(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
 . . . . . . . . . .



A Fool and his Data are Soon Parted [1]

['View Profile', 'The Boss Gomez', 'Joe Democrat']

Date: 2026-04-30  

Compelling evidence indicates that PocketOS, an automobile-fleet management company/product, had its working user database and backup (at least) deleted last week by an “AI” agent using Claude. https://letsdatascience.com/news/pocketos-founder-reports-ai-agent-deleted-database-cd21d9af is the best summary I found. Needless to say, their customer service experienced a big interruption.

A jaundiced eye might assess that PocketOS founder/proprietor Jeremy Crane naively attempted to eliminate human labor hours from his firm by replacing them with “AI” without testing, backups, firewalls, or limiting permissions. This totally predictable mistake resulted in large losses.

The excuse noises of the decision-maker, Mr. Crane, are both clueless and predictable: blame someone else. “Everyone needs to know that these infra-providers and LLM tooling companies say that they have heavy safety guards, but they are not there,” Crane told PC Gamer.

Well, duh. The entire LLM and “AI” industry has completely abandoned any regard for and, much more importantly, any spending for safety assurance. Not just the SkyNet, kill-all-the-humans safety concerns. ANY safety concerns. Whatsoever. It is hard to believe Mr. Crane did not know this. Any responsible decision-maker would know this.

ALL of the leading “AI” multi-national corporations have quite publicly eliminated their Safety teams in the last couple years. Rumor and resignations indicate the last holdout, Anthropic, finally gave up a few months ago and is laying off or reassigning their entire Safety division, in order to “stay competitive.”

Any jaundiced eye can see that the entire “AI” industry has established an industry-wide standard that Consumer Safety is never their problem, and that “they are not responsible for anything.” (Sound familiar?) All responsibility/liability falls upon some human shoulders, somewhere. “AI” is merely a tool, like a wrench or a drill, and hence blameless! Look at Tesla’s “Self-driving” accidents and their legal consequences. More than anywhere else, caveat emptor applies in the “AI” tool industry. Any awareness whatsoever of news and current trends in LLMs, “AI” agents, and the “AI” industry would make one aware of this.

Nevertheless Mr. Crane blames his vendors for doing exactly what ALL vendors in that industry do now, and feigns outrage that his foolishness was costly.

If his eyes hadn’t been dazzled by the prospect of getting rid of those reliable (but parasitic!) human employees, he might have looked a little bit at current events and given some thought to risk management.

But then, he wouldn’t be the Fool of the title.

[END]
---
[1] Url: https://www.dailykos.com/stories/2026/4/30/800031039/community/a/

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/