We have all heard the hype around ai, its supposed to be dangerous. In other threads we have discussed ways it could become harmful vs the benefits it could bring. One topic has been mentioned without being explored, the ability to crack passwords. I say this is an easy threat to stop.
If indeed a super computer using modern cutting edge technology can crack any password, is an important question. If passwords became useless then identities can be stolen, bank accounts emptied and secrets stolen. We would have to go in person to make a payment or collect a salary. It wouldn't end civilization but it would make life more difficult
Brute force hacking seems to be the go to method. Other methods use known data about the person since this is often used in creating a password. Names, dates, titles of books and so on might give a clue. But can brute force really crack any password? It first uses a dictionary attack, might try several languages. Then, it tries making lists of possible combinations but random letters, numbers and symbols can make a list of trillions of possibilities or much more.
It can generate such a list in a short time so what is the problem? The problem is that even if your password is on the list, they still have to try it out to see if its the right one. That is where the bottleneck comes. If it takes one second to verify a password, then if its not on the first easy lists they try, the dictionary list and random list is going to take centuries or even millions of years to try
Most secure systems will allow only a few wrong tries and the account is shut down for a period of time. Therefore, a brute force attack will only work in a few limited cases where the system has no safeguards in place and will process a request rapidly or when social engineering pays off
So don't use short easy to guess passwords, don't use your personal info to make it and use some symbols along with letters and numbers. Do that and the mighty AI monster is slain, he can not crack your account. The nuclear launch codes are probably the most secure codes and would be long, would use all those tricks and may require a fingerprint or other safeguards that they don't publicize.
If someone calls and its a relative in trouble asking for money, simply ask them something that others would not know. What happened at your graduation from college, what is the name of your dog, what is your girl/boy friends last name, when did you call last? and so on. If they can't answer and just keep begging, its a scam but many people will not think of that and it will take a while before people learn about deep fakes but this is relatively easy to stop.
