🎮The world’s first AI-generated, playable game: Oasis

Hey there!

I’m back, and this time I’ve been exploring something mind-blowing.

The world’s first AI-generated playable game, Oasis by Etched.

If you’ve ever dreamed of stepping into a game world that creates itself in real time, you’ve got to check this out.

But that’s not all!

In today’s edition, I’ll also be sharing some exciting new tools and updates you won’t want to miss.

So stay tuned and enjoy the read!

What’s Inside Today’s Newsletter

🪰 Buzz Around AI

😀 AI Recap

🪄 AI Creation

🔹 AI Learning

♨️ What’s Hot In AI

📰 AI News

Read time: 10 mins

🪰 Buzz Around AI

🎮 Oasis AI Creates Open-World Games in Real-Time

If you enjoy AI video games, check out Oasis by Etched – the first AI-made, playable game!

What’s Happening: Decart and Etched launched Oasis, an AI that builds game worlds in real time.

It also includes a Minecraft-style demo. Check Demo

What is more exciting is that you can upload any image you want.

After playing, you can even download a video of your adventure.

The Details:

• Oasis creates game environments, including physics, item interactions, and lighting, based on keyboard and mouse inputs.

• It runs at 20 frames per second (FPS) on current hardware—100 times faster than older AI models.

• The companies are releasing the code:

  That is a smaller 500M-parameter model for local testing and a demo of a bigger version.

• Future versions will support 4K resolution and run on Etched’s Sohu chip. They’ll handle 10 times more users and massive AI models with over 100 billion parameters.

Why It Matters: Oasis shows AI can create real-time, interactive game worlds. This could change how games and virtual worlds are made. Highly likely it will even replace traditional game engines.

📑 Claude Adds PDF Vision Capabilities

What’s Happening: Anthropic has launched PDF support for its Claude 3.5 Sonnet model in a public beta. This update lets Claude analyze both text and images, like charts, in large documents.

The Details:

• The system processes PDFs in three steps: extracting text, converting pages to images, and combining both for analysis.

• It supports documents up to 32MB and 100 pages, including everything from financial reports to legal papers.

• The feature can be used with other Claude tools like prompt caching and batch processing.

• It’s available through Anthropic’s Claude platform and via API for integration into apps.

Why It Matters: Claude could already analyze large documents. Now, with the ability to understand images, it has become an even stronger tool for fields like healthcare and finance, where visuals often contain key information.

😀AI Recap

• Chinese Military Uses Meta’s Llama Model for ChatBIT

  

  Image Source @Lexica.art

  
  Chinese military researchers have used Meta’s open-source Llama model to build ChatBIT, an AI tool for military intelligence and strategy.

• Microsoft’s ‘Copilot Vision’ Coming Soon

  

  
  Microsoft teased that its Copilot Vision feature will launch soon. It will allow the AI assistant to see and understand a user’s browser content and actions.

• Google Adds ‘Grounding with Google Search’ for Gemini

  

  
  Google has released Grounding with Google Search for its Gemini API. This helps developers integrate real-time search results into model responses, improving accuracy and reducing errors.

• Amazon Delays AI-Infused Alexa to 2025

  

  
  Amazon has delayed the release of its AI-powered Alexa until 2025. Testing has faced issues like hallucinations and performance problems with basic tasks.

🪄 AI Creation

The user asks ChatGPT to generate the weirdest thing it can come up with

🔹AI Learning

Create Stunning Presentations in Minutes!

Follow this simple step-by-step guide to craft captivating presentations effortlessly with Decktopus:

1. Sign Up: Visit Decktopus.com and create your account.

2. Choose a Template: Select from a variety of professionally designed templates tailored to your needs.

3. Customize Your Content: Edit each slide and use Decktopus's smart content prompts to help organize your ideas.

4. Add Visuals: Upload your own images or pick from Decktopus’s extensive library of visuals.

5. Fine-tune the Design: Personalize themes, colors, and layouts for a sleek, cohesive look.

6. Preview & Share: Review your presentation, then easily share it via link, PDF, or present it live.

With Decktopus, creating polished presentations is quick, easy, and enjoyable!

♨️ What’s Hot In AI

Runway Adds 3D Control to AI Video Generation

What’s Happening: Runway has launched Advanced Camera Control for its Gen-3 Alpha Turbo model, bringing more precision to AI-generated video. The new features mimic traditional filmmaking techniques.

The Details:

• Users can now control camera movements like panning, zooming, and tracking shots with adjustable intensity.

• The system keeps 3D consistency as users move through the generated scenes, maintaining depth and space.

• This update shows Runway’s progress in developing “world models”—AI that can simulate realistic physical environments.

• The release comes after Runway’s partnership with Lionsgate, hinting at future use in major film production.

Why It Matters: AI video quality has improved dramatically, but tools to control it haven’t kept up—until now. This update makes AI video generation more reliable, moving from random results to a tool creators can use with confidence.

📰 AI News

1. xAI has released the Grok API in public beta. Researchers get $25 in credits each month for multi-modal integration.

2. Standard Intelligence has released Hertz-dev, an open-source model for generating conversational audio.

3. ElevenLabs has launched the Voice Design API, which lets users create custom voices from text prompts.