I recently have began studying more in-depth towards my insatiable interest for AI; through Coursera. I also just today finished a video made by Mr. Beast in collaboration with Team Seas and The Ocean Cleanup on the machines and other cool devices and methods they’re using to help clean up the ocean. Through their fundraiser (which you can conveniently donate here 
https://teamseas.org/ ) their using methods such as trash tracking through aerial methods. However through my understanding of image datasets, wouldn’t we technically (using the rule of thumb of AI can do things that require only second thought) be able to teach an AI through hundreds of thousands of images of trash to detect and find trash in the Ocean? Possibly even engineering the necessary robots to make these collections? Bear with me here, realistically this probably has only a very limited economic rationale in the sense it isn’t a marketable industry, it’s purely charitable. But couldn’t one technically say… protecting our oceans is a matter of global importance. Our Oceans provide a whole means of transportation, food, etc. Being that this trash can cause a lot of death in millions of sea life, as well as the trash is poisoning a lot of wildlife, hence damaging eco-systems can somebody confirm this would be somewhat feasible or am I crazy?
I’m not sure finding it is the core problem. The trash field in the Pacific Ocean is 1.6 million square kilometers…roughly 3x the size of France.
