Hacker News: dinarino

New comment by dinarino in "Doxa – Economical-constrainted ABM simulator with Ollama"

dinarino — Mon, 27 Apr 2026 07:38:37 +0000

We conceived Doxa as an emerging simulator with Ollama (therefore open source and free) that attempts to emulate even complex realities in economics, game theory, and sociology, imposing constraints, operations, and limits on each agent within which to operate.

Currently, with models like Qwen3.5:4B, we obtain fine strategies in contexts of geopolitical conflicts and social tensions. The project is open source.

Doxa – Economical-constrainted ABM simulator with Ollama

dinarino — Mon, 27 Apr 2026 07:35:19 +0000

Article URL: https://github.com/VincenzoManto/Doxa

Comments URL: https://news.ycombinator.com/item?id=47918750

Points: 6

# Comments: 1

New comment by dinarino in "Show HN: Implit – Catch fake AI-generated dependencies"

dinarino — Sun, 26 Apr 2026 09:37:05 +0000

Is AI hallucination on packages/deps really widespread?

New comment by dinarino in "Show HN: CSP Benchmarks – Go vs. core.async (Clojure) vs. libgoc (C)"

dinarino — Fri, 24 Apr 2026 16:13:55 +0000

I did something similar on CSP, it was pure hell.

Show HN: Doxa – Open-source emergent simulator for geopolitical scenarios

dinarino — Fri, 24 Apr 2026 13:52:52 +0000

Hi! We, Vincenzo and Riccardo, built Doxa as an agnostic engine for emergent simulations with agents for constrainted scenarios (like geopolitical, economics, ...) and work well with LLMs like Qwen2.5:7B, Llama but also cloud models such as Gemini.

Doxa can be configured in minutes with YAML scenarios, where you can assign agents' persona, portfolios, constraints.

Here a ready-to-use Colab: https://colab.research.google.com/github/VincenzoManto/Doxa/...

Let us know your thoughts!

Comments URL: https://news.ycombinator.com/item?id=47890337

Points: 2

# Comments: 0

New comment by dinarino in "Simulation of the Hormuz crisis after Iran seizure of 2 ships with 6 LLM agents"

dinarino — Wed, 22 Apr 2026 15:03:21 +0000

I simulated the Strait of Hormuz crisis where agents (US, Iran, Israel) manage quanlitative resources like "political capital" and "strait control" via local LLMs. The simulation showed emergent US military intervention on ships: the operation works but it couldn't balance the narrative spin against the economic blockade. Built with a custom YAML framework to test how trade utility and military operations impact automated diplomatic decision-making.

Simulation of the Hormuz crisis after Iran seizure of 2 ships with 6 LLM agents

dinarino — Wed, 22 Apr 2026 15:03:21 +0000

Article URL: https://colab.research.google.com/github/VincenzoManto/Doxa/blob/main/notebooks/doxa.ipynb

Comments URL: https://news.ycombinator.com/item?id=47864648

Points: 5

# Comments: 7

New comment by dinarino in "Show HN: Doxa – open-source platform for multiagent simulations using easy YAML"

dinarino — Tue, 21 Apr 2026 11:03:15 +0000

The idea behind the platform is to model complex economic-geopolitical scenarios with LLMs and simple YAML like this:

global_rules: epochs: 1 steps: 12 execution_mode: sequential maintenance: corn: 1 markets: - resource: gold currency: credits initial_price: 6.0 min_price: 1.0 max_price: 40.0 clearing: per_step - resource: corn currency: credits initial_price: 2.4 min_price: 0.5 max_price: 15.0 clearing: per_step actors: - id: player provider: google model_name: gemini-2.5-pro persona: Farmer-trader. Your core business is converting gold into corn. Keep enough corn to survive maintenance and monetize surplus. trading_mode: lob initial_portfolio: credits: 45 corn: 12 gold: 5 panic: 0.0

Show HN: Doxa – open-source platform for multiagent simulations using easy YAML