NVIDIA’s new text-to-video

PLUS: The Kinds of Generative AI

Good morning human brains, welcome back to your daily munch of AI news.

Here’s what’s up:

  • NVIDIA’s AI-generated videos 📹️ 

  • Investigating the leaders of AI 👑 

  • Elon Musk was on Fox News 🦊 

  • The browser integrating AutoGPT 🤖 

Follow @BotEatBrain on Twitter for breaking news.

APPETIZER

NVIDIA’s new text-to-video 📺️ 

Nvidia just unveiled a model based on Stability AI’s Stable Diffusion that can generate high-res videos of both art and real-world simulations.

Here’s what it can do for you:

1/ generate several minutes of full-length videos.

2/ generate videos with objects and scenes that were not part of the original training data set (thanks to Dreambooth technology).

3/ render your videos at 511 × 1024 pixels (major step up from previous renditions).

4/ be customized for different use cases (since its fundamentally just a lightweight skin on Stable Diffusion).

Take: from autonomous driving research to TikTok feeds with personalized videos, the next adaptions of this are going to radically change how we use and interact with video. We’re expecting to see exponentially more innovation here as more open-source implementations keep rolling out.

BUZZWORD OF THE DAY

Classification

A learning technique that allows machines to assign categories to data points. Classification models predict things like whether a customer will return or not, whether a certain transaction represents fraud or not, or whether a certain image is a car or not.

FROM OUR PARTNERS

Build MVPs without Dev Headaches

So you have a clear product vision, but how can you deliver the right software on time and within budget? Get your product playbook.

Enter stage left: The first product workshop to build better MVPs.

  • Choose only features that matter and add value

  • Create a roadmap for all development phases

  • Receive a professional document at the end of the workshop.

Save hundreds of hours in development costs with fully fleshed-out plans, templates, and SOPs crafted for non-techies.

MAIN COURSE

Investigating the kings of Generative AI (Chapter One) 👑 

Generative AI tech is changing the world, and its most promising leaders are startups with tall goals. Here’s a rundown of the biggest players:

1/ Midjourney: working on what are arguably the best text-to-image tools shaking up the creative industry. Currently at version five. Founded 2021. Valued at $1B+. 11-50 employees.

2/ HuggingFace: AI community promoting open source contributions. Acts as a hub of models for Natural Language Processing, computer vision, and other AI fields. Used by 5k+ orgs. Founded 2016. Valued at $2B. 51-200 employees.

3/ Cohere: building secure LLMs for developers and big companies. Powering Spotify, HyperWrite, etc. Founded 2019. Valued at $6B. 51-200 employees.

4/ Eleuther: non-profit AI research group following in OpenAI’s footsteps, building LLMs and training text-to-image tools. Founded 2020 from a Discord group. Valuation unknown. 36 employees.

5/ Stability AI: creators of Stable Diffusion. Working on breakthrough AI models for text-to-image tools, language, code, and video (latest image model). Founded 2019. Valued at $4B. 100 employees.

6/ Anthropic: AI research company building reliable and easily steerable AI systems. Creators of ChatGPT competitor Claude. Working on deep learning research and LLMs. Founded 2021. Valued at $5B. 100 employees.

7/ OpenAI: private research lab currently leading the AI movement. Creators of DALL-E & ChatGPT. Founded 2015. Valued at $29B. 375 employees.

P.S. This is the first part in a mini-series, we’ll be exploring how these communities and companies interact together next. Stay tuned.

MEMES FOR DESSERT

YOUR DAILY MUNCH

Think Pieces

Elon Musk talking AI with Tucker Carlson (April 18).

What machines can teach us about ourselves.

Copy AI vs Human writers.

Can AI kill the greenscreen? (8m)

Startup News

Snapchat launches AR lenses powered by generative AI.

Cortical Labs raises $10M for stem cells (that could eventually power AI).

Research

A multimodal group activity dataset: predicting student engagement in classrooms via recordings.

Tools

Bloks: the AI-powered productivity assistant (for notes, tasks, and meetings).

Dora: design and publish 3D and animated websites.

Oasis: generate emails and texts with voice commands.

GodMode: the best UI for playing with AutoGPT (no API key required).

Vocaloid: add lyrics and vocal melodies to your music compositions.

TWEET OF THE DAY

Tag us on Twitter @BotEatBrain for a chance to be featured here tomorrow.

AI ART-SHOW

Slumdog Billionaires - @withgokul

Until next time 🤖😋🧠

What'd you think of today's newsletter?

Login or Subscribe to participate in polls.