Assume 90% of all emails sent is spam. Based on this information, a machine learning startup Spamalot Inc. claims to have solved the email spam detection problem: for any new email, with probability 0.9 they categorizes the email as `spam' and otherwise the email is marked as `not spam.' Will the method correctly categorize all future emails? If yes, brie y explain why. If not, what will be the accuracy of the method? Briefly explain your answer.