Hello Hollywood AI Reader!

Welcome to this week’s edition of the Hollywood AI newsletter, where we bring you the latest & greatest in AI technology news for the entertainment industry. Here’s what’s on tap today:

• 🎤 Fake Drake Pulled From Music Platforms

• 🗣 Amazon Prime Releases New Dialog Boost Feature

• 🎬 NAB Show Dominated By AI Talk

• ✍🏼 AI Tool Spotlight: Descript

• 🍿 AI Kernels

🎤 Fake Drake Pulled From Music Platforms

The entertainment industry was faced with a new type of problem last week when an AI-generated song by “Fake Drake” appeared on multiple music streaming platforms and became an unexpected hit. The song, titled “Heart On My Sleeve,” was posted to TikTok by a user called Ghostwriter977. The song replicates vocals by the artists Drake and The Weeknd, and sounds pretty spot on.

The song has since been pulled from multiple platforms following a complaint by Universal Music Group. It was removed from Spotify and Apple Music on Monday. By Tuesday it was pulled from Amazon, YouTube, SoundCloud, Tidal and TikTok—where it was streamed over 15 million times.

The song’s success could be the start of a new era where music listeners and platforms both need to confirm the artist is really who they claim to be. The masked producer of this Fake Drake hit announced that this is just the beginning of more AI-generated music to come.

UMG released a statement saying “the training of generative AI using our artists’ music” represented “both a breach of our agreements and a violation of copyright law.” The company also added that platforms had a legal and ethical responsibility to prevent the use of their services in ways that harm artists. Just last week, UMG asked the major streaming services to block AI companies from “scraping” lyrics and melodies from their sites.

This leads to the question: If a song is an original composition, can you copyright a voice? Universal is clearly arguing yes, although it seems this issue will end up in the hands of the courts.

This is the start of what's likely to be an ongoing and complex conflict between the use of artificial intelligence in pop culture and efforts to protect copyright.

🗣 Amazon Prime Releases New Dialog Boost Feature

A new accessibility feature called Dialogue Boost is being launched by Amazon Prime Video, which aims to improve the audio experience of its movies and shows. Using AI, Dialogue Boost analyzes audio to identify points where dialogue may be difficult to hear and isolates the speech before enhancing its audio to make dialog clearer. Using this AI-driven approach, only certain sections of dialog are targeted, instead of applying a general amplification to the entire dialog track.

While Dialog Boost was built with the needs of hard of hearing customers in mind, anyone can use the feature to suit their personal listening preferences.

During playback, viewers can choose the level of Dialog Boost they’d like to apply by accessing the Audio and Subtitles drop-down menu. The audio tracks are listed as “English Dialog Boost: Medium” and “English Dialog Boost: High.”

Currently, Dialogue Boost is only available on select Amazon Originals, but Amazon plans to expand its availability in the future. Previously, the feature was only available to customers with high-end audio setups, but it can now be used on all devices that support Prime Video playback.

 🎬 NAB Show Dominated By AI Talk

The National Association of Broadcasters (NAB) Show, an annual event that brings together professionals from the broadcast industry to showcase the latest advancements in technology, concluded on Wednesday. The event, which takes place in Las Vegas every year, was dominated by AI talk.

NAB president and CEO Curtis LeGeyt spoke about the potential dangers and benefits of AI during his state of the industry presentation, noting that the relevance of AI to the broadcast industry has gone from an amorphous concept to a reality. LeGeyt cautioned about concerns he has for local broadcasters, such as the degree to which big tech platforms use their platforms to access broadcast content without providing fair compensation for the content, and called for legislation to address the issue.

Many exhibitors at the NAB show demonstrated AI-driven technology, including IMAX, which made its first appearance as an exhibitor this year. The company demoed current and prototype technology from SSIMWAVE, an AI startup that it acquired in 2022, showcasing tools aimed at bandwidth and image quality optimization. Adobe also presented a new beta version of Premiere Pro that includes an AI-driven, text-based editing tool developed to analyze and transcribe clips.

While there were many advancements in technology showcased at this year's NAB show, it was viewed as an evolutionary year rather than a revolutionary one. Read more about it here.

✍🏼 AI Tool Spotlight: Descript

Descript is an AI-powered editor for videos, audio, recording screens and transcribing. Upload media or record directly in Descript to instantly transcribe your file into text, then tweak the text to directly update your media clips. Editing your media is as easy as editing a Word doc.

One of the best aspects of Descript is it’s intuitive interface. The tool is easy to learn and is a valuable, time-saving solution for podcasters, YouTubers and interviewers.

Transcriptions are fast and somewhat accurate. You’ll still need an editor to go through and edit your transcript if you need it to be 100% accurate.

Editing out filler words and silence gaps can be done with a single click. This function is pretty impressive.

You can also record your screen and webcam, edit out mistakes in seconds, and publish to a link for sharing right away. This is super useful for creating online presentations.

Pro editing features are also available, such as multi-track editing, live collaboration, auto-captioning, and much more.

🍿 AI Kernels

Here’s a round-up of interesting news and stories that caught our attention this week:

New Adobe Lightroom AI innovations empower everyone to edit photos like a pro. New AI-powered features for its Lightroom ecosystem, including Denoise and Curves for masking, enables everyone to make precise and highly customized edits. Plus a new black & white feature for video. → adobe

Google’s big AI push will combine Brain and DeepMind into one team. The company believes this move will significantly accelerate their progress in AI. → google

Introducing video editing with VidStyleODE. VidStyleODE is a tool that uses AI to help with video editing. It can change the appearance and motion of video content, create animation from still images, and even make videos longer or shorter by adding or removing frames. → github

If Pixar made the Harry Potter series. The fun images generated by Midjourney appeared on Reddit this week. Check them out → reddit

That’s a wrap for today’s newsletter. Thanks for reading Hollywood AI!

If you enjoyed today’s edition, we appreciate you spreading the word! New to Hollywood AI? We’ve got you covered. Click the Subscribe button below and we’ll make sure you don’t miss out on any of the fun.