S11E27 – Hard Drive Failure, LG TVs and the AI Doomsday Hype

Episode Summary

This episode had no listener calls. Mike and co-host Howard started with the Cox and Spectrum merger and a report that LG TVs contain microphones, which led into a talk about privacy and why hiding information is not the same as securing it.

They then covered a federal employee accused of swapping processors in work computers, and a hard drive failure comparison that backed up Mike’s long dislike of Seagate. The second half was a long discussion comparing today’s AI apocalypse headlines to Y2K, in which Mike argued the risk is real but overstated.

Topics Covered

Cox and Spectrum merge

Mike said Cox and Spectrum have combined, so Tucson cable customers will end up with Spectrum. He said Spectrum has a lower customer service rating than Cox, which has historically been the better local provider, and he uses Starlink at home. Howard noted that the parent company of Spectrum was hit by a social engineering attack the week before, where someone phoned support and posed as IT.

LG TVs and privacy

Howard described an in-depth video from the channel Gamers Nexus saying LG TVs record audio and store it until they reach the internet. Mike said he could not confirm how much any TV or smart speaker keeps. He said he thinks privacy has been eroding for years and that hiding information, or security through obscurity, does not work now that AI can find what humans would miss.

Hard drive failure rates

Howard cited a study that he said ranked Toshiba and Seagate as the worst drives, with Western Digital well behind them. These figures are as Howard cited them and were not verified. Mike said his own experience matched, with Seagate laptop drives tending to die suddenly, and that enterprise drives such as IronWolf are better than consumer ones. Howard said the study found each extra ten degrees of heat adds about 2 percent failure chance, and that Toshiba failure rates quadruple past five years. Mike said spinning drives older than four years are a ticking time bomb, so keep backups and cool the room.

Y2K and the AI doomsday headlines

Mike compared AI warnings to Y2K, saying that crisis was avoided because people spent years fixing code before a known deadline. He said AI has no set date, so preparation matters, especially keeping critical infrastructure such as water plants off the internet. He said AI cannot supply its own electricity or water, that a copy spread across ordinary computers would be too slow to be dangerous, and that spawned AI agents are just short-lived working sessions. He said the post from the Anthropic employee who quit was a warning about racing without safety, not a claim that the world is doomed.

Calls

There were no listener calls this week.