David Chico Pham writes on technology, labor, democracy and power.

I'm a Senior Software Engineer (Privacy). I studied history and sociology in undergrad. I taught myself to code and program while working full time at the Library of Congress. History is a wonderful teacher that has taught me a way to understand the systems we've built, the way they are, and who they serve. Sociology showed me how people actually behave inside those systems. Those turned out to be exactly the right questions for building digital systems. I write about technology, labor, democracy, and power because these are the hidden forces that shape our world and our lives.

As a first-generation Mexican-Vietnamese American, Buddhist from the Midwest, and working-class person, I often felt like an outsider. The concept of "community" seemed abstract and elusive to me. It was something I didn't quite get. I had a hard time grasping its true meaning. However, over time, I've come to understand that "community" is fundamentally about our relationships with one another. This personal site is an entry to the community of indie web creators and thinkers.

This is my next is a new section on this site where I share what I'm learning about technology, concepts, and ideas that shape our world and my own thinking. With Agentic AI like Claude Code, what were once side projects collecting dust can now be fully built personal apps. Currently, I'm exploring NodeJS API Design

I'm simultaneously terrified and elated by Agentic Engineering. Its ability to scaffold and build out apps that fit a person's needs and execute ideas is incredible.

Chico was my nickname in school.

StoryStream

Findings on the web, interesting stuff, or beautiful things to share

Google is set to use your photos to train it's AI. As ravenous information thieves, they are counting on you to leave your photos in place. In turn they will take their long straw and drink your milkshake. You can stop them from using your photos of your children and prevent them from having them forever.

Except bulk deletion is really time consuming. Below is a script you can copy and paste into the developer's console ( cmd + option + i). What it does, as you scroll through your photos, it will select all of them that are visible on the screen.

I personally have over 64,000 photos. I have been deleting in 5k batches at a time. It takes about 5 minutes to scroll that many photos and select them. The real drawback is having to look through the photos. Stopping too long will end the script. You can fire the script and again by hitting the up arrow and enter to execute it.

Be sure to click on the trash button. It will take a few seconds before you see it going through and deleting the photos.

(async () => {
const delay = (ms) => new Promise(r => setTimeout(r, ms));

// Find all unchecked photo checkboxes
const getUnchecked = () => document.querySelectorAll(
'div[role="checkbox"][aria-checked="false"][aria-label^="Photo"], ' +
'div[role="checkbox"][aria-checked="false"][aria-label^="Video"]'
);

let total = 0;
let lastHeight = 0;
let stableCount = 0;

while (true) {
const unchecked = getUnchecked();

// Click each unchecked checkbox
for (const cb of unchecked) {
cb.click();
total++;
// Small delay to avoid overwhelming the UI
if (total % 50 === 0) await delay(100);
}

console.log(`Selected so far: ${total}`);

// Scroll down to load more photos
window.scrollTo(0, document.body.scrollHeight);
await delay(800);

// Check if we've reached the bottom (page height stopped changing)
const currentHeight = document.body.scrollHeight;
if (currentHeight === lastHeight && getUnchecked().length === 0) {
stableCount++;
if (stableCount >= 3) break; // Confirm we're really done
} else {
stableCount = 0;
}
lastHeight = currentHeight;
}

console.log(`✅ Done. Total selected: ${total}`);
})();

Its important to take the view that Google and the rest of the rotten lot are in the business of extracting as much data out of you as possible. They are strip miners. When you use any of their "free" products, you are performing free labor. Take your power back.


AI has limits, even if many AI people can't see them. Henry Farrellgoes beyond the good and bad of AI, and instead shares his notes on the book, Irrational Decision: How We Gave Computers the Power to Choose for us by Benjamin Recht. One is how do we think about machine learning:

"And machine learning itself is no more and no less than a powerful statistical tool. I found this passage maybe the most clarifying explanation of what it does that I’ve ever read."

To frame the prototypical machine learning problem, I like to think about a hypothetical spreadsheet. Each row of the spreadsheet corresponds to some unit or example. But I don’t care what the units mean. I just know that I have a bunch of columns filled in with data. And I’m told one of the columns is special. I am about to get a load of new rows in the spreadsheet, but someone downstairs forgot to fill in the special column. Management has tasked me with writing a formula to fill in what should be there. For whatever reason, I don’t get to see these new rows and have to build the formula from the spreadsheet I have. The formula can use all sorts of spreadsheet operations: It can assign weights to different columns and add up the scores, it can use logical formulas based on whether certain columns exceed particular values, it can divide and multiply. … I’ll do an experiment. I’ll take the last row of my spreadsheet and pretend I don’t have the special column. I’ll write as many formulas as I can. … But why single out that last row? I can do something similar for every row! I’ll invent a set of plausible functions. I’ll evaluate how well they predict on the spreadsheet I have. I’ll choose the function that maximizes the accuracy. This is more or less the art of machine learning.

Finding the sweet spot in AI and its limits:

As per the quote at the beginning of this post, Ben doesn’t really engage with the question of whether AI is good or bad in any general sense. Instead, he proposes that it can carry out many tasks, including tasks that we might not anticipate right now, but that there are limits. AI, like mathematical rationality more generally, has a sweet spot: problems that are complicated enough that they can’t be solved by other computationally cheaper approaches, but that have enough regularities to be workable. Within that sweet spot, it can do extraordinary things. Outside the sweet spot, it may be redundant or completely useless. And there is an ambiguous zone in between, where it can do stuff but imperfectly.

I found this pull quote very much related to the Software Brain that Niley Patel argues it has taken across my profession:

"Equally, there are challenges that appear to be fundamentally resistant to mathematical rationality, including bureaucracy and politics:"

societies are not computer chips. While I noted in chapter 2 that computer chips were often analogized as microscopic cities, chips were always designed to be hermetically sealed and perfectly controlled. This is what made them optimizable. Real societies, on the other hand, had people. While it’s convenient to model and view the population, its health, and its market flows as mathematical abstractions, these run into the limits of the messiness that people bring to bear.

AI hype machine really wants you to believe they can solve for the messiness in human enterprise, creativity, and ingenuity. At the same time these propagandists know very little what normal people do in general and really out of touch with everyday people. It's why they haven't made a good product out of billions of dollars in AI.

This piece is the most insightful thing I’ve read so far about when AI is useful and when it isn’t. This piece assuages me a little in that AGI is not going to be coming from a fancy statistical model. I ordered the book soon after I read the piece. Here's an affiliate link to it.


Fujifilm xt-3 with 27mm

Photography has been a hobby for me since I was a teenager. I recently purchased a Godox im30 iflash module. Most of my photography has been in iPhone era and if there was something to avoid was to turn off flash.

While scrolling on Instagram, a few accounts posted photos that look straight out of a 90's magazine. I love that aesthetic.

Here's one of my favs of my azaleas

Photography brings me so much joy. I don't know why, but I really like reviewing the shots I took and finding real excellent shots.


Several things that I am pretty excited for.
  • The architecture in taking your voice recording and translating that to text using a local LLM is a perfect fit for AI
  • It's very unobtrusive in taking a note on an ephemeral thought.
  • The concept to tools in AI engineering is using this Model, Context, Protocol (MCP) to provide different actions an LLM can do. We think of them as as apps, like alarm or weather. But to an LLM with no understanding beyond a pattern recognization and token generation, it can pick up on "alarm" in the context of a "setting a alarm".
  • The ring, Index 1, is entirely hackable, open source, and private.
  • That means I can build something on it.

We Made $75 External Memory For Your Brain! [Tick Talk w/ Eric Migicovsky]

Lucas discovered meditation without calling it meditation.

I tried this out and it does work in reseting your brain.


Noam Scheiber and Adam Conover discuss the tech layoffs in the past 3 years are untenable. The call for organizing and unionize has increased in places that were really resistant to it. The best time to organize was yesterday. The right time is now.


It's a Saturday night. Hearts are going to shine her eyes. She can make it. She's got it. She can't lose.


Roxy by Yuko Rade [Yuko Rade - Youtube]

Zack Beauchamp, coming back from months of looking how autocrats lose, writes a really fascinating piece in stopping fascism before it usurps democracy. Places like Brazil, South Korea, and Poland identified threats to their self-governance and democracy. They all defeated would be autocrats. But how? Autocrats deeply care about democratic facade and work very hard to appear to be democratic, including popular support.

This argument for strategic moderation raises a question: Why do would-be dictators care so much about having a pretext? If Chávez wanted to seize power, why didn’t he just do it?

Because when they do, Gamboa notes, the backlash is overwhelming. She cites, as one example, Peruvian President Pedro Castillo’s 2022 declaration of a state of emergency. The move was so obviously authoritarian that it galvanized Peruvians, and the international community, to act — leading to Castillo’s impeachment and arrest on the very same day as his attempted power grab.

Smart authoritarians, Gamboa notes, “have learned that they can avoid this kind of backlash and maintain a democratic facade by undermining democracy gradually instead.”

But if enough people see through the facade, the fact the regime is fascistic or authoritarian, making legible to mainstream, creates a weakness that an opposition party can go after.

Virtually every expert on backsliding has observed that would-be authoritarians value a democratic facade. But too few appreciate the implication: that autocrats have good reason to believe their project will fail if too many people see it as authoritarian, creating a point of vulnerability their opponents can exploit.

Beauchamp goes on to say that our modern day tyrants burn a lot calories and spend a lot time actively trying to convince their supporters that they are playing by the rules. This has important implications. If keeping up a facade is so important, that would mean that people really care about living under a democracy.

But I think the government’s success at building and maintaining a democratic facade is one of the most important factors, simply because many people genuinely care about preserving democracy.
...
But the legibility of the threat matters much more than most people give it credit for, and it has played a decisive role in some of the biggest cases of democratic resilience in recent memory.

This is a must read piece and we have a chance to cancel the fascist takeover of this 250 year old republic.


They are main vehicle to which people can achieve a middle class life. From the 1950's till the 1980's America saw the biggest expansion in the middle class. Nearly half of the population was in it. It had a lot to do with unions, the power unions brought to the workplace, and earning a fair wage.


Subscribe to the Storystream RSS Feed or view the entire archive for more.

This Is My Next

A developer's notebook — to learn is to disrupt the inertia in one's own mind.

  1. NodeJS API Design
    April 27, 2026
  2. Edmund Edgar - Bluesky, atproto, how to hack on it and why it matters
    March 31, 2026
  3. Day 1: Setting Up This Is My Next
    March 15, 2026
  4. Why I'm Starting This Section
    March 10, 2026

View more entries...

Essays

Loose, vague feelings on things I don't entirely understand yet.

  1. Power at Work Podcast - Washington Post Tech Guild's Hazel Court and David Pham
    March 1, 2026
  2. You Will Awaken
    November 28, 2025
  3. Sleeping Deeply and Dreaming
    September 26, 2025
  4. Above the Law as intended
    August 18, 2025
  5. AI is theft of the mind
    June 22, 2025
  6. Washington Post Tech Guild
    June 11, 2025
  7. Where Do We Go From Here
    November 29, 2024
  8. Propaganda Primes for Cruelty
    November 13, 2024
  9. The Fifth Risk is Here
    November 9, 2024
  10. Hopeful Heartbroken Man
    November 7, 2024
  11. The Expensive Education of Jeff Bezos
    October 29, 2024
  12. Erin Kissane's Work
    October 12, 2024

Subscribe to the RSS Feed or view the entire archive for more.

Turbocharged hyperlinks

Chaos desires order -- links to bring clarity

Subscribe to the turbocharged hyperlinks RSS feed or view the entire archive for more.

You can find out more about me, what I use, the books that I have been reading and what I am up to now.