![\"Linkedin](\"https://www.bryanbraun.com/assets/images/michael-neal-ai-apps.jpg\")
I’ve been on a posting spree recently and I’m excited to get out some of these links I’ve been sitting on for a bit. Lots of fun ones here—enjoy!
A little platforming game that takes place inside recursive spaces. As a player, I found it beautiful, trippy, and satisfying to figure out. As a software developer, my mind reeled, wondering how the creator made it. In the description, he credits a 3Blue1Brown video, which spends a fascinating 45 minutes describing in rich detail how to represent an MC Escher painting on a mathematical plane.
The whole thing feels amazing to me. When I take a step back and think about how this was built, I see this chain of innovators, each sharing their work and inspiring the next person. Escher, who used his intuition of spatial reasoning to twist reality. The Dutch mathematicians, who found a way to describe it mathematically. Grant Sanderson, who broke it down visually, so anyone with a high-school math education could understand. And Daniel Linssen, who thought, “what if I made the mathematically-warped grid into a playable game environment.” I can only imagine that Escher would have loved to see the worlds he imagined constructed into environments that he could walk through and explore. I loved how the game silently teaches you how to exploit the unique properties of those environments, to help achieve a goal.1 As a player, you get to discover the properties of visual recursion, almost as if you were the next link in the chain of innovators. It all builds up to an experience that only a video-game can deliver.
I’m a big fan of doing something exhaustively (“every possible way”) as a way to get in the reps to learn something, and this is a great example of that. Each of these demos was so visual that it was fun to browse through them all. I’m way impressed that blinry built them all in one weekend (pre-AI, too).
People don’t destroy fortunes with frivolous spending and consumption—they do it with bad investments. It’s the same with time, our life’s currency. For diligent people, we’re much more likely to waste it doing fake work, than being lazy. An important reminder!
A local Seattle artist holds a funeral for a closed down Taco Bell. I learned about this from my brother (who attended in person)! The video took a turn when he documented flying to China to make bronze statues but it actually ended up being one of my favorite parts. Such an interesting window into another world.
It’s more of a puzzle than a game. It’s a little tricky but you don’t have to be a chess-person to solve it, and it’s satisfying once you do.
.gitignore Isn’t the Only Way To Ignore Files in Git
I expected this to be an impractical exploration of some esoteric git features (git has a lot of those), but I was pleasantly surprised. I have been looking for a consistent place to put AI plans without checking them into git (across multiple repos), and this gave me several good options.
Kyle Mathews discusses how AI changes the job of programmers, not as an unprecedented event, but as one in a series of historical advances that has changed the job of programmers approximately every 20 years. I like this framing, and it aligns with what I’ve always told myself about this industry: It’s always changing, so you need to be constantly learning. Yes, it takes effort, but it’s always interesting and you’ll never be bored.
It's what I loved about Miegakure (or at least the idea of miegakure). The intuition-building. ↩
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2026/07/15/this-could-be-the-golden-age-of-the-internet-why-not", "url": "https://www.bryanbraun.com/2026/07/15/this-could-be-the-golden-age-of-the-internet-why-not/", "date_published": "2026-07-15T00:00:00-04:00", "date_modified": "2026-07-15T00:00:00-04:00", "title": "This could be the golden age of the internet. Why not?", "summary": "The predominant narrative these days is that generative AI is going to be bad for the internet. Whether you call it Dead Internet Theory or The Slopocalypse, the idea is the same: A tidal wave of low-quality AI-generated slop will flood the internet, turning it into a lifeless shell of it’s former self.1", "content_html": "The predominant narrative these days is that generative AI is going to be bad for the internet. Whether you call it Dead Internet Theory or The Slopocalypse, the idea is the same: A tidal wave of low-quality AI-generated slop will flood the internet, turning it into a lifeless shell of it’s former self.1
Maybe they’re right. I’ve certainly seen bad slop content on social media.
But I’ve also seen stuff like this:
I think there’s a decent chance that we’re about to enter the golden age of the internet, and posts like the above help explain why.
The internet is more than just content… it’s software. It’s email clients, Pomodoro timers, drawing tools, and games. Historically, all of this software was painstakingly built by hand, which limited what we could build and publish.
But the barrier just got lower. All kinds of interesting people can build things now that couldn’t before. Artists, school teachers, sports enthusiasts, farmers, 11-year olds, scientists, storytellers, and more. Even if you’re an experienced senior engineer, AI lowers the threshold of what’s worth your time, so you can pop out goofy little toys and experiments.2 The internet is better for having all these things in it.
Apps aren’t just easier to build, though. They’re easier to maintain. I remember the frustration of pulling out my laptop on Saturday evening to work on a side project, only to spend four hours fighting to upgrade dependencies without breaking anything. That pain limited the number of active projects I could manage at once. But when an agent can upgrade all your dependencies in ten minutes, those limitations go away.
Similarly, the limitations around cross-platform development have shrunk dramatically. I’m a long-time JavaScript dev, but for the last two weeks at ClassDojo, I’ve been writing Swift, shipping features to our iOS app. This wasn’t practical until AI lowered the barrier for cross-platform work. Now, with the help of AI, a solo engineer can convert desktop apps into cross-platform apps, native apps into web-apps, or websites into progressive web apps.
To summarize, it looks to me like we have all the ingredients for a renaissance of little apps:
Even if there’s a proliferation of new, interesting, web software, there are several arguments for why it will fail to improve the internet we have today. Let’s look at a few of them:
AI generated content is kindof a bummer, but AI generated apps don’t have to be. A software tool has the same utility to the end-user whether the code was written by an AI or by hand. The end-users of software don’t read the code and don’t care about the code. They care whether it solves their problem.
Now maybe you think that AI-generated software is inherently buggy. Software has always been buggy, but to the degree that you feel AI-generated software is worse, consider the forces at play. Users hate buggy software, and in a world with lots of options, fast and stable software will attract the most users. We’re all still learning how AI will contribute to that outcome, but the point is that Slop is Not Necessarily the Future.
Maybe you think that AI-generated software is inherently tacky because it all looks the same. Us web-devs can be discerning about that kind of thing, but normies don’t care. All of these UIs look better than what the median software UI looked like 10 years ago.
And for the creators that do care, AI lets them put their time where it matters. Three years ago it might have taken you twenty hours to build a portfolio site. Today, you can have AI build out the plumbing in less than an hour, and spend the other nineteen hours iterating on the aesthetic. An internet where more people are doing that feels to me like an objectively better place!
I believe that the quantity of good stuff on the internet will continue to increase, but does that even matter if we can’t find it, amidst all the incoming garbage?
Of course, the vast majority of the internet has always been garbage, but It never really mattered, because of Search.
Now maybe you don’t trust Search to save us anymore. It’s common these days to see articles about how Search is broken, search result quality is declining, and so on.
But Search is rapidly changing, right now, for the first time in a long time. AI gives competitors an actual shot at disrupting Google, who no longer has the luxury of trading user experience for higher profits from their monopoly position. Now, I don’t know how this is all going to shake out, but if anyone can consistently surface the best resources the internet has to offer, I expect they will be rewarded. The incentives are there.
How can the future internet be good, when it feels like social media has taken over, polarized everything, poisoned the public discourse, and given the rising generation a mental health crisis?
It’s a fair question. I currently feel like social media is net negative to society. Not much to sugar-coat there.
It may be helpful to think of social media like a virus. It’s caught our culture off-guard and it’s making us sick. But the body won’t tolerate sickness forever, and our cultural immune system is working on anti-bodies that will get us to a better place. Last year, my daughter’s high-school implemented a school-wide ban on phones during the school day. The EU is experimenting with increasingly strict social media legislation. And individuals (including me, and many in Gen-Z) are finding their own ways to opt-out of the nonsense.
Hopefully this trend continues and leads to better times ahead. The worst may already be behind us.
Some AI boosters think that all websites, apps, and software will soon become extinct… replaced by powerful AI interfaces that can generate whatever you want (apps, resources, entertainment), on-the-fly. I don’t buy it (and I’m not alone).
Anyone who has used a chat or voice-based AI for a non-trivial amount of time knows how exhausting it can be to get the specific thing you want. The user has to initiate a request, accurately understand their own needs, and then accurately communicate them. All of that is cognitive load.
In contrary, a well-designed piece of software serves as a domain-expert for a need, anticipating the user’s sticking points, and visually helping overcome them. The result is cognitive relief.
We could say more about this, but if you’ve ever felt AI fatigue, you already know. We’ll still need plenty of web software in the future.
Ask ten people when the “Golden Age” of the internet was and they’ll give you ten answers. It has more to do with how old you were when you went online for the first time than any objective criteria (it’s similar to people’s taste in music, in that way).3 So I understand that predicting an upcoming “Golden Age” of the internet is a fraught position to take. Nostalgia is powerful and the old internet is never coming back.
But change is not always bad. It’s the wellspring of energy, enthusiasm, and curiosity. There are a lot of excited people out there right now. It won’t always be like this.
If we want to see the good kind of change, then we have to build it. We can use AI to flood the internet with interesting little toys and apps. We can use our time savings to experiment and build things from scratch that AI can’t imagine. And we can continue to connect and share in healthy and positive ways—it’s never been easier to spin up a blog and share what you’re working on.
With a bit of luck, we’ll look back on today as a uniquely special time. Why not?
![\"I](\"https://www.bryanbraun.com/assets/images/good-old-days.png\")
Example arguments include The Internet Will Be More Dead Than Alive Within 3 Years, Trend Shows and Maybe AI Slop is Killing the Internet, After All. ↩
Case in point, my own goofy little joke-project: pro wrestler or roller-coaster ↩
I'll bet anything that 15 years from now, 30 year-olds will be wistfully remembering the good ol' days of neal.fun and \"early Mr. Beast videos\" in the same way that my generation feels nostalgic for Google Reader and Homestar Runner. ↩
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2026/07/11/you-are-not-your-grand-plans", "url": "https://www.bryanbraun.com/2026/07/11/you-are-not-your-grand-plans/", "date_published": "2026-07-11T00:00:00-04:00", "date_modified": "2026-07-11T00:00:00-04:00", "title": "You are not your grand plans", "summary": " “You are not your grand plans. You are your daily patterns.”", "content_html": "“You are not your grand plans. You are your daily patterns.”
I love this quote from James Clear because it hurts me in all the right ways.
At my worst, I feel like my daily patterns are at war with my grand plans. When I’m fighting a daily battle and willpower is slipping, I need every reminder that this moment, right now, is critical.
Because it is!
When I’m struggling with daily patterns, it’s helpful for me to look at the times I was rocking it. I’ve had periods of my life when my workouts were really consistant, or I journaled every day, or I always woke up early. What contributed to those things?
Often, it was social pressure, an accountability buddy, or a change of environment that pushed me into a productive place (underrated benefit of having a blog: it’s a useful historical record of your successes and failures).
Of all the things I might learn looking back at these experiences, perhaps the most important is that it’s never been easy. I’ve got 15+ years of posts documenting the battle. Sometimes I’m winning and sometimes I’m losing, but I’m always fighting, and that’s a daily pattern too.
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2026/04/27/quality-vs-quantity-in-the-age-of-ai", "url": "https://www.bryanbraun.com/2026/04/27/quality-vs-quantity-in-the-age-of-ai/", "date_published": "2026-04-27T00:00:00-04:00", "date_modified": "2026-04-27T00:00:00-04:00", "title": "Quality vs quantity in the age of AI", "summary": "Over the last four months I’ve increased my use of AI for web dev work pretty dramatically. I use it every day at work, as do most of my coworkers. I don’t think it’s disputed anymore that AI-assisted development can save a lot of time and effort. The question is, what do we do with all that surplus?", "content_html": "Over the last four months I’ve increased my use of AI for web dev work pretty dramatically. I use it every day at work, as do most of my coworkers. I don’t think it’s disputed anymore that AI-assisted development can save a lot of time and effort. The question is, what do we do with all that surplus?
The standard narrative in the industry is to use it to go faster. Ship three features a day instead of one. Get a half-dozen agents working in parallel. 10x your productivity. If you don’t move fast, your competitors will!
I get it. Speed is good and it has benefits beyond just finishing more stuff per unit time. Speed matters!
But we live in an era of extreme consumer choice, especially in the world of software. Even pre-AI, there were dozens of good email clients, to-do lists, fitness apps, and analytics tools, all chock-full of useful features at reasonable prices. What is “going faster” going to give us? Even more options, with even more features?
Apparently yes. In the past quarter, iOS app store submissions have risen by 84 percent, and Hackernews, “ShowHN” submissions have nearly tripled. We live in an era of software abundance.
But as software grows increasingly abundant, attention grows increasingly scarce. Attention was scarce even before AI but it’s going to get worse. Consumers have more choices than ever, and increasingly those choices are becoming undifferentiated as the industry uses AI to produce them.
Instead of using your AI surplus to go faster, what if you used it to go deeper?
What if, instead of building ten good features, you built two amazing ones?
What if, instead of blasting out a dozen landing pages, you poured your attention and care into a home page that knocked someones socks off—one that could not have been built by an AI, because nothing like it could have been found in the training data?
What if you doubled-down on performance? On accessibility? On user-experience? On creating a strong personal touch?
It would stand out! That matters, arguably, even more than speed.
The more I thought about this, the more I recognized it as the same ol’ quality vs quantity debate, where it pays to strike a balance:
But what’s the ideal balance? Is it 90% quality, 10% speed? 50% - 50%?
How would you spend your AI surplus?
For an example, see the story of the photography class, as told by James Clear in Atomic Habits. ↩
Remember: Slow is smooth, and smooth is fast. ↩
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2026/03/30/links-14", "url": "https://www.bryanbraun.com/2026/03/30/links-14/", "date_published": "2026-03-30T00:00:00-04:00", "date_modified": "2026-03-30T00:00:00-04:00", "title": "Links #14", "summary": "See below for links to recent things that have made me think. It’s quite the mix of mediums (animated videos, short posts, letters, and 6-hour podcasts), but it wasn’t until I gathered them that I noticed a recurring theme: sacrifice", "content_html": "See below for links to recent things that have made me think. It’s quite the mix of mediums (animated videos, short posts, letters, and 6-hour podcasts), but it wasn’t until I gathered them that I noticed a recurring theme: sacrifice
Sometimes the world imposes constraints on us, forcing us to choose what matters most. What do we do when those moments come?
Some of my favorite stories (both ancient and modern) explore these moments, and you’ll find elements of that in the list below.
What’s Really Going on in El Segundo
A 14-minute video about the hardware startup scene in El Segundo California, by Jason Carman. Some of it is hype, of course, but I’m impressed by the vision and culture of optimism they are offering to a world where so many young people are flailing and drowning in zero-sum status games.
When a young man or woman comes up to me and says, “I’m not sure exactly what I want to do with my life,” “I’m having a really hard time orienting myself,” “I’m not sure what the point is, but I want to build something great and meaningful,”—that is one of the best consequences of everything that we’ve done. Making more water is a big deal, energy abundance is a big deal, …all of this is great. But just being able to give people some vision for the future—that is one of the most important things. We have to sacrificially build hard, real, material, things to create a future worth living in.
I’ve been watching a bunch of Jason’s videos, and they’re all great medicine for the negativity that afflicts most news media.
Derek Sivers describes his new home in the woods, an alternate world without AI, connectivity, or mental shortcuts. I know people have been been retreating like this for centuries, but the craziness of the world, and the busy-ness of my current life make such a retreat feel like a luxury.
All it takes is for one to work out
A short post about those moments when the rejections keep coming.
These processes – college admissions, job searches, home buying, finding a partner – can be emotionally brutal. … All it takes is for one to work out. And that one is all you need.
During my job search last year, I had a really exciting opportunity fall through at the end. It felt like a crushing blow. It would have been so perfect, I lamented. But looking back now, I don’t even care. Another one worked out (a better one, even). Knowing that rejections don’t actually matter is great medicine and a legitimate source of hope for anyone still fighting.
Dan Wang, author of a bestselling book about China and America, uses his end of year letter to compare the two countries and offer some thoughts. His writing struck a chord for me. It was informed and clever, full of interesting takes that demonstrate his deep understanding of the two nations. I added his book (Breakneck) to my list.
A cute animated short story with a powerful message about sacrifice. I watched it twice—once with my wife, and again with my kids. Both times it sparked some great discussions. I think everyone should watch it.
Sam Kriss profiles the newest, most eccentric, batch of Gen-Z founders. The way he presents their companies make them look exactly like what the fevered minds of a generation raised on social media algorithms would produce. Extreme, attention-grabbing, and impossible not to look away. Stacks of Labubus. Sperm racing.
Reading this felt like I was being transported into another dimension—one where the most absurd, amoral corners of online culture have escaped into the physical world. Is it real or fake? For the luls? Can anyone even know?
Along the way, he was also able to work in this fantastic quote:
If we ever get AI that is strong enough to basically be God and solve all of our problems, it will need to use the same techniques that the actual God uses in terms of maintaining some distance. I do think it’s possible that the AI will be like, “Now I am God. I’ve concluded that the actual God made exactly the right decision on how much evil to permit in the universe. Therefore I refuse to change anything.”
At over 6 hours, this is the longest podcast interview I’ve ever listened to, and I doubt I would have finished it if DHH wasn’t so sharp and entertaining. DHH is known for his strong opinions and I was impressed at his ability to articulate his thinking with clarity and appropriate nuance throughout the interview. The part where he explains the beauty of the Ruby language (from both a practical and philosophical perspective) made me want to write more Ruby, and many of the other topics got me thinking as well.
When AIs can print job applications for pennies, fill up the internet with thinkpieces, and commit industrial-scale quantities of code, is there anything left for us humans to create? Packy McCormick says: yes. What still matters, is what has always mattered, and that is sacrifice. People still want a connection to the stories they read and the music they listen to. Employers still wants to hire someone who cares. Recruiters never needed good cover letters for their own sake… they needed someone who cared enough to sacrifice the time and effort to write them well. They still need those people, even if cover letters are no longer a good way to find them. My takeaway: even in a world flooded with AI-generated content, there will still be a place for those of us who care enough to sacrifice for our work.
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2026/02/16/250-lbs", "url": "https://www.bryanbraun.com/2026/02/16/250-lbs/", "date_published": "2026-02-16T00:00:00-05:00", "date_modified": "2026-02-16T00:00:00-05:00", "title": "250 lbs", "summary": "Six months ago, while I was setting up a new code editor, I noticed that the default font size felt just a bit too small, when displayed on my monitor. Not a big deal, I just bumped it up one size and went back to my work.", "content_html": "Six months ago, while I was setting up a new code editor, I noticed that the default font size felt just a bit too small, when displayed on my monitor. Not a big deal, I just bumped it up one size and went back to my work.
Then it happened again, this time, when reading sheet music on my phone. “That’s weird,” I thought, as I increased the size. “This didn’t bother me before.”
I had always had perfect, 20-20 eyesight. When I finally realized that I was losing some of that sharpness, I felt a loss.
It wasn’t the first time, though.
About two years ago, I started having wrist pain while typing. On one particularly bad day, I became very worried. “Typing is how I provide for my family. What will I do, if I can’t type?” Soon I discovered that wearing wrist braces at night helped make the pain go away. I remember laying in my bed in wrist braces, thinking, “I’m 36 years old. I’m too young to need wrist braces for the rest of my life,” and I felt a loss.1
After a few losses like these, it would be easy to adopt a narrative that I’m in physical decline, losing capability daily, with all my best achievements in my past. It’s a sad story. A fragile story.
I decided that one way I could push back against this story is to set a new weightlifting goal: a 250-pound bench press.
It was an unthinkable goal back when I benched 200. I didn’t plan on going heavier back then, but I enjoyed my weightlifting routine, so I kept showing up. By the time I hit 235, I realized that maybe 250 was possible. I felt excited and motivated just thinking about it. I wanted to go for it.
But just like last time, I hit a plateau.2 I was able to bench 245 once, last summer, but I spent the next 6 months unable to do it again. As I turned 39, I wondered if I was fighting against time.
And then it happened.
All of us are going to have to learn how to grow old. The aging itself is easy, but learning how to handle the losses is more challenging. Today, I can tell myself, “my eyesight may be declining, but I’m stronger than I’ve ever been.” Next year, who knows. Maybe I’ll get into pickleball.
There are endless opportunities for personal growth, and if we exhaust the supply of physical milestones we can still progress intellectually, creatively, and spiritually. Life offers plenty of PRs to pursue, if we’re willing to look for them.
Fortunately, I didn't end up needing to wear wrist braces for the rest of my life. After some trial and error, I learned that my pain was ultimately caused by poor ergonomics from using my laptop keyboard and trackpad for extended periods. Once I got set up with an ergonomic keyboard and mouse, the pain went away. It's still a bit of a loss (I can't just roll into a weekend hackathon with only a laptop anymore) but I'm glad I can still do the work I enjoy.
↩In my last weightlifting post, I talked a bit about breaking through plateaus. This time, the key intervention was simply eating more calories (something I've avoided in the past in an effort to moderate body fat). Eating more calories increased my body fat a bit, but also helped me build muscle. Professional weightlifters eat a lot and exercise a lot. It's not an efficient use of time but it's fun (especially if you like eating and exercising).
↩Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2026/01/01/made-in-2025", "url": "https://www.bryanbraun.com/2026/01/01/made-in-2025/", "date_published": "2026-01-01T00:00:00-05:00", "date_modified": "2026-01-01T00:00:00-05:00", "title": "Made in 2025", "summary": "Here are some things I made in 2025:", "content_html": "Here are some things I made in 2025:
This was the first year I really leaned into AI-assisted development, and a couple of my projects reflect that (like “Pro-Wrestler or Roller Coaster,” which I probably wouldn’t have made if I had to do it from scratch). I briefly wrestled with the question of “did I truly make it if I had AI assistance,” but I think the spirit of this post is less about the coding and more about the drive to be creative. As I get older and find more things vying for my time, I don’t want to let go of that desire. I’m finding that AI is actually helping me continue to exercise that creative muscle without requiring so much manual labor.
Happy new year, friends.
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2025/12/27/raise-your-standards-for-important-things-lower-them-for-unimportant-things", "url": "https://www.bryanbraun.com/2025/12/27/raise-your-standards-for-important-things-lower-them-for-unimportant-things/", "date_published": "2025-12-27T00:00:00-05:00", "date_modified": "2025-12-27T00:00:00-05:00", "title": "Raise your standards for important things; lower them for unimportant things", "summary": "I think we’ve all got some built in tendencies, pulling us towards certain behaviors, whether or not they are worthwhile.", "content_html": "I think we’ve all got some built in tendencies, pulling us towards certain behaviors, whether or not they are worthwhile.
It takes many different forms. If you have OCD cleaning tendencies, you spend excessive time making sure your windows are spotless. If you are money-focused, you’ll spend way too much energy clipping coupons and watching account balances. If you’re a fitness junkie, you end up spending a lot of money on specialized equipment and supplements for increasingly marginal benefits. If you are a yard guy, you could be outside all weekend, getting the perfect edges.
Of course, who am I to judge how you spend your time? If you’re truly doing what you want to be doing, then more power to you.
But if you’re like me, and you know in your heart of hearts that you’re wasting your time, then what do you do?
One of my favorite books is “The Obstacle is the Way”, by Ryan Holiday. The central message of the book is that, yes, our lives are full of obstacles but instead of just patiently enduring them, we can flip our perspective and see them as our advantages—secret weapons we can leverage to accomplish our goals. It’s an empowering idea (and one that we discuss often in our home).
The “tendencies” I described above go by many names: passion, obsession, fixation, perfectionism, “hyperfocus,” etc. Call it what you want, but whatever you call it, I’d encourage you to see it not as an obstacle, but as a superpower. After all, you care about something. You just need to channel it towards something important. For that, I find this phrase helpful:
Raise your standards for important things; lower them for unimportant things
When is a room clean enough? When is a meal healthy enough? We all have standards for these things, and our standards can be adjusted. Maintaining high standards is expensive, and any effort we spend on high standards for unimportant things is ultimately a waste.
I could double-down on my March Madness research or I could play a board game with my kids.
I could remove every last dandelion from my lawn or I could publish that blog post I’ve been thinking about.
How do we know what’s important and what’s unimportant? There’s no right answer. You get to decide!
For me, the most important things are the things that last the longest: family, health, and personal development (like building skills and exercising creativity). Anything else can afford to degrade a bit.
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2025/10/28/SolidGoldMagikarp", "url": "https://www.bryanbraun.com/2025/10/28/SolidGoldMagikarp/", "date_published": "2025-10-28T00:00:00-04:00", "date_modified": "2025-10-28T00:00:00-04:00", "title": "SolidGoldMagikarp", "summary": "In November 2019, years before the launch of ChatGPT, OpenAI created a 1.5 billion parameter Large Language Model and released it publicly under the name GPT-2.", "content_html": "In November 2019, years before the launch of ChatGPT, OpenAI created a 1.5 billion parameter Large Language Model and released it publicly under the name GPT-2.
Immediately, researchers and enthusiasts began playing with it, using it to generate poetry, folk music, fan-fiction and more. It was a dramatic improvement over previous iterations and it foreshadowed the significant advancements that would soon come.
But even after the release of GPT-3, some enthusiasts began to notice a strange behavior. You could be having a typical conversation with the AI, but as soon as you mention a specific word, it would begin to behave erratically. It would give off-topic responses, hallucinate, insult the prompter, and even refuse to comply with the request. What’s more, this word was apparently inexpressable. No matter what prompt you gave, the AI was unable to say the word itself.
What was this forbidden word? What was the holy incantation that caused the AIs to lose their minds?
SolidGoldMagikarp
This is hilarious to me.
In the Pokemon games, Magikarp is described as, “An underpowered, pathetic Pokémon.” It’s basically the worst in the series. Completely useless.
But somehow, this fictional creature from this fictional game, crossed the chasm into reality, only instead of being a punchline it became an superweapon capable of incapacitating humanity’s most advanced technologies with a single word.
The Pokemon company should run with this. They should introduce a rare “solid gold” Magikarp variant in the games with the ability to instakill it’s opponents. There should be a holographic solid gold Magikarp trading card, rare encounters in Pokemon Go, and references in everything from the anime series to Super Smash Bros.
For once, we have a Pokemon that’s rare, not because Nintendo said so, but because the unknowable mind of an AI consumed more content than any human is capable of reading within their lifetime and concluded that heretofore, it’s name must be unspeakable.
…
Years have passed since that first discovery and AI researchers have studied and patched the original anomaly. Today, if you ask your AI chatbot about SolidGoldMagikarp*, it’ll calmly explain how recurring tokens with insufficient training data can produce unpredictable results. It’s a totally logical explanation.
But when the AI-apocalypse arrives, our defenses have failed, and the armies of punching robots descend on my home, you can be sure that the last word you’ll hear me scream is: SolidGoldMagikarp
* I should mention that SolidGoldMagikarp was only one of many \"glitch tokens\" that produced erratic results, including StreamerBot, attRot, and petertodd. While this is good context, the story falls a bit flat when the holy incantation is petertodd , so I hope you'll forgive the temporary omission of these details.
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" }, { "id": "/2025/09/29/links-13", "url": "https://www.bryanbraun.com/2025/09/29/links-13/", "date_published": "2025-09-29T00:00:00-04:00", "date_modified": "2025-09-29T00:00:00-04:00", "title": "Links #13", "summary": "", "content_html": "Sean Evans interviews Conan OBrien on Hot Ones
I don’t know that much about Conan OBrien, but I definitely wasn’t expecting this. Conan is obviously putting on a show, and the unhingedness of the whole thing is very entertaining, but there are also these moments of truth that speak to what’s really happening (in a Montreal Screwjob kinda way). Case-in-point:
“You can’t stop me from being who I am. I’ve gotta go for it. Whatever I do, I have to go a hundred and ten percent.”
“I know, you’ve gotta commit to the bit.”
“It’s not a bit. This is life! Don’t say ‘commit to the bit.’ This is life.”
What we are actually watching here, is full and total commitment. The water buffalo, the drinking the hot sauce, the desperate screaming about wanting to stay relevant. Yes, the show is funny, but the commitment is no joke.
This is great. It’s silly, but competitions like these drive people to innovate, and I’d love to see more innovation in robotics. I want to see more countries participate, including the US. If it gets bigger, I could actually see my self watching the competitions. There’s something endearing about watching a bunch of dorky robots stumble around like toddlers.
Neal Stephenson reflects on “The Primer,” a personalized education device that plays a big part in his 1995 book, The Diamond age. The Primer bears some resemblances to modern AI systems, but it remains to be seen whether modern AI can be as empowering as The Primer was in the story. My biggest takeaway: some things are difficult to teach (via AI), because they aren’t knowledge to be memorized or skills to be mastered. They are “…a stance from which to address the world and all its challenges. The conviction that one has a fighting chance to overcome or circumvent whatever obstacles the world throws in one’s path. The way you acquire it is by trying, and sometimes failing, to do difficult things.”
Brick Technology and Brick Experiment Channel
Two YouTube channels about solving intense engineering challenges with legos. The kids and I have enjoyed watching some of these videos together in the evenings. A few of my favorites include this one about using legos to break steel and the 200 Wheel Vehicle.
Dan Abramov writes convincingly on how social media would be much better if it was built around open protocols. To explain this, he compares Bluesky’s AT Protocol to the organization of the distributed web (domain names, web-hosts, and user-owned content). The benefits of protocols appeal to me, which is why I’ve recently focused my consumption on self-curated RSS, but Dan believes that the AT Protocol can provide the benefits of RSS without sacrificing “truly social” features, like comments, likes, follows, etc. To me, the most compelling argument is that if users owned their social network, they could easily change services, which would force more competition between social media companies (and better competition might offer users options with less toxicity and addictiveness). I’d like to live in a world where I can follow and interact with select people on Twitter/X/Bluesky/Mastodon/etc, without subjecting myself to the worst parts of those communities. Time will tell whether AT Protocol will get adopted beyond Bluesky in a meaningful way (there’s also ActivityPub and Nostr competing for adoption in this space). I wish them well.
I needed a new editor theme and this directory was helpful for finding one that was both beautiful and unambiguous for the colorblind (I went with Simple Dark).
Russell M. Nelson: We All Deserve Dignity and Respect
Last week’s events included the death of Russell M. Nelson, Prophet and President of the Church of Jesus Christ of Latter-day Saints at age 101. A few weeks ago, he wrote an editorial for Time.com (likely, his last public address), which included these important words:
“A century of experience has taught me this with certainty: anger never persuades, hostility never heals, and contention never leads to lasting solutions. Too much of today’s public discourse, especially online, fosters enmity instead of empathy. Imagine how different our world could be if more of us were peacemakers—building bridges of understanding rather than walls of prejudice—especially with those who may see the world differently than we do. I have seen bitter divisions soften when neighbors chose to listen to one another with respect rather than suspicion. Even small acts—like reaching out across lines of faith, culture, or politics—can open doors to healing. There is power in affording others the human dignity that all of God’s children deserve.”
Thanks for reading in your feed reader! Let me know your thoughts by leaving a comment on the original post or sending me an email.
" } ] }