Module 10: The Persian AI Community

← Previous | 🏠 Home | Next →

Overview

As artificial intelligence rapidly evolves globally, the Persian-speaking community has also seen a massive surge in AI adoption, development, and education. This module explores the vibrant ecosystem of Iranian AI startups, academic institutions, Telegram channels, and content creators. Whether you are a Persian speaker looking to connect with peers, or an international observer interested in the regional AI landscape, understanding this community is invaluable.

The growth of AI in Iran is particularly notable given the unique challenges of the region, including international sanctions, limited access to global payment gateways, and local internet restrictions. Despite these hurdles, Iranian developers, researchers, and tech enthusiasts have built a robust and self-sustaining ecosystem. They actively contribute to open-source projects, publish in top-tier international AI conferences (like NeurIPS, CVPR, and ACL), and create localized solutions tailored to the Persian language and domestic market needs.

This module will guide you through the various hubs of activity within the Persian AI community, providing you with a curated list of resources, communities, and thought leaders to follow.

---

Telegram Channels and Groups

Telegram serves as the primary digital town square for the Iranian tech community. It is where news breaks, discussions happen, and tutorials are shared. Due to the lack of access to platforms like Twitter/X for many ordinary users without VPNs, Telegram remains the most accessible and active platform.

General AI News & Trends

1. AI Forward (25k+ members): The leading channel for daily AI news. It translates announcements from OpenAI, Google, and Anthropic into Persian within minutes. It also shares prompts and use cases.
2. AI Tools Daily (20k+ members): Focuses strictly on new tools. Every day, they review 3-5 new AI websites, categorized by use case (e.g., video editing, SEO, coding).
3. Persian AI Mag (15k+ members): Acts like a digital magazine, offering long-form translated articles from sources like The Verge, Wired, and TechCrunch regarding AI policy and ethics.
4. Future of Tech Farsi (12k+ members): Explores the intersection of AI with other emerging tech like Web3, AR/VR, and quantum computing.
5. AI for Everyone (18k+ members): Aimed at non-technical users. Explains how teachers, doctors, and marketers can use AI.

Technical & Academic Channels

1. Deep Learning Iran (DLI) (15k+ members): The most serious technical group. Discussions revolve around PyTorch, TensorFlow, CUDA optimization, and paper implementations.
2. NLP Persian (8k+ members): The hub for computational linguists. Frequent discussions on fine-tuning Llama or Mistral for the Persian language, tokenization challenges, and dataset curation.
3. Data Science Iran (30k+ members): A massive community covering traditional Machine Learning (scikit-learn, XGBoost), Data Engineering, and MLOps. Excellent for job hunting.
4. Vision AI Tech (5k+ members): Specialized in Computer Vision. Topics include YOLOv8, medical image segmentation, and OpenCV projects.
5. AI Research Network (6k+ members): Focused on academia. Members help each other understand complex math in AI papers, share LaTeX templates for thesis writing, and discuss PhD opportunities abroad.

Generative AI & Prompt Engineering

1. AI Prompt Crafters (12k+ members): A massive repository of tested prompts for ChatGPT and Claude.
2. Midjourney Persian (10k+ members): Dedicated to AI art. Users share their prompt parameters, seeds, and post-processing techniques in Photoshop.
3. Audio & Video AI (7k+ members): Focuses on tools like Suno, Udio, Runway, and Pika Labs.

Tips for navigating Telegram AI communities:

• Use the Search Function: Before asking a basic question in a group of 10,000 people, search the chat history. Chances are, someone has already answered it.
• Beware of Scams: The hype around AI attracts scammers. Be extremely cautious of channels selling “AI Crypto Trading Bots” or “Guaranteed AI Income Courses.”
• Join the Associated Groups: Most channels have a linked discussion group. The real value is often in the group chats rather than the broadcast channel.

---

Discord Communities

While Telegram is for the masses, Discord is the preferred platform for gamers, developers, and collaborative projects.

1. Iran Tech Hub: A massive general tech server with dedicated, highly active #artificial-intelligence and #machine-learning channels. Great for voice chats and screen-sharing debugging sessions.
2. Persian AI Creators: Specifically for generative AI artists. They host weekly “prompt battles” and collaborative art projects.
3. PyFarsi: The official Iranian Python community server. Since Python is the lingua franca of AI, their data-science and ML channels are top-tier.
4. CS50x Iran Alumni: The community built around the localized Harvard CS50 course. They have specific study groups for the AI track.

---

Persian AI Newsletters and Blogs

For longer, more thoughtful content, several platforms host high-quality Persian articles.

Virgool (ویرگول)

Virgool is the Iranian equivalent of Medium. By searching for tags like هوش_مصنوعی (Artificial Intelligence), یادگیری_ماشین (Machine Learning), and پردازش_زبان_طبیعی (NLP), you can find hundreds of comprehensive tutorials.

• Notable Virgool Publications: Look for publications run by coding bootcamps like Quera or Maktabkhooneh, as they frequently post high-quality AI guides.

Persian Medium Writers

Many Iranian researchers studying abroad or working at FAANG companies maintain Medium blogs where they write in both English and Persian. They often provide “behind the scenes” looks at cutting-edge research.

Personal and Corporate Blogs

• Part AI Blog: Offers deep dives into the challenges of building AI for the Persian language.
• Snapp! Tech Blog: The technical blog of Iran’s leading ride-hailing app. They frequently publish articles on how they use Machine Learning for pricing algorithms, route optimization, and ETA prediction.
• Bazaar Tech Blog: Cafe Bazaar (the leading Android marketplace in Iran) publishes detailed posts on their recommendation systems and search algorithms.

---

Content Creators & YouTubers

YouTube is an indispensable resource. The Persian YouTube community has exploded with high-quality educational content.

1. Jadi Mirmirani (جادی): The most famous tech educator in Iran. While his focus is Linux and Python, his videos on neural networks and ML basics are legendary for their clarity and humor.
2. AI with Ali: Focuses on practical day-to-day applications of ChatGPT. He excels at breaking down complex prompt engineering techniques for the average user.
3. Deep Tech Persian: Run by a PhD student, this channel dives into the math and architecture of models like Transformers and Diffusion models. Not for beginners.
4. Sina ML: Provides hands-on coding tutorials. You watch him build models from scratch using PyTorch.
5. Prompt Ninja: The best Persian channel for Midjourney and Stable Diffusion. Covers advanced topics like ControlNet and LoRA training.
6. Data Science with Sara: Focuses on the data aspect of AI. Great tutorials on Pandas, SQL for AI, and data visualization.
7. TechTube Farsi: Covers the hardware side of AI—which GPUs to buy, how to set up a local LLM, and reviews of AI laptops.
8. Code Farsi: Excellent beginner-friendly Python tutorials specifically geared towards transitioning into AI.
9. AI Business Hub: Discusses the economics of AI. How to start an AI agency, automate workflows for clients, and monetization strategies.
10. NLP Farsi: Deep dives into linguistic challenges in AI. Essential viewing for anyone wanting to train models on Persian text.

---

Persian Podcasts

Podcasts are perfect for staying updated during a commute.

1. Radio Geek (رادیو گیک): Hosted by Jadi. It’s a general tech podcast but dedicates significant time to major AI breakthroughs and the societal implications of AI.
2. AI Cast (پادکست هوش مصنوعی): Interviews with Iranian researchers, startup founders, and engineers. It provides a realistic view of the domestic AI industry.
3. Data Talks Farsi: Focuses on data science careers, interviewing data scientists working both in Iran and at international tech giants.
4. Tech Talk Farsi: Explores the philosophical and ethical sides of AI. Will AI take jobs? What does AGI mean for humanity?
5. Silicon Valley Farsi: While broader in scope, it frequently covers the AI startup scene in California from a Persian perspective.

---

Iranian AI Startups & Companies

The domestic market in Iran relies heavily on local startups due to sanctions and localization needs.

Enterprise AI & APIs

• Part AI (پارت): The behemoth of Iranian AI. They offer a comprehensive suite of APIs including OCR (Optical Character Recognition) highly optimized for Persian script, Speech-to-Text, and localized NLP services.
• Roshan (روشن): A major player specializing in deep learning for text and voice. They are famous for their highly accurate Persian TTS (Text-to-Speech) engines used by visually impaired users and audiobook platforms.

AI in Tech Giants

• Snapp! / Tapsi: These ride-hailing platforms rely on massive data science teams to manage dynamic pricing, predict traffic patterns, and optimize dispatch algorithms.
• Digikala: The “Amazon of Iran” uses advanced AI for product recommendations, review sentiment analysis, and supply chain forecasting.

Open Source & NLP Initiatives

• Hazm (هضم): The cornerstone of Persian NLP. It is an open-source Python library for text preprocessing, stemming, and POS tagging in Persian.
• ParsBERT: A monumental achievement in Persian AI. ParsBERT is a monolingual language model based on Google’s BERT architecture, pre-trained on massive Persian corpora. It is the foundation for many local AI applications.
• Hooshvare: A research group dedicated to advancing Persian NLP, responsible for creating several important datasets and models.

---

Academic Resources

Iran has a strong tradition in STEM education, and its universities are producing world-class AI researchers.

Top Universities

• Sharif University of Technology: Often referred to as the MIT of Iran. Its AI lab is highly competitive and produces research published globally.
• University of Tehran: Boasts strong programs in computational linguistics and robotics.
• Amirkabir University of Technology: Known for its applied AI research and data mining labs.

Online Learning Platforms

• Maktabkhooneh (مکتب‌خونه): The Coursera of Iran. It hosts high-quality recorded lectures from top professors at Sharif and Tehran universities. You can find complete, semester-long courses on Machine Learning, Deep Learning, and AI.
• Quera: An interactive coding platform that offers specialized bootcamps in Data Science and Machine Learning.
• FaraDars (فرادرس): Offers thousands of specialized, low-cost video tutorials on specific AI frameworks and mathematical concepts.

---

Common Pitfalls & Advice

1. The “Get Rich Quick” Illusion: The Persian internet, like the global internet, is flooded with influencers claiming you can make $10k/month easily with AI. Treat these claims with extreme skepticism. Building real value with AI requires hard work and learning.
2. Relying Solely on Translated Content: While Persian resources are excellent for beginners, the cutting-edge of AI operates in English. To truly master the field, you must be comfortable reading English documentation, papers, and GitHub repositories.
3. Ignoring the Fundamentals: Don’t just jump into prompting LLMs. Take the time to understand basic statistics, Python, and how neural networks generally function. This foundational knowledge separates power users from novices.

---

Hands-on Workshop Exercise

Objective: Map out your personal Persian AI learning network.

1. Discovery: Open Telegram and join at least 3 of the channels listed in the “Telegram Channels” section. Spend 10 minutes scrolling through their recent posts.
2. Deep Dive: Go to YouTube and watch one video from “Jadi” or “AI with Ali” regarding a topic you are curious about (e.g., “What is ChatGPT?”).
3. Action: Find one open-source Persian AI project on GitHub (like Hazm or ParsBERT). Read its README file.
4. Share: In the workshop discussion group, share the most interesting tool, article, or video you discovered during this exercise and explain why it caught your attention.

← Previous | 🏠 Home | Next →