged community and a hundred voices amplifying your latest feature and converting other users. On the other hand, rampant viral growth has the downsides. You don’t need it when unit economics are strong and there is a real cost to maintaining a community.

The Aspiration of Sovereignty

The quiet of this morning represents the freedom I am working toward. After years on the corporate treadmill, I find myself aspiring to build a lifestyle business: a “hearth” rather than a “high-rise.” It is a choice to optimize for sovereignty over scale, seeking a niche of users who look beneath the superficial churn of visual updates to understand the structural integrity of their tools. I want to build things that don’t require me to be beholden to venture capital or the noise of a constant pivot.

Yet, as I sit here with my coffee, I have to wonder: is this a truly rational path, or am I merely rationalizing a desire for peace? There is a seductive comfort in the idea of a self-contained, local-first business. In a world of hedonistic adaptation, where users are trained to expect a hit of dopamine with every new feature, stasis can feel like a death sentence. Am I projecting my own need for stability onto a market that only values momentum? This is the fundamental gamble of the lifestyle business. It requires setting an honest expectation with the user: you are paying for a tool that works, not a promise that it will constantly change.

Right after I hit publish, Raycast rolled this out

They are becoming a software house – the new product looks very interesting. I am glad they were able to work on this. However, similar to the Arc browser saga and Dia, one can’t comment that Alfred will die while Raycast will survive.

I’ve templatized my solution and call it gurupanguji/simulacrum. It is a clean, open-source template for LLM-independent personal context management. I am releasing it today on GitHub.

The Prompt-to-Project Leap

The creation of simulacrum itself was the “meta” proof of the protocol. I had the idea for a clean, open-source version of my context-management system at 10:00 AM. By 10:15 AM, the entire repository was live on GitHub, fully formed with protocols, placeholders, and a personal-context skill. This wasn’t a feat of typing speed. It was the result of a system that finally stopped treating me like a stranger. My agents knew the “earned frame” of my life: my skepticism, my preference for short sentences, and my mandatory branching workflow. The time it took from idea to generation was the length of a single prompt.

Ending Agent Amnesia

Model amnesia is the tax we pay for fragmented context. I often start a task in Gemini and hit a token limit or a reasoning wall. I switch to Codex or Claude to finish the work, but the new agent is a stranger. We are currently living through a period of “Agent Amnesia” where we switch models and lose the last hour of our lives.

I need a system that defaults to remembering who I am and the work I’ve done. simulacrum not only continues the context because of its compounding philosophy, but also ensures my context stays mine, independent of whichever model is currently holding the pen.

The architecture is minimal: a simple scaffolding that guides agents to write relevant context into shared files across models.

I am not sure if this setup will stay relevant as models evolve. But even if the companies themselves aren’t motivated to preserve my context, simulacrum ensures it stays mine.

The Runway Stack

I treat my environment as a “runway stack.” Each layer is designed to remove drag. I use Ghostty for a jitter-free terminal. I use a minimal zsh setup. The core architecture is a hidden ~/.agents directory. This is the digital nervous system of my workshop.

It contains three primary components:

1. The Personal Directory: Markdown files that store the grit of my actual process in personal/.
2. Specialized Skills: Functional primitives that reside in skills/.
3. The Compounding Protocol: A mandate that forces the agent to synthesize what it learned and propose updates to the core files.

By symlinking these files into the configuration folders of tools like Codex and Gemini, I create a shared memory. If I update a preference on my laptop, a git pull on my desktop ensures my agents are perfectly calibrated within seconds.

Sovereignty and Scale

The value of simulacrum is independence. You are no longer beholden to the memory features of a specific provider. This local-first approach ensures your context is portable and private.

This architecture also allows for easy integration with exciting developments like openclaw and nanoclaw. Whether you are using a heavy-duty CLI or a lightweight agentic script, the source of truth remains the same. You are building a cognitive auxiliary that grows with you rather than resetting every morning.

Open Source and Contributions

This is the first step. I am making the simulacrum template public to see what others build when they stop repeating themselves. Hopefully it’s as useful to you as it is to me.

You can find the repository here: gurupanguji/simulacrum

I invite you to try it, star the repo if you find it useful, and contribute your own skill templates. We are building the primordial primitives of a duct-taped life, one session at a time.

I’ve been thinking about the intersection of Skills and LLMs. I keep coming back to one question: can I use this to improve my own thinking? I write to think. If I think, I eventually want to publish. I realize I never really describe my writing process. I never think about how I write. I just type. On lucky days, coherence appears. The rest stays as drafts in a folder. I wonder if I can build a Skill to partner with me in co-authoring. I want a system that helps me translate my internal monologue into finished work. Documenting this process provides a blueprint for others.

I looked online and found this tweet to be quite useful to learn the basics of Skill construction.

I’ve maintained a “working-with-me” for colleagues for years now. So, I used it to get started to setup the primitives – about-me and my working-style. I used passages from this blog post and some old professional documents to feed good examples of my-voice.

I watch the LLM as it reads my primitives. I hear the clack of the keys as I refine them. It scans my history as a “terminal spelunker,” my vegetarianism, my preference for Swift and Python, and my camera gear. It adapts. It starts with a vague imitation of my voice, and it slowly grows into a partner that attempts to predict how I might finish my sentences. It is not perfect. It is a hazy simulacrum. This confirms that context is the new currency. If the model doesn’t understand your perspective, it can’t help you tell your story.

I decide to push past these basic boundaries. I need a better instrument for writing. I do not just ask for a blog post. I still have this (irrational?) belief that my voice is unique. So, I decide to build a blog-coauthoring skill.

This is the longest part of the process because I have to figure out “how” I write. It is not easy. Wait, “not easy” is too mild. It is exhausting to audit your own subconscious habits, even if it is interesting. I break them into stages: The Spark, The Hook, The Editor’s Desk, and The Resonance Test. But structure is just a generic frame. I have to define what makes the writing feel like “me,” the journal-like tone, the active voice, and the way I use examples to bridge the gap between an observation and a hypothesis. The manual isn’t a finished document. It’s a dynamic blueprint for a shared intelligence. It is living in a weirdly anthropological and ontological sense.

Skills are like shell scripts for agents. They’re the levers you use to turn raw potential into something more kinetic. I am reminded of pottery. The clay is the context; the wheel is the model. One needs to know how to mould the clay and make a vase. It’s the skill (small s). Then examining that and then writing that down into a program that a robot can then try its hand at making similar vases – that’s what a Skill does.

It’s time to test it. I use it to brain storm my key takeaway. That skills are like man-pages for the LLM world. The discussion moves past nostalgic shell scripts to debate the future of sandboxed intelligence and the risks of data exposure. I discuss vocabulary options testing the ontological abysses to see if the structure holds. The truth is that the co-authoring is a feedback loop. The LLM acts as a mirror, reflecting my thoughts back at me, sometimes with a sharper edge that I started with.

There’s a strange recursion here. The SKILL.md file I just wrote is the tool I’m using to write this post. I make edits to the post to make it more of my voice and then I ask the LLM to analyze the difference. It suggests that I use different types of sentence lengths; it calls out my use of first personal singular point of view and my use of active of voice. I prompt that I like the writing as if I journal my observations.

In some ways, I’m using the manual to write the manual. Maybe, the process is the point! Is this the process of building a partner that can actually synthesize my intent. The “Aha!” moment isn’t finishing the post. It’s realizing I might be able to build an engine for more!

I’m back at the blinking cursor now. I don’t know if this skill will stay useful over the long term.

The process is full of conflicting feelings, questions about identity and authenticity mixed with a lot of curiosity about what happens if this actually works.

I don’t have a co-author in the sense that I started out with. Yet, the process did help me introspect and learn about my own process more. And maybe with enough iterations, there might be an aid that can convert more drafts into posts.

I’m still irrationally protective of my voice, but I can see the value in separating the “thinking” from the “drafting.” I wonder if I should break this down further into separate skills for thought, editing, and voice. It’s an exciting place to be. I am ready to keep tinkering!

I realize that my brain when left idle spawns unlimited threads. One can say that I am almost overloaded with signal. It could be sound, touch, smell, sub conscious, unprocessed thoughts. I have NEVER had “nothing” go on in my brain.

There are a few times my brain quiets down. When I am enthralled by a particular problem and I am processing through the various aspects of it. When listening to music, when I am working out.

In almost all other times, my brain is constantly churning.

I can articulate it only now but I’ve done this since I was a kid. Writing is one of the ways I slow my brain down and focus on a problem. Writing slows down the infinite thread spawn and it helps me ignore signals from the outside world. Signals that are otherwise like siren calls for my attention.

Writing helps me slow it down to the speed of my pen as I craft my thoughts into words. It’s honestly, relaxing. I think this is why I write. I write to process. I write to think. I write to not go to my default setting. I write to clarify. I write for the unprocessed thoughts to shrink. I write for relief – a respite for my churning brain. I write for peace.

It feels cathartic. I didn’t realize writing is this therapeutic to me.

While I also write digitally, like this post, I want to talk about writing in my journal. I restarted writing in a journal in 2024. I’d gone digital since 2003-ish and it stayed that way for 20 years.

In an attempt to help process my unprocessed thoughts, I picked up writing using pen and paper in 2024.

I grew up to see my dad be particular about things. He always used a Pilot Hi Tech Point v5 pen. He would purchase a blue, a black and a brown (and have inks available to refill as needed). He was also particular about writing on good paper. While the choices were low, what I remember most is the JP Bond paper that he used to prefer along with his diaries.

So, maybe that’s why I also grew up to liking pens and good paper. I use a Pilot Frixion Knock 0.5mm pen. It’s perfect for the type of glide I like on the type of slightly creamy paper that I prefer to write on. I still have a soft spot for 2 more pens – the Pilot Precise v5 (the equivalent of the Pilot HiTech Point that my dad prefers) and the UniBall One. The Precise has more watery ink and can hence smear on certain types of paper even though it can often be smoother to write. The UniBall One at 0.38mm is one of the finest point pens that maintains a good balance between a fine point and the scratchiness associated with finer points.

I usually write early in the morning. I wake up at 3:30 and do my journaling from about 3:30 – 4:30 with a warm mug of coffee.

I don’t go in with a specific intent on what to write. Some days my brain starts with a plan for the day. Other days, might be a feeling I want to process. And other days might be recalling a concept or a thought and then looking at it from different angles. I also almost always doodle along with the writing. I find that visualizing concepts that way helps it stick a little better too.

I follow the flow. I find that it often brings out unprocessed feelings or connections that my brain might have formed subconsciously, but can now articulate.

This post is a perfect example of such writing. I started with no specific intention. The flow took me here. I can make a connection though. A close friend had asked me how I journal and I told him very similar words.

p.s. it’s interesting that how I journal is also how an LLM operates. Will have to spend some time thinking about that in a future post.

AI building a C compiler is not truly revolutionary, but it does reveal how far AI coding has progressed and where it may be heading next.

• Before diving in, here are my main take-aways:
• AI has moved beyond writing small snippets of code and is beginning to participate in engineering large systems.
• AI is crossing from local code generation into global engineering participation: CCC maintains architecture across subsystems, not just functions.
• CCC has an “LLVM-like” design (as expected): training on decades of compiler engineering produces compiler architectures shaped by that history.
• Our legal apparatus frequently lags behind technology progress, and AI is pushing legal boundaries. Is proprietary software cooked?
• Good software depends on judgment, communication, and clear abstraction. AI has amplified this.
• AI coding is automation of implementation, so design and stewardship become more important.
• Manual rewrites and translation work are becoming AI-native tasks, automating a large category of engineering effort.
• AI, used right, should produce better software, provided humans actually spend more energy on architecture, design, and innovation.
• Architecture documentation has become infrastructure as AI systems amplify well-structured knowledge while punishing undocumented systems.

The implications for engineering teams are real and immediate. At the end, I share how I’m translating these insights into concrete expectations for my team at Modular.

I respect Chris Lattner. I have a long-running bias toward LLVM, so when he talks compilers, I pay attention. Pair that with Adam Neely’s analysis of AI music and Suno, and I had the same reaction twice: these tools are making real gains.

The announcement cadence is still relentless. A lot of it is incremental, and not every release changes my day. But the recent jump did. It felt like the compounding finally showed up in a way I could feel: better reasoning, longer follow-through, and stronger models built on better training and deeper knowledge.

The thread I keep pulling from both Lattner and Neely is: I do better with these tools when I stay in the loop. When I treat them like a collaborator I steer, not a machine I merely ask for answers, the output gets better and I do too. 

So my current practice is simple. I keep playing with the tools. I give myself permission to be bad at them at first. I learn what kinds of prompts actually move the work forward, what kinds create noise, and where I still need my own judgment. Over time it has started to feel less like “using AI” and more like building a new kind of workflow for creative work.

I also understand the fear that this stuff automates the human away. I feel it sometimes. When I am already anxious, it is easy for every new capability to read like another door closing. The more out of control life feels, the more convincing that story becomes, whether the lack of control is real or my ego’s rationalization to protect itself.

But I keep noticing a counter-effect in my own life. The same tools that shrinks parts of my work also expands my choices. When I use them well, I feel more capable. I try more ideas. I  iterate faster. I take swings I would not have taken. The cost of “starting” drops, the cost of “being stuck” drops. It feels like a small, but real increase in my locus of control. 

And that is where my video game brain kicks in.

I love games. They are one of the few places where failure is clean. You die, you respawn, you try again. No shame spiral required. Over time I started to notice that the most useful way to play, for me, is hard mode. In retrospect, I saw this with dear friends who enjoyed the process and not the outcome. Yet, it wasn’t obvious to me, at first. 

Hard mode changes what I am doing. I stop trying to “finish” and start trying to learn. If I wipe on a boss, I do not read it as proof I am bad. I read it as data. I ask what I missed, what pattern I did not see, what tool I ignored, what timing I rushed. Then I take another run with one change.

That loop taught me something I keep borrowing outside games. Repeating the same approach and hoping the environment changes sometimes works, but it is not reliable. Changing my approach is usually the lever.

That is how resilience shows up for me.

It is less “positive mindset” and more acceptance: reality is messy, plans break, timelines slip, and the outcome I want might arrive late or look different than I pictured. The only part that stays available, even when things go sideways, is choice. Every respawn is a decision point.

So when I am stuck, I try to shrink the problem down to one move: the next best thing. Not the perfect thing. Not the final thing. The next thing I can do that keeps me moving.

Another thing games keep teaching me, against my will, is that grinding is not always progress. Sometimes I get more unstuck by walking away. I will put the controller down, go live my life, and then some new angle shows up later. Not because I tried harder, but because I changed state. 

That pattern has carried over to how I use AI too. When I hammer at the same prompt and demand the same outcome, I get frustrated fast. When I treat it like a loop I can steer, I get options. If I step away, I often come back with a better question, a different approach.

tl;dr: Play in hard mode; Enjoy new approaches; and touch grass to progress. 

Because these stories are told and retold, in meetings, messages, rituals, and daily interactions, they become a form of control.

In communities, people regulate themselves through stories of solidarity and care.
In partnerships, through stories of reciprocity and equality.
In marketplaces, through stories of generativity and proportionality.
But these moral stories do more than regulate behavior.

They give it meaning.

They explain why people act the way they do.

As such storytelling replaces supervision. They make abstract principles tangible and remind people what it means to be a good member of the organization.

And what it means to be a “good member” takes three very different shapes.

In a recent discussion about how money is the agent of coordination between humans, I went down a rabbit hole of trying to uncover what might be other ways in which humans can be coordinated.

It led me to this very interesting discussion around organizations and how they can be organized with culture – a set of shared stories that reinforce the behavior you want from society.

Great read.

Found this really cool video of an iOS engineer showing real time specular shading based on the lights around you. Look I am not a fan of Liquid Glass in its current instance. However, I am still a fan of technical feats. Apple first introduced this in Apple wallet to cards and esp the Apple Card. Now they’ve made it part of the OS.

Design considerations aside, the technical feat is impressive.

Much like the genetic engineering behind Jurassic Park was impressive.

“I began to realize how important it was to be an enthusiast in life. He taught me that if you are interested in something, no matter what it is, go at it at full speed ahead. Embrace it with both arms, hug it, love it and above all become passionate about it. Lukewarm is no good. Hot is no good either. White hot and passionate is the only thing to be.”

The Devotion-Friction Matrix of devoted action has four states:

Flow (high devotion, low friction) where action feels natural and repeatable;
Strain (high devotion, high friction) where caring is high but the cost of doing it is high too;
Coasting (low devotion, low friction) where you keep going mostly because it’s easy;
Avoidance (low devotion, high friction) where the task feels both unrewarding and hard to start, so it gets postponed or completely dropped.

This is a perfect visual to the state of play.

It’s a teenage rite of passage to explode into rage at your parents. While the usual outburst is sparked by some combination of hormones, insecurity, and authority issues, for me it was a popping sound in my father’s jaw. I first noticed it at the dinner table. Every time he took a bite, the disc of cartilage that cushioned his jawbone would slip out of place and snap back. Chew, click, chew, click. Like a drum, his mouth reverberated the sound, which changed in pitch each time he opened to take a bite. Layered beneath all of this was the wet percussion of normal chewing. The trio — jaw pop, meat squish, fork scraping teeth — became inescapable. And it drove into me, first through my chest, a surprising shock of affront and disgust that then suffused through my whole body. It was the first time I ever got scared that I wasn’t in control of what was inside my own head.

It hits fast.

A second I am fine. The next, a wet, grinding chew or a hard break of someone taking _in_ their sniffle cuts through everything. It is not “a little annoying.” It is a five alarm buzz in my body. My skin goes tight. My jaw locks. My stomach turns like I tasted milk gone bad. Heat crawls up my neck and into my face. My heart starts doing that stupid, hard thump that feels louder than the sound itself.

The noise feels intimate, in the worst way. Too close. Too detailed. Like someone put a mic really close to the other person and a speaker next to my ear and turned up the gain. Each smack is sharp and slow at the same time. Each click lands with a tiny pause after it, like my brain is forced to replay it and confirm, yes, it happened again.

Disgust shows up first, quick and clean. Earlier my face used to contort into this mess. These days, my eyes immediately lock to the origin of the sound. Then anger follows like a door slammed in my chest. Not “I’m irritated” anger. The kind that wants to shove the world away, the kind that makes my hands want to do something, now, just to make it stop. And straddling both is panic, because I can feel myself losing control. I can watch myself getting pulled into a reaction that I do not want to share.

Then the shame comes in, right on time, and whoo!, it makes everything worse.

Because it is a person I love.

So now the sound is not just a sound. It is a test I am failing in real time. I am angry at them for making the noise, and angry at myself for being angry. I start bargaining with my own brain. Be normal. Be kind. Don’t be weird. Don’t be cruel. Just sit here. Just breathe. And the harder I try to be calm, the more the sound digs in, like it knows where the soft parts are.

I can feel my face trying to hold a neutral shape. My teeth press together. My shoulders creep up. My eyes lock on some point that is not them, because if I look at them I might snap, and I would rather bite my tongue than hurt them. I get this sick, trapped feeling, like being stuck in a room with a dripping faucet that is also an insult.

I am not choosing any of it.

That is the worst part.

It is not a preference. It is not a pet peeve. It is my nervous system grabbing the wheel.

So I reach for escape the way you reach for a railing when you miss a step. Headphones. Noise canceling. Leaving the room. Anything that puts a wall between me and that sound before it turns me into someone I do not want to be.

And then I hate myself for needing that.

I hate that love is sitting right there, chewing, and my brain treats it like a threat.

I hate that I have to protect both of us from a sound.

I am heaving a sigh of relief. Much like the author says, this is the first time I can put a word to the feeling that I get.

It’s misophonia.

The Quicksilver inspiration is clearly visible and that makes me so excited. I am downloading it as type this post and I am genuinely thrilled to see if it is a replacement for Alfred 5, which is my favorite launcher.

Right now, I will admit that it’s the quicksilver – subject > verb semantics that is different from Alfred, which allows is clearly via workflows, but is limited to pre-set subjects and actions.

The “voice” mode is also very interesting as a great way to quickly enter transcribe text.

A “sinister diagonal” experiment.

I really enjoy people showcasing their capabilities especially their creative skills. The only thing that will make me happier is if I get a chance to meet the person behind the craft to share how amazed I am to their face and that there are dozens of us who appreciate their craft, their creativity and their skills.

My grandfather drilled this into me, “Sing like no one is hearing. Draw like no one is watching. Write like no one is reading. A skill is best done when the act is innocent.”

I wonder how he might process our age of social peacocking. Whatever his faults, lacking authenticity or conviction wasn’t one of them. Maybe, that in itself, is the skill I practice when I follow his words.

NetNewsWire 1.0 for Mac shipped 23 years ago today!

Here’s where things are on this particular February 11: we just shipped 7.0 for Mac and iOS, and now we’re working on NetNewsWire 7.0.1.

After a big release, no matter how careful we are, there are often some regressions to fix and tweaks to make right away, so we’re working on those. Here’s the milestone with the current to-do list.

NetNewsWire

@NetNewsWire@micro.blog

NetNewsWire Turns 23: netnewswire.blog

February 11, 2026, 17:35

I’ve tried miniflux and it was great. However, I always keep coming back to NetNewsWire because the Mac is my platform of choice and NNW is the best RSS reader for the Mac. Yes, even after 23 years. This is especially true as my home server is not necessarily something that I can keep running once I no longer have a home.

OpenAI’s first hardware release will be a smart speaker with a camera that will probably cost between $200 and $300, according to The Information. The device will be able to recognize things like “items on a nearby table or conversations people are having in the vicinity,” The Information says, and it will have a Face ID-like facial recognition system so that people can purchase things.

This could also help explain why OpenAI hinted that they weren’t interested in the Apple deal anyway

OpenAI declined to comment. But a person close to the company said it had taken “a conscious decision to not become the custom model provider for Apple” in the autumn of last year and instead focus on building its own AI device to leapfrog the big tech companies.

https://www.ft.com/content/8033b1bc-4ffe-47ed-baf0-5abea6a1322a

While this is going to be all very interesting, I am not really sure what the best product here will even look like given we are early in the LLM era.

After planning to include the new capabilities in iOS 26.4 — an operating system update slated for March — Apple is now working to spread them out over future versions, according to people familiar with the matter. That would mean possibly postponing some features until at least iOS 26.5, due in May, and iOS 27, which comes out in September.

There are 2 things to takeaway from this.

1. Apple’s likely over correcting – once smitten, twice shy style.

However, I like Apple products because they are polished. As a patron, I will accept the delay as the cost to pay for polish. I am not a patron of Siri, so I will reserve judgement. Yet, I do trust Apple’s judgement of when a product is “done.”

Apple is very set in their old processes and release cycles. But OpenAI ships major new features multiple times a month. I don’t see how Apple can be competitive in AI unless they rethink how they work and release software.

Manton Reece

2. I don’t know what to think of this. It is true that just like Google, facebook, Amazon and their ilk, Open AI is also a fast shipper.

Yet, I also akin their product polish to that of Google and their ilk. They are good enough, for most purposes, but they often show their jagged edges a LOT more than Apple products (exception: Siri).

I think LLMs allow gains to develop features faster. However, based on my experience I don’t see any time gained in the chasm between “good” and “polished.”

I am open to change my mind if you have counter examples. Please let me know.

I’d wondered why one of my favorite shows suddenly disappeared after a blisteringly good season. This video details how Swat Kats came together, how it was shivved because of corporate politics and American hypocrisy of “violent TV.”

Swat Kats came at a special place in my life and I credit it to sparking my love for heavy metal and grunge music.

However, much like most great things – they are making a comeback. In a surprise twist, the creators managed to hold on to the rights of the original IP.

I am eagerly looking forward to the comic and what it might ignite. Add this to my 2026 anticipation list.

Probably one of my favorite pictures I’ve taken in a while. Although I’ve been shooting quite a lot of keepers lately.