YSK long noise videos cant effectively be compressed and as such take quite a lot of storage space and bandwidth. So if you want to keep the hosting costs of social media platforms low, for example when going public, you definetly wouldnt want disenfranchised users to upload them to your platform. This is the kind of noise you would want to avoid: https://stackoverflow.com/questions/15792105/simulating-tv-noise#15795112
So say you have a picture, and it’s made up of pixels, and you want to send that picture to someone but in order to do so, you have to make it smaller. You could send the most important bits and allow reconstruction on the receiver’s end, or you could some how make it smaller without changing the information. So if your picture is four blue pixels, followed by 3 red, and 2 yellow you could send the entire string like that, versus blue, blue, blue, blue, red, red, red, yellow, yellow. This would be lossless and are generally GIF, PNG, etc. JPEG is lossy compression, and it would be like telling your friend receiving the picture “I have a picture of a bird, here’s part of a beak, one wing, a tail, and one foot.” Your friend, being smart, can reconstruct the data that wasn’t sent (other wing and foot, body) because they have a good idea how the rest of the bird should look based on the parts they see. Lossy is better for smaller compression, but lossless is important if all the information needs to reach the receiver. Hope that helps.