Comments URL: https://news.ycombinator.com/item?id=49929970

Points: 49

# Comments: 61

Comments URL: https://news.ycombinator.com/item?id=49927212

Points: 105

# Comments: 26

Comments URL: https://news.ycombinator.com/item?id=49927100

Points: 196

# Comments: 151

Comments URL: https://news.ycombinator.com/item?id=49925036

Points: 112

# Comments: 48

Comments URL: https://news.ycombinator.com/item?id=49915126

Points: 265

# Comments: 110

Comments URL: https://news.ycombinator.com/item?id=49893144

Points: 72

# Comments: 55

Comments URL: https://news.ycombinator.com/item?id=49890226

Points: 423

# Comments: 139

Comments URL: https://news.ycombinator.com/item?id=49886195

Points: 130

# Comments: 48

Comments URL: https://news.ycombinator.com/item?id=49882781

Points: 282

# Comments: 113

This is a great wee device! BLE, Wi-Fi and LoRa on one board makes it flexible for way more than a shopping list, this is just what I built first. This firmware only uses Wi-Fi.

Fully vibe-coded with Claude Code, I didn't hand-write this. I wanted to see how Claude would get on with building something useful for a new hardware device.

As it's stuck to the fridge it's easy for the family to use, and it's already being used day to day which I'll admit is a first for a home built project like this

Comments URL: https://news.ycombinator.com/item?id=49875801

Points: 149

# Comments: 62

Comments URL: https://news.ycombinator.com/item?id=49870295

Points: 356

# Comments: 117

Comments URL: https://news.ycombinator.com/item?id=49869574

Points: 316

# Comments: 137

I was frustrated with writing apps, and like most I used Apple Notes. After 2+ years immersed in Cursor, Codex and CC, I wished some interactions existed in a text editor.

The Markdown renders in place as you write it, including tables (which you can drag and resize without lining up pipes), KaTeX math and Mermaid diagrams. A lot of thought went into typography and the writing experience.

Your notes on the Mac are plain .md files in a folder with images next to them, so they open in anything else too. Each note keeps 200 versions. If you share a note with a link it syncs peer to peer over WebRTC while someone has it open, and the server that connects you only ever sees a random room id.

Most of the ~1,000 commits went into the editor itself. It's a contenteditable, and I don't let the browser restructure the document at all, which was a lot more work than I expected. Building this to be performant and cross-platform required some novel approaches, which I'll later open source. The editor is free with no account or ads, and there's an optional subscription for sync through your own iCloud.

https://www.markdown.beauty

Comments URL: https://news.ycombinator.com/item?id=49866597

Points: 69

# Comments: 52

Comments URL: https://news.ycombinator.com/item?id=49863600

Points: 128

# Comments: 69

Comments URL: https://news.ycombinator.com/item?id=49859682

Points: 27

# Comments: 17

Comments URL: https://news.ycombinator.com/item?id=49857536

Points: 27

# Comments: 2

Comments URL: https://news.ycombinator.com/item?id=49836979

Points: 27

# Comments: 3

Comments URL: https://news.ycombinator.com/item?id=49828120

Points: 67

# Comments: 32

Comments URL: https://news.ycombinator.com/item?id=49822970

Points: 52

# Comments: 43

Comments URL: https://news.ycombinator.com/item?id=49809935

Points: 53

# Comments: 17

With Sitefire (YC W26), we help marketing teams get recommended by AI Search (ChatGPT, Google AI Overviews, AI Mode, Claude, etc.). Our software monitors prompts, sees which web pages get cited, and uses these insights to help marketing teams take action, e.g. create YouTube videos or write the right blog posts.

This means we have a commercial stake in AI-generated web content. And for now, high-information, AI-generated content works great to get cited and recommended in AI Search.

But after talking to hundreds of marketing teams, it became clear that everyone despises AI-generated content (“AI slop”). And yet, everyone still wants to leverage AI to create content. So we asked ourselves: what characterizes AI slop? Can we train a model to identify it from human-generated web pages?

Researchers from the University of Maryland and Google DeepMind already asked this question for fiction. Their paper StoryScope (Russell et al., 2026) showed that you can tell AI-written stories from human ones by their structure alone, without looking at the words.

We ported their pipeline to commercial web pages. Using the Wayback Machine, we collected 2,250 blog posts from 268 B2B company websites that were written before ChatGPT existed. For each blog post, five AI models (GPT-5.4, Claude Sonnet 4.6, Gemini 3 Flash, DeepSeek V3.2, Kimi K2.5) wrote their own version.

Instead of looking at the words, we looked at how each post is built. We had an AI model answer 214 questions about every post, e.g. how hard it pushes its own product, whether it backs up its claims with sources, or whether it quotes a named expert. Then we trained a classifier on these answers.

On blog posts it had never seen before, our classifier told AI-generated and human posts apart with 98% accuracy, getting only 19 of 1,740 wrong.

Why does it work so well? Because all five AI models write in a similar shape. Mapping every AI model’s values for these features, we see they cluster together, while the human values sit apart and spread out much more. Of the 1% most unique blog posts in our data set, 149 are human, only 4 are AI.

So what characterizes AI slop? It tells you the same thing three times. The title already promises what you'll get ("How to Cut Onboarding Time in Half"), the intro lays out what's coming, and the ending says it all again. 77% of the AI posts end by repeating their main point, compared to only 12% of the human posts. We call it the tidy, self-announcing blog post.

Still, each AI model has its own accent. We trained a second classifier to tell which of the five AI models wrote a post, or whether a human did. It picks the right author 79% of the time, where random guessing (1 in 6) would get 17%. Almost all of its mistakes are mix-ups between the AI models, not between human and AI.

The cool thing about structural features is that you can't simply reword your way out of it. We had each AI model rewrite its own posts until, on average, 73% of their original 13-word sequences were gone, and the AI slop classifier still worked just as well.

We're building this into Sitefire: our agents get a structural understanding of text, so the posts they write go deeper and vary the way human writing does.

There's a lot we haven't tested yet, like the myriad of humanizer tools, human rewriting, restructuring a post, or prompting an AI model to explicitly avoid these habits. And our human posts are mostly from 2020 to 2022, while the AI posts were generated in August 2026. Structure can't really tell when a human post was written, but it's still not a same-year comparison.

We published the study with all the figures on arXiv: https://arxiv.org/abs/2609.15369. The code is on GitHub: https://github.com/pulse-energy-eu/slopshape

We're pretty sure your own blog isn't AI slop, is it? We built a checker that runs one of your posts through the ten features from the paper, so you can see for yourself (the full report asks for a work email): https://sitefire.ai/slop-checker.

Think you can tell AI slop from human writing? We also made a little game to see if you can keep up with our model, which gets all five rounds right: https://sitefire.ai/spot-the-slop.

Comments URL: https://news.ycombinator.com/item?id=49800566

Points: 74

# Comments: 26

Comments URL: https://news.ycombinator.com/item?id=49791125

Points: 163

# Comments: 49

Comments URL: https://news.ycombinator.com/item?id=49783568

Points: 57

# Comments: 26

Comments URL: https://news.ycombinator.com/item?id=49775700

Points: 70

# Comments: 118

Comments URL: https://news.ycombinator.com/item?id=49773624

Points: 35

# Comments: 2

Comments URL: https://news.ycombinator.com/item?id=49769676

Points: 221

# Comments: 162

---

Hello. It's 2026, we're training simulated fruit fly brains to play Beat Saber, do we still have to be stuck with internet (re)search as fn: natural language -> black box we can't do anything about -> ranked_list/summary?

There is a long history of people trying to do very fancy things that end up being done in relational databases and a little SQL. There is a gravity to them, a bitter lesson, just like scaling of generalized ml training methods. I mean many, many information products can be built off essentially giant real-time OLAP databases and frontier LLMs writing brilliant SQL+Datalog+vector+Jev etc. queries.

Google Search, Tavily, Exa essentially have the problem of mapping your agents' context you are willing to provide, to a tiny subset of their index. You pay a fixed cost to an extremely hard problem that has a distribution of hardness, which means YOU eat the downsides when they are running out of budgeted compute to help you out.

Their algorithms are opaque to the caller, there's really not much user control, and there's not a serious opportunity to communally improve search recipes, like the lexical+Jev recipes you trust to select bleeding edge AI builders.

Furthermore, search companies aren't even pursuing text-to-SQL anymore (several have talked to me)... they made up their minds during the traumatic 2024 text-to-sql days. They were just too early.

I hope you enjoy. I'm intent on scaling this paradigm on differentiated hardware over much more data, so any compelling use cases or queries I could show off, would be much appreciated!

Comments URL: https://news.ycombinator.com/item?id=49748041

Points: 61

# Comments: 26

Comments URL: https://news.ycombinator.com/item?id=49746963

Points: 27

# Comments: 7

Comments URL: https://news.ycombinator.com/item?id=49737326

Points: 50

# Comments: 9

Comments URL: https://news.ycombinator.com/item?id=49735979

Points: 93

# Comments: 14

Comments URL: https://news.ycombinator.com/item?id=49735049

Points: 31

# Comments: 2

You start in the Halo 3 main menu, go through the lobby, spawn in the map, and explore. There are photos, stories, quests, mini-games, and Easter eggs scattered around. Movement is via WASD and mouse, no signup.

Guardian was my third place growing up. Starting in 9th grade, my friends and I would get home after school, get on Halo, and hang out in a Guardian free-for-all with MLG settings until everyone came online, then play customs or ranked for hours.

On weekends we'd bring our Xboxes and monitors to each other's houses, stay up all night, and play in local tournaments. Guardian was probably the closest thing I had to a digital home, so I wanted my site to live there. Playing it and hearing the theme music for the first time in 10+ years in a place I knew by heart felt a bit like coming home.

The map itself isn't AI-generated. It uses an original Guardian asset export distributed through the fan-run Halo Archive: an OBJ with 115,250 faces, the original UV coordinates, and 42 referenced diffuse textures, loaded through Three.js.

What was AI-generated was everything on top needed to turn it into a complete game: shaders, lighting, scenario tags, physics, animations, collision, movement, sky and fog, lifts, pickups, weapon handling, and quest routing.

Every pickup was placed by hand against the geometry using callout maps and a dev-only coordinate locator. This was surprisingly tedious to get right - I think it took me 10 iterations just to get the blue lift on Camo to line up parallel with the ledge.

The Guardian geometry and texture assets are 21 MB, so the load is hidden inside the lobby countdown and scorecard screen while the shaders warm up. It targets 60 FPS and feels smooth on a recent MacBook in Chrome.

Touch controls work on mobile, but the field of view is roughly 20 degrees in portrait, so desktop is definitely the best way to play it.

I built it with Codex using Astra. It wrote essentially all of the code: 20,552 lines of TypeScript across 118 files, 973M tokens, 37 hours of runtime, and about 25-30 hours of my time.

Coding wasn't the hard part. The time-consuming part was deciding which memories belonged in the map, where each should live, what game mechanic fit each one best, and then playing it over and over to fix bugs and get the movement and user flow to feel right.

I'll write up more about the Codex/Astra side in a comment since I had a lot of interesting learnings from building something this large with it.

This is an unofficial fan project and I'm not making money from it. The geometry, textures, UI, and many of the sound effects are Halo material and belong to their respective owners.

Some things are still rough. There's occasional lag when picking up a power weapon. The lifts don't feel quite natural - it was hard to replicate the vacuum effect of the real lifts. On some browsers like Firefox, audio needs a click before it starts.

Would love feedback on how it could be improved, especially from anyone else who spent a lot of time on this map.

Comments URL: https://news.ycombinator.com/item?id=49732259

Points: 29

# Comments: 9

Comments URL: https://news.ycombinator.com/item?id=49726501

Points: 34

# Comments: 23

Comments URL: https://news.ycombinator.com/item?id=49721548

Points: 125

# Comments: 41

Comments URL: https://news.ycombinator.com/item?id=49715113

Points: 234

# Comments: 171

I always had the problem that building HTML pages is really simple now, but trying to save data required hosting it somewhere, and sharing it afterwards was not easy. Over the last few months, I've been building an app called Capsule (it’s also the file extension name) written in Rust with Tauri 2.0 that allows packing an HTML app and its data into a single SQLite file.

The HTML file and any related assets are directly embedded in the database. User data can either be saved as a localStorage key/value store or via a MongoDB-inspired collections API as documents, saved in a table in the file. You can also save other assets, like PDF files or images, directly in the database to keep different documents together. All data can be easily exported to CSV or JSON if needed.

Privacy and security were a big priority for me, so documents cannot do anything out of the box. They don’t have direct access to the file system and they require permission to access the internet. The permission model is still something I’m working to improve. Capsule documents can also use local or remote AI models for document specific AI features.

One downside with this approach is that multiple people working on it will create different copies. To make it possible to merge different copies of the same file, each data entry has a unique UUID and timestamp.

I’m planning to open up the file format specification for the 1.0 version of the app so other apps can read or write Capsule files.

You can try it out in the web preview at https://withcapsule.app/preview with pre-built templates or use any AI provider of your choice to create a custom, Capsule-optimized app by using the following prompt:

"Please read the app wizard instructions at https://withcapsule.app/prompt.txt and help me design an app.“

I’m still working on the file format but there are migrations for each new version, so data should never be lost when using newer versions of the app in the future. Please let me know if you have any ideas or use cases where this might make sense or does not work.

Comments URL: https://news.ycombinator.com/item?id=49712278

Points: 380

# Comments: 167

Comments URL: https://news.ycombinator.com/item?id=49706708

Points: 63

# Comments: 100

Comments URL: https://news.ycombinator.com/item?id=49705868

Points: 106

# Comments: 90

Comments URL: https://news.ycombinator.com/item?id=49700138

Points: 27

# Comments: 3

Comments URL: https://news.ycombinator.com/item?id=49695805

Points: 66

# Comments: 24

Comments URL: https://news.ycombinator.com/item?id=49693226

Points: 42

# Comments: 21

Comments URL: https://news.ycombinator.com/item?id=49688391

Points: 109

# Comments: 15

Comments URL: https://news.ycombinator.com/item?id=49681724

Points: 28

# Comments: 51

Comments URL: https://news.ycombinator.com/item?id=49679685

Points: 48

# Comments: 34

Comments URL: https://news.ycombinator.com/item?id=49664574

Points: 26

# Comments: 6

Comments URL: https://news.ycombinator.com/item?id=49653311

Points: 110

# Comments: 23

- SQLite is now the default database.

- Datastar is used for server-side rendering.

- Introduces biff.graph[2] and biff.fx, two libraries that both help structure your application logic.

- Biff is now "just" a collection of independent libraries rather than a single monolithic framework, which makes it much easier to swap out the various parts. e.g. using a non-default database is much less work now.

I'd been experimenting with a good chunk of this work (the database change and the biff.graph / biff.fx stuff) starting 2 years ago. The Datastar and independent libraries stuff goes back 6 months or so. Datastar's been on my radar for a while and has been picking up steam in the Clojure community, for good reason IMO. The independent library stuff was instigated by me trying out LLM-driven coding for the first time earlier this year: I wanted to try making a Clojure web app without using Biff to see if web frameworks still seemed helpful (my conclusion: absolutely) and how I might do things from first principles were I to start over. So I did that and then started packaging up/rewriting various Biff code as smaller individual libraries as I went.

Note that I consider Biff to be mostly "hand-coded"; I start with LLM-written drafts of code but often by the time I'm done editing it's been completely reworked. The initial drafts of docs are all hand-written. I'm interested to try using it for vibe-coding apps though. e.g. how much of the code that requires good judgment can I pull into the framework so that the apps are then easier for LLMs to not screw up.

[1] https://news.ycombinator.com/item?id=23921220

[2] https://news.ycombinator.com/item?id=48820361

Comments URL: https://news.ycombinator.com/item?id=49647635

Points: 29

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=49642845

Points: 30

# Comments: 4

Comments URL: https://news.ycombinator.com/item?id=49639104

Points: 185

# Comments: 181

Comments URL: https://news.ycombinator.com/item?id=49636345

Points: 184

# Comments: 8