Poor Man’s AI — Building While the Servers Are Down

With my servers and bots currently down while I’m in the middle of moving, I’ve still been working hard behind the scenes.

Right now, most of my focus has been going into what I’ve been calling Poor Man’s AI.

👉 https://peakebot.onrender.com/

This project is exactly what it sounds like: an attempt to build an AI-style knowledge system using whatever free tools I can get working together.

No expensive cloud setup.
No paid AI infrastructure.
No massive budget.

Just code, persistence, and a lot of trial and error.

The Goal

The goal is to build a system that can keep learning by saving information into its own knowledge banks.

It is not a full artificial intelligence yet. I would not call it that.

Right now, it is more like a self-growing knowledge system. The more information it receives, the more useful it should become over time.

The hope is to make something that can:

Accept input from users
Save useful knowledge
Recall stored information later
Keep building its own knowledge base
Run using mostly free services
Stay lightweight enough for someone like me to actually maintain

Python, Node.js, FTP, and Static Hosting

This has been one hell of a process.

One of the biggest challenges has been trying to make Python work smoothly with Node.js.

The Node.js side handles the web server and user interaction, while Python is being used for the thinking, processing, and knowledge-handling side of things.

Then comes the storage problem.

Since I’m trying to keep this as free as possible, I’ve been working on ways to save information through an FTP service connected to a static website that allows a large amount of storage.

That sounds simple until you actually try to wire it all together.

Every piece has to talk to the next piece correctly:

The user sends information
The Node.js server receives it
Python processes it
The system saves it
The stored knowledge has to be readable later
The whole thing has to stay organized enough not to collapse into a mess

That part alone has been a serious learning curve.

Looking Into IPFS

Something I would really like to figure out next is IPFS.

I want a free or very low-cost way to store the knowledge so it does not ever truly get lost.

I have not read deeply enough into IPFS yet to say I fully understand it, but the idea is interesting.

If I can store important knowledge files in a decentralized way, then the project would not be fully dependent on one server, one website, or one hosting company.

That would fit the whole point of the project:

A poor man’s AI should not need rich man’s infrastructure.

Feeding the Knowledge Banks

Right now, what the system needs most is consistent input.

The more useful information it receives, the better its knowledge banks become.

I am also trying to force-feed it knowledge myself at a steady rate. The goal is to build up enough stored information that the system can start giving better and more useful responses.

It is still early.

It still needs work.

It still needs better memory handling, better structure, and probably a cleaner way to sort through what it has learned.

But it is moving.

Building Through the Chaos

Between moving, keeping things together for my children, and dealing with my servers being down, I have not been able to run everything the way I want.

But I am still building.

That matters.

Poor Man’s AI is not being built in perfect conditions. It is being built in the middle of real life, with limited resources, using whatever free tools I can stitch together.

That is part of what makes it meaningful to me.

This project is about proving that even without money, even without perfect hardware, and even without a professional AI lab, someone can still build something useful.

One step at a time.

One file at a time.

One piece of knowledge at a time.

Hopefully, over the next few weeks, I can get it running smoother and continue expanding what it can remember, process, and return.

Poor Man’s AI is still rough.

But it is alive.

🤖 PeakeBot — Autonomous Trading System (RC-AWARE)

Independent multi-token trading bot featuring:
RC-aware execution, adaptive delay logic, and self-regulating trade cycles.

📊 Trading bot details:
👉 https://geocities.ws/p/e/peakecoin/trading-bot/peakebot_v0_01.html
💻 Open-source repositories:
👉 https://github.com/paulmoon410