Google's Video-to-Audio Model (AI Tech update) 

NOTE: There's an echo on the audio playback that was caused by MY OBS recording, NOT Google's model. Still the quality of their model speaks for itself.

I eagerly look forward to when we can plug in thousands of classic memes of the ages (preferably unprompted) and seeing what the AI comes up with adding audio to it.

This is so useful, gonna be interesting to experiment with this at least

Great nuanced breakdown again, Jesse. We can expect tV2A will rapidly become better and better, as with any AI tool. When the output quality is let's say twice what it is already, the extent to which this will be a threat to music creators will mostly depend on legal developments. If courts rule these tools and its generated output legal, it would probably will get used more and more commercially and composers and music rights holders will see a decline in work or income. But, I love your 10000ft zoom-out reframe - how can this work FOR us and focusing (even) more on being original, unique, innovating!

I dont understand if this "Google's Video-to-Audio" is avaiable to users or not. It make a big difference because, lately, a lot of AI sites announce new AI services but without a precise date for the release. Is the case of SORA (we are waiting since months), the so called "udio killer" form Eleven Labs, the incredible real time ChatGPT "Omni" etc... It is becoming an unfair business practice... Anyway, you can be sure the examples are always ultra-cherry picked...

Game on.

The reverb in all of those examples is nasty and annoying. Getting better but still very poor quality IMO. 😎