🔥Breaking: Alibaba’s New AI Video Model Crushes Sora

PLUS: ChatGPT Gets Major Upgrades & Microsoft Lifts Copilot Limits

Welcome, AI pioneers!

Alibaba’s open-source AI video suite is challenging Sora, Anthropic’s Claude is playing Pokémon (slowly), and OpenAI is rolling out major ChatGPT upgrades. Meanwhile, Microsoft just removed key limits on Copilot’s advanced reasoning. Here’s what’s making waves in AI today.

In today’s Generative AI Newsletter:

  • Alibaba’s Open-Source AI Video Is Shaking Up the Industry

  • Claude Plays PokĂ©mon—One Slow Move at a Time

  • OpenAI Rolls Out Two Major Updates for ChatGPT

  • Microsoft Removes Copilot Limits

Lead the AI Revolution—By Doing, Not Just Learning

AI is transforming industries, yet 84% of professionals feel behind. Why? Traditional courses focus on theory—not real impact.

At GenAI.Works, we take a hands-on approach with guidance from experts at Stanford, Google, and Amazon to help you:

âś… Build real-world AI solutions

âś… Tackle industry challenges through practical projects

âś… Gain recognition and lead the AI future

🎯 Your learning. Your pace. Your success.

 Alibaba’s Open-Source AI Video is Shaking Up the Industry

Image Source: Wan AI/Alibaba

Alibaba’s Tongyi Lab just released Wan2.1, an AI video model suite that outperforms both SOTA open-source and closed models like OpenAI’s Sora, while running 2.5x faster.

đź’ˇ Why this matters:

  • New industry leader – Wan2.1-T2V-14B tops the VBench leaderboard, excelling in complex motion dynamics, real-world physics, and text rendering.

  • Multilingual AI video – The first AI model to generate readable English & Chinese text directly in videos, eliminating a key limitation.

  • More accessible than ever – A light 1.3B version runs on consumer GPUs, generating a 5-second 480p clip on an RTX 4090 in just 4 minutes.

  • Advanced video editing – Features inpainting, outpainting, multi-image referencing, and structural consistency tools for seamless AI-driven content creation.

Alibaba’s open-source push is reshaping the AI video race—with China rapidly iterating, will closed models like Sora keep up?

Claude Plays Pokémon—One Slow Move at a Time

Image Credits: Claude Plays Pokémon on Twitch

Anthropic just launched Claude Plays Pokémon on Twitch, where its Claude 3.7 Sonnet AI model is attempting to beat Pokémon in real time. Unlike basic game-playing AIs, Claude “thinks” through its moves, analyzing obstacles, planning ahead, and learning from mistakes—though not without hilarious missteps.

đź’ˇ Why this matters:

  • Unlike basic reinforcement learning models, Claude analyzes its environment, recalls past actions, and strategizes future moves instead of relying on preprogrammed patterns.

  • Claude stores key information like character locations and objectives, allowing it to recover from mistakes and adjust its approach dynamically.

  • The AI uses structured reasoning techniques similar to agentic AI systems, interpreting game mechanics as real-world problems.

  • Claude processes pixel-based game visuals rather than reading raw game data, making it function closer to a human player than traditional AI-driven bots.

Claude’s Pokémon journey is equal parts entertaining and frustrating—but it’s also a glimpse into a world where AI plays games, reasons through problems, and maybe even becomes a gaming personality itself.

OpenAI Rolls Out Two Major Updates For ChatGPT

Image Source: OpenAI

OpenAI has rolled out Deep Research access to all ChatGPT Plus, Team, Edu, and Enterprise users, while also launching a free Advanced Voice Mode powered by GPT-4o mini for all users. These updates mark a major expansion of ChatGPT’s capabilities, bringing enhanced research and conversational AI to a wider audience.

đź’ˇ Key updates:

  • Deep Research Expansion – Paying ChatGPT users now get 10 queries/month, while $200 Pro-tier users see an increase to 120 queries/month. This tool provides in-depth, structured insights for researchers, finance professionals, and policy analysts.

  • Advanced Voice Mode (Free) – OpenAI’s real-time, natural-sounding voice AI is now available to free-tier users, offering near-instant, conversational responses at a fraction of the cost of full GPT-4o voice.

  • AI Research Arms Race – OpenAI, Google, and Perplexity are competing to dominate AI-driven research, as companies push users toward premium AI subscriptions.

By making deep research and AI-powered voice interactions more accessible, OpenAI is widening the gap with competitors while reinforcing its presence in AI-driven productivity tools.

Microsoft Removes Copilot Limits—Here’s Why

Image Source:

Microsoft has lifted restrictions on Copilot’s Voice and Think Deeper features, making advanced AI reasoning and real-time conversations free for all users. This move comes as demand surges and users hit frustrating usage caps.

đź’ˇ What’s Changing & Why It Matters:

  • Voice Mode is now unlimited – Users can engage in longer, interactive conversations, improving language learning, brainstorming, and hands-free assistance.

  • Think Deeper removes limits – Powered by OpenAI’s o1 model, this feature helps analyze complex decisions like big purchases, home investments, and career planning.

  • Scaling AI for the masses – Microsoft is ramping up accessibility in response to Copilot’s growing adoption, aiming to compete with ChatGPT and Google Gemini.

  • Copilot Pro still gets perks – Pro users retain priority access to new models, experimental AI tools, and Microsoft 365 integrations.

With AI assistants becoming an essential tool, Microsoft is pushing to make Copilot a go-to AI companion—and it’s now free to use at full capacity.

🚀 Boost your business with us—advertise where 10M+ AI leaders engage

🌟 Sign up for the first AI Hub in the world.

📲 Our Socials

Reply

or to participate.