Abed Khooli

Introducing Claude Haiku 4.5 Anthropic released Claude Haiku 4.5 today, the cheapest member of the Claude 4.5 family that started with Sonnet 4.5 a couple of weeks ago. It's priced at $1/million input tokens …

My notes (and a pelican) for the new Claude Haiku 4.5 model, released by Anthropic this morning simonwillison.net/2025/Oct/15/...

الذكاء الاصطناعي الصحي والأخلاقي لجيل البيانات | Abed Khooli الذكاء الاصطناعي الصحي والأخلاقي لجيل البيانات يُعرف الجيل المولود بين عامي 2010 و 2024 باسم "الجيل ألفا"، لكنني أُطلق عليه اسم جيل "D" (أي جيل البيانات أو الجيل الرقمي). هذا الجيل...

الذكاء الاصطناعي الصحي والأخلاقي لجيل البيانات
Healthy and Ethical AI for Generation D
www.linkedin.com/posts/akhool...

Abed Khooli on LinkedIn: ALLaM Language Model and Revisiting Arabic Tokenizers A few days ago… ALLaM Language Model and Revisiting Arabic Tokenizers A few days ago, NCAI/SDAIA published a 7B instruction version of ALLaM (https://lnkd.in/diNUwrt2). The…

ALLaM Language Model and Revisiting Arabic Tokenizers
www.linkedin.com/posts/akhool...
#ALLaM #tokenizer #NLP #AI #LLMs

Report: composite-indicators.jrc.ec.europa.eu/explorer/ind...

Global Knowledge Index 2024
مؤشر المعرفة العالمي 2024

Qwen2.5-Max from
Alibaba
and Deepseek V3 from DeepSeek AI
are two recent LLMs that are likely to disrupt #GenerativeAI
I asked both about "Fanar chat" (recent LLM from
QCRI). Qwen 'hallucinated' an answer and DS said nothing as of cut-off date (July 2024).

Abed Khooli on LinkedIn: #ai #misinformation #disinformation #factchecking #media #journalism… The Future of Media and Journalism in the Age of AI MENA (Jordan, Lebanon, Morocco, Palestine and Tunisia): desk research, questionnaire, and in-depth…

The Future of Media and Journalism in the Age of AI
MENA (Jordan, Lebanon, Morocco, Palestine and Tunisia): desk research, questionnaire, and in-depth interviews
مستقبل الإعلام والصحافة في عصر الذكاء الاصطناعي: الأردن، تونس، فلسطين، لبنان والمغرب
www.linkedin.com/posts/akhool...

Research paper on Fanar LLM
الورقة البحثية الخاصة بنماذج فنار
arxiv.org/pdf/2501.13944

Abed Khooli on LinkedIn: A Toy ModernBERT-based Arabic Sentence Transformer Following the earlier… A Toy ModernBERT-based Arabic Sentence Transformer Following the earlier release of a toy Arabic ModernBERT model (https://lnkd.in/ddGS3uPR), here's an…

A Toy #ModernBERT-based Arabic Sentence Transformer and #HateSpeech Classifier
www.linkedin.com/posts/akhool...

Abed Khooli on LinkedIn: The Need for A Standard Arabic Tokenizer While working on the… The Need for A Standard Arabic Tokenizer While working on the experimental Arabic ModernBERT model (https://lnkd.in/ddGS3uPR), and based on earlier Arabic…

The Need for A Standard Arabic Tokenizer
www.linkedin.com/posts/akhool...

Abed Khooli on LinkedIn: Here's an experimental tiny Arabic ModernBERT base model:… Here's an experimental tiny Arabic ModernBERT base model: https://lnkd.in/gHc64DVM trained on 300K items from Arabic FineWeb2 (filtered for language score…

The first Arabic tiny #ModernBERT model (with custom tokenizer):
www.linkedin.com/posts/akhool...

Don’t stop talking about Palestine

That didn't take long! Nomic AI has finetuned the new ModernBERT-base encoder model into a strong embedding model for search, classification, clustering and more!

Details in 🧵

Are there any tutorials or starter scripts on training a new #ModernBERT models (or fine-tuning the Answer.ai's model for a language from a different family)? In both cases, a new tokenizer will be trained.

The Oxford Insights Government AI Readiness Index 2024 is out.
Top 10: United States, Singapore, S. Korea, France, United Kingdom, Canada, Netherlands, Germany, Finland, Australia.
MENA has 5th rank as a region with 48.50 average score (world average is 47.59).
Report and data: lnkd.in/dp6ERDMi

Abed Khooli on LinkedIn: حوار مع فنار أطلق معهد قطر لبحوث الحوسبة، مع مجموعة من المؤسسات الأخرى،… حوار مع فنار أطلق معهد قطر لبحوث الحوسبة، مع مجموعة من المؤسسات الأخرى، نموذجا لغويا كبيرا يدعم اللغة العربية بشكل أساسي واسمه فنار ، والفنار ضوء في أعلى…

Chatting with Fanar - the new LLM from Qatar
www.linkedin.com/posts/akhool...

Attending lnkd.in/dW_DxekM ...
Protecting rights in the age of #AI

OpenAI’s new defense contract completes its military pivot A new partnership with Anduril, announced today, will deploy AI on the battlefield. It represents an overhaul of the company’s position in just a year.

OpenAI in 2024:

“No AI for weapons or military”

“Do use our AI to make weapons to hurt yourself or others”

“Military is fine, but no AI for weapons”

“Sure put it on battlefield drones”

www.technologyreview.com/2024/12/04/1...

🌍 AI, Power, & Human Rights Symposium | Dec 10–11
How does AI shape power, amplify oppression, and protect rights?
📅 3 sessions:
🤖 AI in Power Structures
💥 Weaponized AI
🌟 Building an AI Alliance
🔗 Register: [birzeit-edu.zoom.us/webinar/register/WN_qzM7jCiCQPqmsi0exN7NxQ]

#EthicalAI #HumanRights

Abed Khooli on LinkedIn: #opendata #datataxation #dataphilanthropy #ai Data Philanthropy and Data Taxing Data is not just a resource, it is a national asset and should be treated as such. It has been said that data is the new oil…

Data Philanthropy and Data Taxing
www.linkedin.com/posts/akhool...
#OpenData #DataTaxation #DataPhilanthropy #AI

bluesky-community/one-million-bluesky-posts · Datasets at Hugging Face We’re on a journey to advance and democratize artificial intelligence through open source and open science.

First dataset for the new @huggingface.bsky.social @bsky.app community organisation: one-million-bluesky-posts 🦋

📊 1M public posts from Bluesky's firehose API
🔍 Includes text, metadata, and language predictions
🔬 Perfect to experiment with using ML for Bluesky 🤗

huggingface.co/datasets/blu...

Two recent reports to look at:
Global Knowledge Index:
www.knowledge4all.com/dashboard
Global AI Index:
aiindex.stanford.edu/report/

Did anybody try R2L text tokenization? It could make a difference there for R2L languages. For math, R2L (ex. Arabic) uses the same order of digits.

@bsky.app How do I view post stats/impressions?

I think a 'critical AI mass' exodus from x (both producers and consumers) is needed before we can have a good diverse menu here. I looked for AI and data governance and could barely find a few.

حوكمة البيانات في الجنوب العالمي - لأجل الصالح العام والخاص
www.linkedin.com/posts/akhool...

Abed Khooli on LinkedIn: Arabic Hate Speech Detection is not an easy task. Available datasets are… Arabic Hate Speech Detection is not an easy task. Available datasets are either not representative or have annotation issues. Context is also an issue as well…

Arabic #HateSpeech Detection is not an Easy Task
Dataset + Model
linkedin.com/posts/akhool...
#NLP #AI
حول صعوبة كشف خطاب الكراهية باللغة العربية

Abed Khooli on LinkedIn: AI Risk Mitigation Tools Along with advances in AI, and generative AI in… AI Risk Mitigation Tools Along with advances in AI, and generative AI in particular, the need for safe and responsible AI is a global priority. There are…

AI Risk Mitigation Tools:
- Garak from NVIDIA (lnkd.in/dgj7U7_E)
- Guardrails from Guardrails AI (lnkd.in/d6_jNV7K)
- PyRIT from Microsoft (lnkd.in/ddruVb2k)
www.linkedin.com/posts/akhool...

Abed Khooli on LinkedIn: #datagovernance #opendata #datasharing #dataeconomy #datalaw Data Governance in the Global South - for Public and Private Good In today's digital age, data has become a critical asset that drives decision-making…

Data Governance in the Global South - for Public and Private Good
www.linkedin.com/posts/akhool...

Hello, world!
Welcome to BlueSky!
السلام عليكم
مرحبا في بلوسكاي