Skip to content

Instantly share code, notes, and snippets.

View moshebeeri's full-sized avatar

moshe beeri moshebeeri

View GitHub Profile
@moshebeeri
moshebeeri / article-the-picture-the-model-will-not-give-up.md
Created October 2, 2026 08:19
The picture the model will not give up: how AI image models carry inherited prejudice into pictures of the past

The picture the model will not give up

2 October 2026 · Moshe Beeri

AI image models have inherited a picture of the Neanderthal that science withdrew decades ago, and they hold on to it when asked to draw something else.

A cultured Neanderthal with a brute's face

A recent ynet article reported a study of a cave on the Turkish coast. Neanderthals lived there from about 77,000 years ago and collected small, inedible sea shells, apparently because they liked them. When modern humans replaced them about 59,000 years ago, the newcomers made the same tools and collected the same shells. The researchers suggest our ancestors learned from the Neanderthals, taste included.

@moshebeeri
moshebeeri / Strategic-Analysis-2026-US–Israel–Iran Conflict.md
Last active April 11, 2026 15:57
A Strategic Analysis of the 2026 US–Israel–Iran Conflict

The IRGC's Empty Gun: Why Iran Lost the War It Started — and What the US Should Do Next

A Strategic Analysis of the 2026 US–Israel–Iran Conflict

By Moshe Beeri - GenAI Specialist


The War That Was Misread From the Start

Agentic Provider-Consumer Pattern

Agent-to-Agent Communication using A2A-MCP protocols

Introduction

My journey into building scalable agentic systems began with a simple challenge: I needed a framework that could handle complex workflows with multiple AI agents working together seamlessly. Initially, I explored the Motia framework, which provided an excellent foundation for orchestrating event-driven workflows with zero infrastructure setup. The code-first approach and built-in observability were exactly what I needed for rapid development.

However, as my agent ecosystem grew more complex, I discovered the need for more dynamic context sharing between agents. This led me to the Model Context Protocol (MCP), which offered a standardized way to provide resources and prompt templates to AI models. I realized I could create a catalog of MCP resources that would automate the creation of specialized agents, each with their own unique capabilities but speaking a common language.

The final piece of the

from openai import OpenAI
import os
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
def get_completion(messages, model="gpt-4", temperature=0.7):
response = client.chat.completions.create(model=model,
messages=messages,
temperature=temperature)