Today, the United States grounded all flights for several hours. There was a telecommunications issue that impacted the entire country as well as incoming international flights. That a single failure could ground the country is inexcusable. Yesterday, it was Microsoft that had the nationwide failure. By consolidating the service and thus making it more efficient and cheaper for Microsoft to run, it created another point of failure that crashed the system nationally. Do the benefits of efficiency outweigh the risks of single point of failure? No. These systems either need more robust backups or should be split into regional hubs.
The problem with flight communication started in the Chicago area and migrated. The Microsoft issue was a series of servers that failed and then were unable to restart without being overwhelmed as people tried to reconnect. In both cases, there is talk of international hacking, which would not be surprising. Recently, Europe saw many grocery store freezers turned into ovens that ruined vast quantities of frozen food. Why do you need the freezers connected to the internet? With the rise of AI - or should I say SI for Superintelligence? - anything connected to the internet will be subject to intense hacking. It might be time to disconnect.
No comments:
Post a Comment