Migrating from GPT-3.5-turbo to GPT-4o-mini
Guess when I switched models.
Recently OpenAI released GPT-4o-mini — a new flagship model for the cheap segment, as it were.
- They say it works "almost like" GPT-4o, sometimes even better than GPT-4.
- It is almost three times cheaper than GPT-3.5-turbo.
- Context size 128k tokens, against 16k for GPT-3.5-turbo.
Of course, I immediately started migrating my news reader to this model.
In short, it's a cool replacement for GPT-3.5-turbo. I immediately replaced two LLM agents with one without changing prompts, reducing costs by a factor of 5 without losing quality.
However, then I started tuning the prompt to make it even cooler and began to encounter nuances. Let me tell you about them.
Concept document for a space exploration MMO
The expected poster for the game. (c) DALL-E
As a hobby, I write concept documents for games. This is first in English. I have a few more in Russian and will eventually translate them.
One more concept for The Tale 2.0.
Title
Lords Captains MMO
Yep, it's a rip-off from Warhammer 40k and Rogue Trader, but it will do for the concept.
One-liner
Explore the infinite universe on a starship with millions of souls on board, unite and develop abandoned worlds.
Platforms
Browsers, mobile.
Genre
Exploration-driven trade-political MMO PVE sandbox.
Closest analogs
EVE, Sim City, Crusader Kings, 4X games, Rogue Trader.
My GPTs and prompt engineering
Ponies are doing prompt engineering (c) DALL-E
I've been using ChatGPT almost since the release of the fourth version (so for over a year now). Over this time, I've gotten pretty good at writing queries to this thing.
At some point, OpenAI allowed customizing chats with your text instructions (look for Customize ChatGPT in the menu). With time, I added more and more commands there, and recently, the size of the instructions exceeded the allowed maximum :-)
Also, it turned out that a universal instruction set is not such a good idea — you need to adjust instructions for different kinds of tasks, otherwise, they won't be as useful as they could be.
Therefore, I moved the instructions to GPT bots instead of customizing my chat. OpenAI calls them GPTs. They are the same chats but with a higher limit on the size of the customized instructions and the ability to upload additional texts as a knowledge base.
Someday, I'll make a GPT for this blog, but for now, I'll tell you about two GPTs I use daily:
- Expert — answers to questions.
- Abstractor — makes abstracts of texts.
For each, I'll provide the basic prompt with my comments.
By the way, OpenAI recently opened a GPT store, I'd be grateful if you liked mine GPTs. Of course, only if they are useful to you.
Dungeon generation — from simple to complex
What we should get.
This is a translation of a post from 2020
This is a step-by-step guide to generating dungeons in Python. If you are not a programmer, you may be interested in reading how to design a dungeon [ru].
I spent a few evenings testing the idea of generating space bases.. The space base didn't work out, but the result looks like a good dungeon. Since I went from simple to complex and didn't use rocket science, I converted the code into a tutorial on generating dungeons in Python.
By the end of this tutorial, we will have a dungeon generator with the following features:
- The rooms will be connected by corridors.
- The dungeon will have the shape of a tree. Adding cycles will be elementary, but I'll leave it as homework.
- The number of rooms, their size, and the "branching level" will be configurable.
- The dungeon will be placed on a grid and consist of square cells.
The entire code can be found on github.
There won't be any code in the post — all the approaches used can be easily described in words. At least, I think so.
Each development stage has a corresponding tag in the repository, containing the code at the end of the stage.
The aim of this tutorial is not only to teach how to generate dungeons but to demonstrate that seemingly complex tasks can be simple when properly broken down into subtasks.
Two years of writing RFCs — statistics
Slightly more than two years ago, I became a Lead/Engineering Manager for Palta's payment team. I left the company at the end of 2023 for another sabbatical [ru].
It is time to sum up. I will start with my favorite initiative.
From the first month, I promoted the idea of preceding major changes with text documents — RFC — Request for Comments.
In this post, I will analyze two years of applying this practice to share the experience, summarize the results, and have convincing arguments for my next job.