last year and earlier this year, ai youtuber matt wolfe was using AI to create his thumbnails. Not sure if he still uses AI, but he's got a website that catelogs all AI tools, including art-related tools. this is the website: futuretools.io his yt channel: matt wolfe (its basically an AI news channel. If new software comes out, he tries it.)

How To (Achieve Goal) (With Constraint) Examples: How To gain 1,000 Subscribers without publishing every week. How to Travel The World With Zero Money

So we just experienced a tough pill to swallow, and that is YouTube’s A/B thumbnail testing. Originally we thought this was a great tool to use and would help with picking the best thumbnail. BUT it comes with a cost, after searching/reading through Reddit we found countless of recent scenarios where other YouTubers that have used the A/B testing feature on YouTube have took a major hit on their analytics, compared to their other videos, they’ve noticed big differences in click through rates and also views in general. After posting our most recent video in almost 24hrs we have only 70-80 views... This is way off than our usual numbers on a 24hr time frame . We honestly think it is the A/B testing doing this and hope you guys don’t “waste” a good video by doing this. We found although it’s a good idea to do testing for bigger channels it actually might not be good for smaller channels. Maybe a better way to test is to have a group of friends (like us) or family that you could send your variations of thumbnails and ask them which they like best. Anyways let us know your thoughts, have you guys used A/B testing before ? What do think?