Find your customers on Roku this Black Friday

As with any digital ad campaign, the important thing is to reach streaming audiences who will convert. To that end, Roku’s self-service Ads Manager stands ready with powerful segmentation and targeting options. After all, you know your customers, and we know our streaming audience.

Worried it’s too late to spin up new Black Friday creative? With Roku Ads Manager, you can easily import and augment existing creative assets from your social channels. We also have AI-assisted upscaling, so every ad is primed for CTV.

Once you’ve done this, then you can easily set up A/B tests to flight different creative variants and Black Friday offers. If you’re a Shopify brand, you can even run shoppable ads directly on-screen so viewers can purchase with just a click of their Roku remote.

Bonus: we’re gifting you $5K in ad credits when you spend your first $5K on Roku Ads Manager. Just sign up and use code GET5K. Terms apply.

Use code GET5K now

Today:

• OpenAI Launches GPT-5.1-Codex-Max for Smarter, Safer Coding at Scale

• Musk’s xAI to Build 500MW AI Data Center in Saudi Arabia

• Google DeepMind Opens New AI Research Lab in Singapore

• xAI Launches Grok 4.1 Fast with New Agent Tools API

• Microsoft Unveils Agent 365 to Manage AI Agents Across Enterprises

OpenAI Launches GPT-5.1-Codex-Max for Smarter, Safer Coding at Scale

OpenAI launched GPT‑5.1-Codex-Max, a powerful coding model designed for long tasks, deep debugging, and multi-step agent work. It’s faster, smarter, and uses fewer tokens than before. It improves collaboration, supports large projects, and enhances cybersecurity tools while keeping human oversight essential.

KEY POINTS

• Long-Horizon Coding & Compaction: The model can handle large, ongoing tasks by smartly trimming past context while preserving progress—enabling uninterrupted work over hours or even days.

• Token Efficiency & Cost Savings: It achieves better results with fewer tokens, especially in frontend design and complex software tasks, leading to lower costs for developers.

• Cybersecurity & Safety Measures: It includes better safeguards against misuse and is OpenAI’s most capable cybersecurity model to date—though human oversight is still strongly advised.

Why it matters

This model brings AI closer to being a dependable coding assistant for big software projects. It can fix bugs, redesign apps, and stay focused for long periods without getting lost. By saving time and reducing mistakes, it helps developers build better tools faster—while staying safe and under control.

Musk’s xAI to Build 500MW AI Data Center in Saudi Arabia

Elon Musk’s AI company xAI is building a 500-megawatt data center in Saudi Arabia, powered by Nvidia chips and backed by the kingdom’s Humain venture. The project was revealed at a U.S.-Saudi investment forum featuring Musk, Nvidia’s CEO, and top political leaders.

KEY POINTS

• Massive Data Center in Saudi Arabia: xAI will build a 500-megawatt AI facility, one of the largest of its kind, to power next-generation AI models.

• Partnership with Nvidia and Saudi Arabia: The project is backed by Saudi state-owned AI venture Humain and will use Nvidia hardware for compute.

• Announced at High-Level Forum: The plan was unveiled in Washington during a joint event with Elon Musk, Jensen Huang, and leaders including Trump and Crown Prince Mohammed bin Salman.

Why it matters

This shows how serious Elon Musk is about making xAI a top player in artificial intelligence. Building in Saudi Arabia gives him money and space to run big AI systems. Teaming with Nvidia means he’ll have the best hardware. This could shape how global AI power is built and controlled.

Google DeepMind Opens New AI Research Lab in Singapore

Google DeepMind is launching a new AI research lab in Singapore to boost AI development and adoption across Asia-Pacific. The lab will focus on inclusive, culturally aware AI, especially advancing Gemini models, and applying them to Google products and cloud services in the region.

KEY POINTS

• New DeepMind Lab in Singapore: Focused on cutting-edge AI research, this lab expands Google’s presence and investment in Asia-Pacific.

• Advancing Gemini & Inclusivity: The lab aims to improve Gemini and other frontier models, emphasizing local languages and cultural relevance.

• Regional Impact & Growth: DeepMind’s APAC team has doubled in size, showing strong regional commitment and collaboration with local partners like AI Singapore and Singapore’s EDB.

Why it matters

AI needs to work well for everyone—not just people in Western countries. This lab helps make sure AI tools understand and respect Asian languages, cultures, and needs. It also means Asia gets more direct benefits from the latest AI tech and research, closer to home.

🧠RESEARCH

Souper-Model: How Simple Arithmetic Unlocks State-of-the-Art LLM Performance

Meta researchers introduced Soup of Category Experts (SoCE), a new method that boosts AI model performance by smartly combining parts of different models instead of retraining them. By averaging the best models for different tasks using weighted techniques, they achieved state-of-the-art results in language, math, and tool use.

Part-X-MLLM: Part-aware 3D Multimodal Large Language Model

Part-X-MLLM is a new AI model from Tencent that understands and edits 3D objects using natural language. It takes a 3D point cloud and prompt, then outputs structured commands to describe, generate, or modify object parts. This unified system enables precise 3D editing and achieves top results across multiple tasks.

AraLingBench A Human-Annotated Benchmark for Evaluating Arabic Linguistic Capabilities of Large Language Models

AraLingBench is a new benchmark designed to test how well AI models truly understand Arabic. It includes 150 expert-written questions covering grammar, syntax, and reading. Results show many models perform well on surface tasks but struggle with deeper language skills, exposing gaps in true comprehension.

🛠️TOP TOOLS

Each listing includes a hands-on tutorial so you can get started right away, whether you’re a beginner or a pro.

AI Tutor Pro – Your Personal Digital Assistant - browser‑based study assistant

AI Voice Checker – AI Voice Detection Tool - web and browser‑extension tool that analyzes an audio clip

AI-coustics — Advanced Audio Enhancement Platform - cleans up voice in recordings and live streams

📲SOCIAL MEDIA

🗞️MORE NEWS

• xAI launched Grok 4.1 Fast and its Agent Tools API, enabling developers to build powerful AI agents with real-time web search, code execution, and file tools. The model excels in tool use, speed, and reasoning.

• Microsoft Agent 365 is a centralized dashboard to manage AI agents across organizations. It offers tracking, security, access control, and analytics—helping companies safely scale agents without rebuilding their current systems or workflows.

• Larry Summers has resigned from OpenAI’s board following new revelations about his connections to Jeffrey Epstein. The move marks the latest high-profile fallout from the ongoing scrutiny of Epstein’s extensive network.

• Meta’s chief AI scientist Yann LeCun is leaving to start a new company focused on “world model” AI. His exit follows internal AI division turmoil and reflects diverging visions on open-source and AGI development.

• Luma AI raised $900 million from Saudi-backed Humain to expand its video-generating AI models. They’ll jointly build a 2-gigawatt supercluster in Saudi Arabia and develop Arabic-focused AI tools for regional businesses and cultural representation.

• OpenAI and Target are partnering to bring AI shopping tools into ChatGPT. Customers can browse, plan, and check out directly in chat. Target also uses ChatGPT Enterprise internally to boost productivity and streamline operations.

• MIT researchers built an AI agent that uses CAD software like a human. Trained on 41,000 videos, it converts 2D sketches into 3D models, aiming to simplify design, boost productivity, and support engineering education.

• Warner Music settled a copyright lawsuit with AI startup Udio and signed a licensing deal for a 2026 AI music platform. The service will let fans create remixes using licensed artist voices and songs.

• The U.S. approved selling 70,000 advanced AI chips to UAE and Saudi Arabia, reversing earlier concerns. The deal, tied to Trump’s diplomacy, includes safeguards to block China and boosts Nvidia and AMD partnerships.

• Microsoft and Nvidia are investing up to $15 billion in Anthropic, raising its valuation to $350 billion. The deal includes massive compute commitments, deep tech collaboration, and signals Microsoft diversifying beyond its OpenAI partnership.