It would be revealing to see some of the e-mail communications this person was sending to MITRE while trying to fight this issue.
We’re all familiar with how some use LLMs to write code and submit PRs, but there’s a growing problem of people using LLMs to tirelessly fight issues with communications like e-mails and even suggesting physical paperwork too.
Now that the effort to argue something against an institution is approaching zero, more people are getting the idea to have their LLM and harness fight some battles for them. It feels like it costs them very little but if there’s a non-zero chance of personal gain they do it. I’m hearing a lot of stories about everything from local governments to university administration offices being overwhelmed by requests that just keep coming in from relentless senders who figure asking for anything is worth a try even if there’s no chance it would be granted.
I think we’re going to have to rethink a lot of our communication and request systems that previously relied on the fact that most people wouldn’t take the effort to argue for something that they didn’t deserve. When the cost of arguing approaches zero, the machine can keep trying to get the non-zero chance of success for them.
There was a recent article in The Economist about people fighting employment disputes with AI (https://archive.ph/mi3zu). On the one hand, democratizing the system, on the other hand filling it with massive quantities of legally questionable slop.
It sounds like the only end result of this is them making the bureaucracy even more bureaucratic by employing their own agents, where the only way to get anything done is to have your agents fight their agents.
Frankly, it feels like Kafka was actually an optimist.