Welcome to VP Land - your daily update for all the latest news in movie tech.

In this issue:

👨‍💻 Amazon’s Bedrock platform adds powerful AI models
🗣️ AI vs. human on voice dubbing technology
🌆 Enhance or create your own virtual sets

And more! Let’s dive in.

PRINCIPAL PHOTOGRAPHY

Meet Bedrock: AWS’ Modern AI Family

Amazon is expanding their generative AI integrations at Bedrock, their AWS service for generating AI-powered apps via 3rd party models (and some of Amazon’s own developments).

This isn’t a surprise - we’re in an AI arms race and the major cloud services have their go-to models. Microsoft’s Azure has OpenAI (creator of GPT-4 and ChatGPT), and Google has its own models.

Who’s hanging out in Bedrock’s quarry?

• Anthropic, which just released Claude 2, their own text generation model that many have said is as good or better than GPT-4 (depending on the task).
  Does Manchester England from Hair pop into anyone else’s head when they hear Claude?

• AI21’s Jurassic-2 model

• Stability AI, which just released its new SDXL model last week (which we covered here), is the closest competitor to Midjourney when it comes to text-to-image generation.

Agents. They’re like Spacely Space Sprockets.

Also in beta are Agents. Swami Sivasubramanian, the VP of machine learning at AWS, describes agents as a new capability for developers to enable generative AI applications to complete tasks in just a few clicks.

He says agents can be used to help configure foundation models automatically and help to orchestrate tasks without having to write any code.

What’s it mean for video?

• Easier to deploy models and automatic image generation. Cuebric has already been using Stable Diffusion for creating 2.5D backgrounds.

• Doing things with the data you already have on AWS, like metadata generation or potentially training your own custom models.

• AWS Agents have been compared to the plugin system in ChatGPT. These are early days, so who knows what other APIs bedrock will be able to tap into. A few ideas: Storyboard creation, script generation, previz models, rough cut assembly.

SPONSOR MESSAGE

Safeguard your video content with Vestigit.

Vestigit’s innovative solutions, including Vestigit AI, Vestigit Sentinel AI, and Vestigit Watermark, provide robust protection against video content piracy and track illegal streams.

Don't let your valuable content fall into the wrong hands. With Vestigit, you can ensure your content remains secure and in control.

Experience the power of Vestigit’s platform firsthand - contact Vestigit now for a free demo.

SECOND UNIT

Aloud, YouTube’s AI Voice Dubbing Service, Brings Another AI vs Human Debate

Announced back in March, YouTube is rolling out its Aloud service, a tool that automatically dubs videos into multiple languages using AI generated voices.

Backstory

• Many big creators have been dubbing their videos into other languages for years (and running completely separate channels, like MrBeast en Español, which has 24 million subscribers)

• In the same March announcement, YouTube announced they’d roll out multi-audio support on videos, making multiple channels less necessary.

• YouTube is available in 80 different languages and less than one-third of YouTube’s most popular channels are in English (28%). So if you’re a big creator and want to reach the most people possible, you need to dub your videos.

Dubbing firms have popped up, offering translation and dubbing services to creators. But the pricing only makes sense for the top 0.1% of creators who have enough reach to make the translations profitable.

For the other 99.9%, Aloud makes sense. But could it eventually wipe out all human dubbing?

The case for human audio dubs:

Novelty in the voice actor. MrBeast has talked about how he gets famous voice actors in local markets to be his voice, so when people watch they get a ‘wait a minute’ moment and start commenting.

But, YouTube has said voice dubbing that sounds like the creator’s voice will be coming in 2024. Is that enough for big creators to switch?

Advertiser accuracy. In a great write-up in Rest of World, there’s this quote from a founder of a dubbing firm:

What do you think about AI dubbing in the future?

VIDEO VILLAGE

Joshua Kerr demos a new plugin for Unreal: SwitchLight from Beeble. Use AI to relight your scenes based on the virtual environment.

AKS

🌐 Around the Web

• How to enhance or build your own "virtual sets" for your videos using AI

• Google strikes gold with ongoing investments in AI for search and advertising

• Upload large files and folders fast with MASV*

• Wuxi Studios launches a virtual production lab studio showcasing China's largest LED volume powered by Brompton Technology and ROE Visual

• Black Eyed Peas' will.i.am launches his FYI (Focus Your Idea) app, designed to promote AI-powered "creative co-piloting"

• ARway.ai licenses its augmented reality spatial computing navigation platform for $500,000 a year

  * Sponsored Link

CALL SHEET

📆 Upcoming Events

August 3
SMPTE - The BIGGEST Virtual Production Show Ever!
XR Studios, Los Angeles, CA
Showcase of theatrical cinematic performances, driven by a menagerie of creative technologies within virtual production

August 4 - 5
frame:work:losangeles
XR Studios, Los Angeles, CA
Conference for conversations about the unique challenges of creating video for physical spaces, live performances, and virtual worlds.

August 
Studio Arts - Technology for Entertainment Creatives
Unreal Engine and virtual production online courses

Have a film-related event to share? Just reply back to this email and let us know the details.

ABBY SINGER

Thanks for reading VP Land!

Have a link to share or a story idea? Send it here.

Interested in reaching media industry professionals? Advertise with us.