Tech Topic 70: Google Vids for free, AI Avatars, Gemini Nano Banana Limitations

In this week's Tech Topic, Nina, Peggy & Bob shared three tips for using Google's tools for creative work:  the newly-free Google Vids video editing app, using AI avatars in Vids, and attempting (unsuccessfully) to use the latest Gemini image editing model to create a labeled photo diagram.


This week's background is a photo of the Dolomites by Nina Trankova.

Peggy Kolm: Google Vids Free Video Editor


Google Vids start page at vids.google.com

Google has expanded access to Vids to free Google Accounts. Vids was originally limited to Google Workspace Business and Education users. 

Vids lets you record yourself or your screen; add images and video clips from Drive, Photos, or your computer; import a Google Slides slide or use a template. 

Currently Google Vids only creates videos in landscape orientation, but soon it will be able to create square and vertical videos as well (a much wished-for option!).

Free accounts do not have access to Google Vids Gemini AI features, such as video clip generation and AI avatars. That requires either a Google Workspace account or personal account with Google AI Pro or Ultra subscription.

More Information

Nina Trankova: Use AI-Powered Avatars in Google Vids


How to use Google Vids AI Avatars (click to enlarge)

Nina demonstrated the new AI-generated Avatars you can use in Google Vids. 
  1. Navigate to Google Vids (vids.google.com or docs.google.com/videos)  
  2. Enter a script.
  3. Select an AI avatar.
  4. Preview and select an avatar, which then generates audio and video.
  5. Review and make changes.
  6. The system performs a "lip synchronization process."
  7. The final video can be exported and downloaded.
OnEBoard is experimenting with the AI Avatars, and Nina notes it will be interesting "whether we reach some new audiences via these automated tools that Google is rolling out for the work accounts."


More information

Bob Danley: Limitations For "Nano Banana" Image Generation  


Details of Bob's attempts to get Gemini to generate an accurately labeled dragonfly image.

Bob attempted to use Google's latest image generation and editing model, Gemini 2.5 Flash Image (nano-banana), to create an image identifying body parts of a dragonfly, a "tedious task in Photoshop."

His first result: "I did get lines and text. however they didn't identify the body part correctly and in some cases actually had the wrong color."

He then tried additional prompts, with no better results. See the slides above for details.

He concludes: "However, the things that it does do are pretty incredible. So, don't give up the ghost yet. I'm just telling you that we're not going to be able to use that just yet to do our tedious tasks that we usually do in Photoshop. Hopefully it's coming."

More information

OnEBoard will be back with 3 Tips in 10 Minutes on September 7.


Comments