Generate music with Lyria

PLUS: New model for video generation by Meta

Anita Kirkovska
November 17, 2023

Hi folks!👋🏻 This is The Prompt!

Here’s what we have for you today:

Music generation model by DeepMind
Live video/image editing with KREA
New video model by Meta
First impressions with the Assistant API
Playbook on building an AI company

Let's read on

FEATURED

Lyria: Generate music with instruments & vocals

Music contains huge amounts of information, like voice+ instruments all playing at once, so it's much harder to create than speech.

But, Google Deepmind has now released their most advanced music generation model called Lyria.

It can create great music that sounds just like the real thing.

They opened up access to the model through two experiments:

DreamTrack: Limited access for a selected number of influencers that can create 30 second soundtrack for their Shorts using 9 artist voices . Here’s are two audio samples generated with DreamTrack using Charlie Puth’s and T-Pain’s voice.
Music AI tools: a set of tools to help musicians/songwriters be more creative. You can hum a tune and the tool will create music for you using the most suitable instrument. Here’s an example!

Any content published by the Lyria model will be watermarked with SynthID, the same technology toolkit that Google is using for their AI generated images.

🚨 What else is going on

Google’s Gemini model that should be multimodal like GPT-4 with Vision is reportedly delayed due to unknown reasons, and they’re now expecting to release it in Q1 of 2024
Real time editing of images and videos is here. Krea AI makes it possible, just look at the demo below. Another demo here.

Meta released new model for image editing called Emu Edit, and another model for video generation called Emu Video. Example below, and you can create yours on this link

📕 Resources

First impressions with the Assistants API
What does AI Agent mean today?
Playbook on starting an AI company by Hubspot’s co-founder

🧰 Tools

GPT for GPTs: 10k+ GPTs in a GPT
Plugbear: Connect Slack with OpenAI Assistant instantly
Outset: AI Voice interviews
AI Shadows: Generate realistic shadows for product photos
Pillar: AI powered checkout

✍🏼 Prompt of the Day

TOOL

Midjourney

PROMPT

amateur film photography, university students , furtive moment between doors, 1990, Leica 35mm --ar 2:3 --style raw --v 5.2

RESULT

https://twitter.com/hugovntr/status/1725423339114471825/photo/1

What'd you think of today's edition?

Subscribe to keep reading

This content is free, but you must be subscribed to ThePrompt to continue reading.

Already a subscriber?Sign In.Not now