And I know I'm just one person and most people aren't like me so my behavior/ contribution to metrics is hardly a drop in the ocean, but I try to live my life as though my actions do have consequences

(right now I kinda try to weigh the pros and cons, like if it's someone with a large platform already, I hold back until the algorithm shows me a "good" version of the title and thumbnail, while if it's a smaller creator I'm more prone to say "whatever". I don't like this system though)

I'd like the option of clicking through and saying "FYI I did this because it's you, despite the presentation, not because of it"

I feel stuck atm, because I'd like to support the creator and boost their early view and interaction count, but I also want to disincentivize click bait

or shows being resurrected (e.g. Scrubs s9 continuing after the s8 finale ended everything, so now suddenly they're teaching at med school instead of working at a hospital)
Heck, they reportedly tried to pitch a seventh season of Supergirl that didn't have Supergirl (thankfully it was rejected)

Not sure if it's exactly what you're talking about, but I think many TV series have a final season that feels "off" for production reasons that affected plot
This can come in the form of major cast members leaving (think Lisa Edelstein leaving House -> suddenly Foreman is Dean of Medicine),

As it is, it seems less related to the nature of the images generated and more the sheer volume of prompts. I wouldn't be surprised if they just did the math and figured it wasn't actually worth it to make available to all.
(Might have even been original roll-out plan, an unofficial free trial)

The ONLY way I could see this as (part of) a solution was if it was framed as "generating such images is not allowed. While we work on a fix to the model, we're limiting its use to paid users and will report violations to law enforcement, identifying culprits by pay method"

(This is not the case)

or wasn't legal
(The company in question is hardly one of the major players, but it was a Silicon Valley tech company developing an AI Voice Assistant. I worked there as a linguist. I can share more details via PM if anyone's interested)

I don't disagree with you on most things you've said, btw

Language Models, not LLMS) was reluctant to download youtube videos for its dataset because of YouTube's ToS, but not because of the creators' rights. I literally sat with the corporate lawyer and told him we couldn't do it, and he replied that his job was risk management, not determining what was

absolutely take advantage of things, but AI companies are 100% willing to consciously use things unauthorized if the cost of obtaining them legally is too high, as long as they're only stealing from "little fish". For example, a company I used to work for back in 2019 (back then we still called them

Thing is, the comparison shouldn't stop at using the end product - we should include training (and dev, though that's a much smaller number), and - in the case of video games - the energy used for game dev

As for plagiarism - I don't know that there's an easy solution, because large IPs will

Thank you!
Thank you for making the video
Thank you for sharing it with us
Thank you for caring

May you and your channel continue to grow 💜

Then, when you need to change it, you just move onto the next line:
N0weverybodygoesbelow!

It's always long enough, easy for you to remember, and decently hard to brute-force

For the record: I've never used this story in particular, my go-to is a different one 💀