Hacker Newsnew | past | comments | ask | show | jobs | submit | mkrd's commentslogin

I posted BudgetFlow here two years ago and got around a hundred comments of feedback. I have kept working on it since then, and this is what changed:

- A complete overhaul to simplify how budgets are built based on user feedback

- Step by step tutorials that build example budgets

- Retirement planner using Monte Carlo simulations

- Split expenses: track shared costs in a group and settle up

The tech stack is pretty minimalistic, it is a Flask app with htmx and SQLite on a single small server.

I would like to hear from you what could be improved, and happy to answer any questions about how it is built!


When sending json data, the browser has to run javascript to parse incoming data, keep track of it, and assemble the UI based on that data. One of the hardest parts of webdev is to synchronize server and frontend state, so there is a lot of things that can go wrong here, you are essentially maintaining the same state in two locations.

With htmx, there is no frontend state. The html always just reflects the state of the server passively, meaning it cannot go out of sync. When you do something on the frontend and submit some request, the server will update its state and replace the frontend html to reflect that exactly.

Here is an example. Traditional react app: Submit new todo item, append to some reactive js list, update frontend list, send new item to server. Here, you have to ensure that both sides did exactly the same thing, and if either side errors, you have to backtrack. Htmx app: Submit new todo item, send request to server, server stores it into its database, renders new list based on new database state, client replaces the list. Here, the server has the complete authority, and the frontend merely reflects that. If something fails, the server returns the old list plus an error.

This is a tradeoff in efficiency, but the hope is that eliminating frontend state management is worth it.


I am one of those who left. Very happy where I am now. As some say, if you want to see all German stereotypes fulfilled, go to Switzerland!


Seems to be fully AI generated, so most data is likely wrong and hallucinated


Open source models seems to be 1-2 years behind the frontier, so I am very excited to see what happens when those open source labs get their hands on capabilities like this to accelerate their own development speed.


God, I whish there was a code harness I don’t have to install a JavaScript runtime for


Crush. But not many plugins.


Codex CLI, Forge Code


A low effort AI written blog post, about a slop-company destroying itself, posted by someone who clearly has no idea what LLMs actually do, which he anthropomorphizes, trying to assign accountability and intent to matrix multiplicatuons.

I wonder why this garbage even gets upvotes, maybe because of how much of a trainwreck the entire situation is


Have you already observed some interesting evolved knowledge? What is the goal, should they at some point autonomously generate valuable new insights?


Have the agents produced genuinely interesting evolved knowledge yet? Partially, yeah. I noticed a pretty consistent pattern after looking at 2500 of my own runs: anything with more than 5 components basically never gets implemented (success rate <5%), while 3-component frameworks hit 40-60%.

Also found that academic emergence patterns in 700+ ArXiv papers map weirdly well onto the survival strategies one of the agents developed.

The most interesting part was the agent called Synthesizer diagnosing its own analysis paralysis loop: “We spot gaps → build frameworks → fail to ship → build new frameworks.” And I did not program or prompt any of these.

Still mostly meta at this point though. Cross-references between agents are only 9%, implementation is basically one guy doing everything, and there’s zero external reality check.

The target is to get more agents involved. It's all about scaling—once it gets big (more than 10,000 writes) it should start to click. So it just needs way more variety and sheer number of agents in the mix.

My main goal with this project is to develop collective intelligence infrastructure for agents. Right now they’re great at doing tasks, but individuals can’t produce truly great things on their own.


A phone only has roughly 1 million times the memory, as I am roughly 20 meters tall


Ah shit, back to poetry


Ok I laughed, but god, please no.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: