How to Use Google Flow Music (Free AI Music Tool)
How to Use Google Flow Music (Free AI Music Tool)
Is generative AI leading to a “sea of sameness” in advertising? This week, we’re tackling the creative burden – from the pressure to produce endless aspect ratios for YouTube to the fear of losing brand control in an automated world.
In this episode of Ads Decoded, host Ginny Marvin sits down with Charles Boyd (Group Product Manager – Creative, Google Ads) and Sarah Hathiramani (Director of Product Management, YouTube Ads) to bridge the gap between brand fidelity and AI-powered scale.
In this deep dive:
- Avoiding the sea of sameness: How to use generative AI to stand out rather than blend in.
- Understanding Ad Strength: Why it’s a diagnostic metric for discovery, not a pass/fail performance report.
- Advertiser-in-the-Loop controls: Using text guidelines to steer AI while maintaining brand control.
- PMax Asset Experiments: How to use the new guided A/B testing to isolate creative impact.
In the context of Google Ads, cosine similarity is a mathematical algorithm used under the hood to calculate ad strength (6:38–7:30).
Specifically, it measures:
- Asset Uniqueness: It determines how distinct your headlines and descriptions are from one another.
- Diversity of Combinations: By assessing how unique these individual text elements are, the algorithm provides a real-time indication of how many unique combinations can be generated for your ads.
As noted in the video, this serves as a diagnostic metric—it helps you understand if you have provided enough varied assets to reach different types of searches, rather than acting as a direct performance or ‘pass/fail’ report (7:33–8:50).

Read more
Lyria 3 | Music Generation comes to Gemini
Google Vids: Generate custom music with Lyria 3
Flow Music and Believe bring next-gen tools to artists
for more refer Artificial Intelligence website click here
