Hi! I know a lot about AI and I would classify an AI as something that alters itself to get better at a certain task. (Mostly Neural networks) I have a few things to clarify: Stable diffusion doesn't have filters for faces or people. A lot of websites that use it can choose to use a filter that is meant for NSFW, but it is just too sensitive. Also, you can think of a neural network as a function that takes each input as an x and does some multiplications and additions to get to a y. At first, the multipliers and adders(weights and biases) are randomized, then a sample(like an image of a cat or a dog as the pixel values) will go through the calculations and they will end with a predicted Y(Maybe a 1 or a 0 for a dog or a cat). Then it calculates what the weights and biases would have needed to be in order for the predicted Y to be equal to the true Y and it slightly adjusts the weights and biases to be closer to what they would have needed to be. And it repeats this for every sample in a dataset Chatbots are neural networks that try to predict the next word from a list of words when given a previous text. Their inputs are words stored as numbers and their output is an index from a list of words/word-parts Sophia is based on chatbots like chatgpt Some companies like google use neural networks to try to get you to click on a video or stay on youtube for as long as possible to make the most Ad money I think Loab is just a consequence of the weights and biases being randomly initialised and the model not being big enough, so it only learned one face But overall this was a great video
I'll be watching this video an entire week since I vacuum pretty quickly, but yeah I used to vibe to some lofi beats but this is more interesting than music when I vacuum
