What do you think of the proposed California bill to regulate the training of large AI models (see https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202320240SB1047 and https://safesecureai.org/)?