> I do the unthinkable and just give it a go for a few minutes
You do this on a societal level and you waste centuries. There's no need for thousands of people to watch a few minutes of a bad DIY video to conclude that it's crap.
Doesn't seem like society is all that concerned with how much time it wastes watching Youtube videos either way. But, in this situation why wouldn't prioritizing videos with more upvotes not work just as well?
Because you don't know whether the video on the top of the queue is one with like/dislike ratio of 90% or 10%. If it's the latter, you are inclined to get off YouTube and start searching elsewhere.
Relying on viewer voting to identify misinformation is a terrible idea. These videos are deceptive by design - and in our polarized environment the voting is probably a better reflection of the audiences's priors than the truthfulness of the content.
Some of us are looking to learn or do a specific thing, not just waste time on whatever videos YouTube is recommending. This weekend I am planning to install a tow hitch on my car. I highly doubt YouTube is going to recommend just the right video for that at exactly the right time. YouTube knows me pretty well, but I doubt they know whether I want to learn about dark matter or modular synthesizers at this exact moment.
YouTube's entire business model is to occupy your time; whether it's a waste or not doesn't matter to their bottom line. They will absolutely not do anything to minimize the waste.
If YouTube notices people keep bouncing from your video it will not perform well due to the algorithm. YouTube wants you to be watching content, and not spending time searching around for content to watch.
You do this on a societal level and you waste centuries. There's no need for thousands of people to watch a few minutes of a bad DIY video to conclude that it's crap.