yoesefab

SWE📍🇲🇦

[MASA @MarwaneAsa]

new week. new youtube video.

https://t.co/lqB9yf2FiD

[Marouane Lamharzi Alaoui @marouane53]

The "editors are dead" posts flooding my timeline are nonsense. Pure hype. None of those tools are changing real editing.

Most of what's being hyped isn't editing intelligence. It's a transcript, silence detection, captions, and a chatbot calling ffmpeg.
That's useful but calling it the future of editing is like calling autocomplete the future of writing.

The hard part isn't exporting final.mp4. The hard part is making video understandable at scale.

A real AI editor has to understand an entire library, not one clip. Scenes, people, dialogue, mood, story beats, continuity, missing coverage, what the editor already tried. It has to find exact moments across thousands of scenes without hallucinating. It has to live inside the real workflow, not a toy prompt box pretending to be a studio.

That's why we built Firassa around the Universal Video Knowledge Graph.
UVKG turns raw footage into persistent, timecoded memory. Not a transcript. Not tags. Not a search index pretending to understand video. A living graph of what exists, how moments relate, where the story is hiding.

Firassa turns that memory into action. Organize the library. Ask the footage questions. Find exact shots. Build first-pass sequences. Generate only when the story needs coverage that doesn't exist.

The future of editing isn't an AI that spits out a file. It's an AI that understands the footage well enough to become part of the editorial process.

We're not replacing editors. We are making them faster, sharper, and freer to do the part only humans can do: tell the story.
Pro editors from Netflix and Beast Games are already using it and love it.

93% on the Long Video MME benchmark. Link to the waitlist below.

[Kaito | 海斗 @_kaitodev]

what a privilege to wake up and try again.

[yoesefab @yoesefab]

@XaMiNeZH 💀

[Claude @claudeai]

Introducing Claude Design by Anthropic Labs: make prototypes, slides, and one-pagers by talking to Claude.

Powered by Claude Opus 4.7, our most capable vision model. Available in research preview on the Pro, Max, Team, and Enterprise plans, rolling out throughout the day.

[Mohammed Daoudi @MIduoad]

You can contribute easily to https://t.co/JUl7JpCPaX by using https://t.co/MWkcvZDuDQ

Access code: MrrakcIsLive!

[Mouad @mouad_builds]

ولات واحد الموجة ديال الـ prompts الموجودين…
هاك Copy هاد البرومبت الى،
. بغيتي دير مشروع؟ هاك prompt
. بغيتي تقرر فحياتك؟ هاك prompt
. بغيتي تصاوب Super App؟ هاك prompt

ولكن شنو كاين من مور هاد الـ prompts؟

[MASA @MarwaneAsa]

@BenGeekly just dropped ProdShort on @ProductHunt, its really cool for busy founders who want to make content but dont have time!

check it out on PH : ProdShort ( its actually #2 today, it should go to #1 🙃)

[yoesefab @yoesefab]

@MIduoad M9bara 🤔 ?

[Marouane Lamharzi Alaoui @marouane53]

In 2019 Sam Altman said GPT-2 was "too dangerous to release". A 1.5 billion parameter text predictor that could barely write a coherent paragraph. That was the big scare. A glorified autocomplete.

Today Anthropic announced Claude Mythos and it's genuinely what Altman was fantasizing GPT-2 would be except he didn't build it.

This thing found thousands of zero-day vulnerabilities across every major operating system and every major web browser. It found a 27-year-old bug in OpenBSD, an OS literally famous for being unhackable, that survived decades of human security audits. It found flaws in the Linux kernel and autonomously chained them together into a full remote takeover exploit on the first try. 83.1% success rate on proof-of-concept exploits with no human guidance.

Anthropic is so spooked by their own model they won't release it publicly. They created something called Project Glasswing and gave access only to Amazon, Apple, Microsoft, Google, CrowdStrike, Nvidia and a few dozen others to defensively scan their own code before this capability spreads to other labs.

THIS is "too dangerous to release". Not a chatbot that writes fake blog posts. A model that can pop every server on the internet.

But don't sleep on OpenAI. GPT-5.4 Pro is already absurdly good. First AI to beat human experts on desktop computer use. 1M token context. Frontier coding baked in. And Axios already reported OpenAI is building a model with similar cyber capabilities for their own trusted access program. People inside the industry say Mythos-level capabilities are 6 to 18 months from showing up across every major lab.

OpenAI has the secret sauce for raw intelligence. They ship relentlessly. GPT-5, 5.2, 5.3-Codex, 5.4, each one a real jump. Their next frontier model will almost certainly match or exceed Mythos on cyber. It's just a matter of when.

[Chaofan Shou @Fried_rice]

Claude code source code has been leaked via a map file in their npm registry!

Code: https://t.co/jBiMoOzt8G

[𝒵𝒶𝒸𝓀 @1zakaria1]

4 سوايع ديال الكاوية والبرمجة بـ ESP32 باش نخرج بهاد النتيجة: نظام إضاءة ARGB باش يحاكي ضوء الشمس (الشروق، النهار، والغروب) بشكل الي . 💡☀️
​النباتات , Moss, Fittonia وزئبق السلام
الكائنات Springtails, Roly Poly و TinySnails. 🌿🐌
اشنو بان ليكم 🤗

[Marouane Lamharzi Alaoui @marouane53]

Calling what we do now vibe coding is like calling flying a plane vibe flying.
Nobody's vibing at 40,000 feet. You have a destination, you make decisions, the autopilot handles the rest. Same thing. The models got serious. The workflows got serious. The results got serious but the name didn't.
It should be called intent coding.
You think clearly, you express what you want, the model executes.
The skill shifted from typing to thinking.

[yoesefab @yoesefab]

@irachdaoui L3id qreb anmshi nstagi m3a chi 7ela9

[Mohammed Aboullaite @laytoun]

هاد الخريطة كتبرز بالواضح أن أكثر المهن المعرّضة للخطر مع الـAI هي الوظائف اللي خدمتها رقمية وتحليلية أكثر من المهن الميدانية: Software Developers, Data Analysts, Editors, Paralegals, وWriters… اكثر مهن مضرة
كلما كانت الخدمة كلها على الحاسوب، كلما كان احتمال التأثير أكبر 😳

[Mouad @mouad_builds]

AI B Darija 🇲🇦 Ep#3

Machine Learning algorithms sound complex ☹️…

But the core ideas are actually simple ;)

In this episode,

I explain the fundamentals of:

• Linear & Logistic Regression
• Decision Trees & KNN
• K-Means & DBSCAN
Etc...

فهاد الفيديو،

شرحت أساسيات أهم خوارزميات تعلم الآلة بطريقة مبسطة
وبأمثلة من حياتنا الواقعية باش نكملو سلسلة الفهم ديالنا للذكاء الإصطناعي.

ما تنساش تشارك feedback ديالك فالتعليقات 👇

#AI #MachineLearning #Ai_Bdarija

[Mouad @mouad_builds]

AI B Darija 🇲🇦 Ep#2

Today I explain the basics of Machine Learning in a simple way: how computers learn from data and start making predictions ;)

فهاد الفيديو شرحت تعلم الآلة ML بطريقة مبسطة
كيفاش الكمبيوتر كيتعلم من data، كيقلب على patterns، ومع الوقت كيولي قادر يعطي prediction على معلومات جديدة.

و خطوة بخطوة غادي نبنيو الفهم ديالنا لــعالم الذكاء الإصطناعي😀

[Mo @atmoio]

I was a 10x engineer. Now I'm useless.

[Mouad @mouad_builds]

AI B Darija🇲🇦 Ep#1

In this first video, I explain what #ArtificialIntelegence is and how it works in a very simple way.
The goal is to make AI easier to understand for everyone.
If you're curious about AI or just getting started, this series is for you.
More videos coming soon!

قد نستهين و نغفل عن تعلم بعض الأمور من البداية لأننا أحياناً نضطر أن نبدأ من الوسط ثم نعود لفهم الأبجديات و الأساسيات تدريجياً!

فأول فيديو من (الذكاء الإصطناعي بالدارجة) حاولت أن أعرّف بمفهوم الذكاء الاصطناعي بشكل مبسّط وكيف يعمل :)

[yoesefab @yoesefab]

Lesgooo 🔥

[Mouad @mouad_builds]

First video :)
First step,

Let’s see where this journey goes 💪