Godmother of AI unveils World Labs

Welcome, AI enthusiasts.

Fei-Fei Li, referred to as the “Godmother of AI,” just launched a startup to create AI that understands and generates 3D worlds.

With a fresh $230M in funding, could this spatial intelligence push AI beyond language models and revolutionize AR/VR, robotics, and more? Let’s get into it…

In today’s AI rundown:

• World Labs goals to create 3D worlds with AI

• AI creates open-world video games from text

• Prompt suggestions and tricks for OpenAI’s latest o1

• OpenAI o1 is smarter than most humans

• 6 latest AI tools & 4 latest AI jobs

• More AI & tech news

Read time: 4 minutes

LATEST DEVELOPMENTS

WORLD LABS

🌎 World Labs goals to create 3D worlds with AI

The Rundown: Fei-Fei Li, renowned AI researcher referred to as the “Godmother of AI”, has launched World Labs, a spatial intelligence company developing AI models able to understanding and generating 3D environments.

The main points:

• World Labs is pioneering “Large World Models” (LWMs) to perceive, generate, and interact with 3D worlds, incorporating physics and semantics beyond 2D imagery.

• Co-founded by Li, Justin Johnson, Christoph Lassner, and Ben Mildenhall, the startup has secured over $230 million in funding from high-profile investors, including Andreessen Horowitz.

• Initial applications concentrate on creating and editing virtual 3D spaces for creative professionals and general users.

• The team goals to bridge the gap between AI’s current text-based understanding and the complex spatial relationships of the actual world.

Why it matters: Immediately, AI relies heavily on text descriptions provided by humans. But determining spatial intelligence and moving AI beyond LLMs and language could greatly impact and speed up the event of intelligent AR/VR, robotics, architecture, game design, and rather more.

TOGETHER WITH DEFINED.AI

🌟 Top-Tier AI data – Flexible, Scalable, Inexpensive

The Rundown: Defined.ai provides probably the most diverse and ethically-sourced marketplace for AI training data, allowing businesses to create smarter, more cost effective AI solutions at scale.

Defined.ai delivers:

• Customizable, bias-free datasets that adhere to strict privacy and ethical standards

• Industry-specific data for sectors including insurance, banking, and retail

• Fast access to pre-built datasets to speed up your AI projects

Claim your 5 free hours of knowledge today, exclusively for The Rundown readers!

TENCENT

🎮 AI creates open-world video games from text

The Rundown: Tencent just unveiled GameGen-O, a groundbreaking AI model able to generating open-world video game content from text prompts and allowing interactive gameplay simulation.

The main points:

• GameGen-O can create diverse game elements like characters, environments, actions, and events from easy text prompts.

• It also allows for interactive control of the generated video game content it creates to construct worlds that gamers can interact with.

• The AI model was trained on OGameData, a brand new dataset of over 4,000 hours of high-quality video game footage and uses a two-stage training process: foundation pretraining and instruction tuning.

• The training data segments were annotated by GPT-4o for validation purposes.

Why it matters: Following the impressive accomplishments of Google’s GameNGen, Tencent’s GameGen-O raises the stakes for AI-generated video games. This technology will dramatically reduce the time and value of constructing open-world games, resulting in higher quality games created by smaller teams of game developers.

AI TRAINING

🪄 Prompt suggestions and tricks for OpenAI’s latest o1

The Rundown: OpenAI’s latest release, o1, unlocks latest reasoning capabilities, but there’s a catch: prompts ought to be fundamentally different than the way in which you prompt GPT-3 and GPT-4, resulting from the brand new Chain-of-Thought (CoT) architecture.

Here’s our top suggestions:

1. Log into ChatGPT (paid subscription required) and choose o1-preview or o1-mini from the model dropdown.

2. Keep your prompts easy and direct, avoiding Chain-of-Thought instructions.

3. Use XML tags for clarity and test with difficult questions or philosophical queries.

4. Analyze its pondering process by clicking on the “pondering” indicator after each response to grasp the model’s reasoning path.

Pro tip: Experiment with different query types to totally grasp o1’s advanced reasoning abilities and take a look at OpenAI’s latest ‘Advice on prompting‘ guide.

PRESENTED BY ASSEMBLYAI

🎙️ AssemblyAI sets the Speech AI standard

The Rundown: AssemblyAI delivers probably the most accurate speech-to-text model with advanced audio intelligence – allowing teams to construct beyond transcription and transform words into actionable insights.

AssemblyAI offers:

• Top performance rankings across major industry benchmarks

• Fully-featured speech understanding models

• Comprehensive developer docs and reliable SDKs

Join today and get $50 in free credits.

AI RESEARCH

🧠 OpenAI o1 is smarter than most humans

The Rundown: OpenAI’s latest model, “o1”, has reportedly scored an IQ of around 120 on the Norway Mensa IQ test, potentially marking the primary time an AI model has surpassed the common human IQ.

The main points:

• o1 appropriately answered 25 out of 35 questions on the Norway Mensa IQ test.

• It demonstrated proficiency in solving complex visual and logical puzzles, each on publicly available tests and latest, unpublished questions.

• This performance suggests the advance isn’t resulting from the questions being within the training data.

• While OpenAI hasn’t officially confirmed these results, ChatGPT Plus members can test o1-preview themselves using the model dropdown inside ChatGPT.

Why it matters: If confirmed, these results would indicate that OpenAI has achieved Stage 2 performance on their 5-tier roadmap towards AGI. This milestone suggests the model can stop, think, and reason about problems to reach at correct answers more ceaselessly — an important prerequisite for Stage 3 (AI agents).

NEW TOOLS & JOBS

Trending AI Tools

• 🎬 Runway Gen-3 Video-to-Video – Video style transfer with the Gen-3 AI video generation model

• 🎵 Suno Cover – Reimagine the music you’re keen on with AI covers

• 👩‍🎤 HeyGen Avatar 3.0 – AI-generated avatars with emotions and tones that match your message

• 📊 Infography – Convert blogs into engaging infographics

• 🚀 FlutterFlow – Quickly generate code and styles for app development

• 📞 AIPhone – Translates and transcribes live calls seamlessly

Recent AI Job Opportunities

• 👨‍💼 Siena – Product Manager

• 💻 Hippocratic AI – Senior Full Stack Engineer

• 📊 Horizon3 – Marketing Project Manager

• 🎥 Captions – Technical Program Manager

QUICK HITS

Runway released Gen-3 Alpha Video to Video, allowing users to rework input videos with AI-generated styles and prompts across all paid plans.

Meta (Facebook) admitted to scraping public data from all Australian adult users to coach AI models, without offering an opt-out option available to EU users.

Google AI Studio launched a brand new model comparison feature, allowing users to simply compare outputs from different AI models and parameter settings.

Researchers developed “g1”, an AI system using Llama-3.1 on Groq hardware that implements reasoning chains to resolve complex problems just like OpenAI’s latest o1 model.

A brand new AI chatbot using GPT-4 Turbo successfully reduced belief in conspiracy theories amongst users, with effects lasting for months after transient interactions.

Montana State University is developing AI methods using neural symbolic regression to assist farmers maximize crop yields through precision agriculture.

Researchers are developing AI-piloted drone swarms to detect and extinguish wildfires, with teams of as much as 30 autonomous aircraft working collectively to forestall fire spread.

THAT’S A WRAP

SPONSOR US

Get your product in front of over 650k+ AI enthusiasts

Our newsletter is read by hundreds of tech executives, investors, engineers, managers, and business owners around the globe. Get in contact today.

FEEDBACK

If you’ve got specific feedback or anything interesting you’d prefer to share, please tell us by replying to this email.