⚡️ Google Releases VideoPoet

PLUS: Midjourney V6 brings revamped prompting methods

Good Morning. In a bold stride for AI technology, Google has just pulled back the curtain on VideoPoet, a new AI designed to revolutionize video generation. From the collective minds of 31 researchers at Google Research, this LLM is poised to redefine our expectations for synthetic video creation. Let’s dive in.

Today’s Highlights:

  • Midjourney V6 brings revamped prompting methods

  • AI predicts mortality with unprecedented accuracy

  • Warehouse automation firm GreyOrange raises $135M

DEEP DIVE

Google’s VideoPoet: A New Era of AI-Generated Imagery

Meet VideoPoet, the latest from Google's AI lab—that creates not just text but videos that wow. From mere words to synthesized sounds, this large language model does it all.

Google’s ambition shines with VideoPoet—it's a one-stop model that neatly sidesteps the need for multiple specialized systems, making it a juggernaut in video generation tasks:

  • Image-to-video transformation? Check.

  • Stylizing videos on the fly? Double-check.

  • Video inpainting, outpainting, and even video-to-audio? You got it.

But VideoPoet’s not just flexing its tech: In head-to-head comparisons, it's already stealing hearts. Some 24 to 35% of participants have shown a preference for VideoPoet-crafted videos over its peers like Phenaki and VideoCrafter.

While it’s currently wrapping sound around visuals like a pro—think a cat charmingly tickling the ivories—it’s the promise of "any-to-any" generation that’s got tech enthusiasts buzzing.

Imagine the Possibilities:

  • A script penned by AI, brought to life in full audio-visual splendor? Done. Google has already dipped its toes into this, crafting a short film via Bard, their scriptwriting AI.

  • Text-to-audio, audio-to-video, and even savvy video captioning? All in the cards for VideoPoet.

Yet, Google is playing coy with VideoPoet’s availability—no word on when (or if) we can get our hands on it, but whispers of a Bard Advanced tie-in linger.

Will AI change the video creation landscape?

Like a director with an AI megaphone, VideoPoet suggests so. It could revolutionize anything from education to entertainment, though for now, we can only drool over the high-res demos waiting on Google’s VideoPoet project page.

PUNCHLINES

Artificial Unintelligence: AI cannot be named as inventor—Supreme Court rules in a patent dispute.

Dirty Data: An influential AI dataset contains thousands of suspected child sexual abuse images.

More than Machine: AI robot ElliQ becomes a chatty companion for seniors seeking solace.

Cash flow: AI-driven warehouse automation firm GreyOrange raises $135M.

AI Gold Rush: Anthropic cozies up with Menlo Ventures for a stunning $750 million fundraise.

TLDR

Midjourney V6 brings enhanced AI art generation: Midjourney's latest update, version 6, introduces new features like in-image text and revamped prompting methods. While some results show marked improvement in detail and photorealism, the upgrade requires users to adopt new ways of interacting with the AI. V6 is currently in alpha testing, with frequent changes anticipated before full release.

New era for video generation with Stability AI release: Stability AI unveils its Stable Video Diffusion model for developers, enabling advanced video generation for apps and services. Amid controversy over training data, the API offers short, customizable video clips with potential for gaming, advertising, and more, challenging competitors with no API options.

AI-developed novel antibiotics combat superbugs: MIT researchers harness AI to identify new antibiotic compounds in the fight against MRSA, a superbug killing 10,000 Americans annually. Their innovative model, which eschews the 'black box' approach, revealed five new compounds with minimal toxicity to human cells, offering hope in tackling the global antibiotic resistance crisis.

AI predicts mortality with unprecedented accuracy: A new AI, trained on Danish citizens' personal data, can foresee the likelihood of death, outshining existing models by 11% in accuracy. The "life2vec" system analyzed millions of life sequences to predict mortality, with implications for personalized interventions, though ethical concerns limit its use by insurers.

TRENDING TOOLS

🎥 Creatify AI: Instantly generate short video ads from a product URL

📱 Create.xyz: Effortlessly create apps in minutes with no code

💻 GPT Engineer: Experience rapid prototyping at unparalleled speeds

🖼️ Photo AI on iOS: Generate unique AI photos on your iPhone

🤖 macOSpilot: Navigate your macOS with a savvy AI assistant

That’s all for today—if you have any questions or something interesting to share, please reply to this email. We’d love to hear from you!

P.S. If you want to sign up for the Supercharged newsletter or share it with a friend, you can find us here.

Reply

or to participate.