Someone recently asked me "when did turning computers off and on again to fix problems start?" A colleague, Bob Doran, considered this and replied: "My guess is that it was not done during the valve computer era because turning off then on would almost guarantee a valve would die. With large computers, too, I am sure that turning the machine off would be avoided because the “cold start” was a long process. Reboot the computer, definitely. The reason that it works is that the problem is likely memory corruption and off/on clears all writable memory. So that means its from the 70s with PCs and DRAM memories. Mind you, I think that the technique was also applied to other devices."
Given the recent scandal concerning Facebook and Cambridge Analytica many people are reconsidering their relationship with Facebook. By all means, delete your FB account, you'll find it's not easy as they prefer you to "deactivate" your account. I’m not deleting Facebook, because it’s a useful way of keeping in touch with friends, particularly overseas. But, please go into the settings and restrict all your privacy settings to "friends only." Deny access to your contacts. Then remove all 3rd party apps that have access to your Facebook, then lock everything else down to friends only. Remove all your personal info, birthday, school, Uni, workplace, home address, mobile number, etc... yeah it takes a while. Then set location to "only when you use the app". Finally, open Facebook enjoy your friends' posts and then kill it every-time you finish, never let it run in the background.
A couple of weeks ago I was approached by a journalist, David Farrier. He had a very interesting story and was looking for my AI opinion. He had come across a company in Christchurch that allegedly had created a groundbreaking AI, called Zach AI, that could create patient-doctor consultation notes. The AI had been created by Uni drop-out Albi Whale and the company went by the unlikely name of "Terrible." Did I think this was for real? I first went online and could find no record of an A. Whale ever publishing at an AI conference or in a journal. It seemed unlikely that a single person could beat teams of researchers at Google, Amazon, Apple and IBM so I asked some colleagues working at these companies if they'd ever come across A. Whale or Zach AI. All came up blank.
You can read David Farrier's (long) article here and a recent follow-up piece here. It would seem that scammers are jumping on the AI band waggon.
AI and machine learning has been in the news almost constantly for the last year, so you'd be forgiven for thinking that the era of decisions being made by computers is now here and people should just let the AIs take over. However, we must be careful. A program in use since 1998 in the US, called Compas (Correctional Offender Management Profiling for Alternative Sanctions), used to assess more than a million US defendants may not be accurate enough for potentially life-changing decisions. The accuracy of the program used for bail and sentencing decisions has been called into question after it was found to be no more accurate at predicting the risk of reoffending than untrained people. Read more about this in this Guardian article,
You may have heard of Google's deep learning system AlphaGo that learned to play Go better than the best human. Well, the latest incarnation, AlphaZero, has learned to play chess better than the best computer chess playing system, Stockfish. However, what is most remarkable is that AlphaZero took only four hours to learn how to play chess! Read all about it in Chess.com. Machine learning is advancing very rapidly.
Mozilla has just announced the initial release of Mozilla’s open source speech recognition model that has an accuracy approaching what humans can perceive when listening to the same recordings. They are also releasing the world’s second-largest publicly available voice dataset, which was contributed to by nearly 20,000 people globally. This looks like it will be a very useful resource for researchers.
