The Deletist

The recent events in the United States are scary to consider, on more than one level. One of the most dangerous is giving the Department of Government Efficiency (DOGE) workers so much access to data! Though we have some idea of what DOGE can access, we don’t really understand the full extent. And probably neither do they, which can be dangerous in its own way.

When we collectively started to go digital and mobile, our data came along. We all generate mass quantities of data about ourselves and others every day. Even worse, in most scenarios, we don’t even realize how much data we’re giving up control to or who is gaining access. Nor do we fully know what will happen to all that data.

Based on my experiences working in information and data governance, the little “g” governance is sometimes missing in big “G” Government. Basically, this means not enough controls or guard rails exist in systems controlling information and data. I’m willing to bet that the US government is similar to a lot of places I’ve worked. They probably didn’t put in enough controls and guardrails on their information systems when they were first created. After all, when the workplace started to go digitally, nobody was thinking this far ahead. Or perhaps not imaging this particular scenario could happen.

For example, an organization should put controls around highly sensitive and personal employee data. This would include information about an employee’s marital status, health issues, personal address and phone number, etc. Controls are a way to protect data and the people behind it. One control may be to restrict access to authorized personnel. Another could be to anonymize data when used in aggregate. Deleting data once it is no longer needed is a control often overlooked. However, it protects both former employees from and the company by limiting and minimizing damage if there is a cybersecurity incident.

With all this technology and advances, one might think implementing these controls is easy. After all, can’t AI do this for us? The short answer is “no.” Often the controls are missing because they weren’t considered part of the information system. It’s usually only after a disaster happens that people start to think about these things. By that time, it’s often expensive, or sometimes not possible, to implement controls. Now we’ll all be learning a valuable lesson about governing the Governed.

Lately, I’ve noticed AI (Artificial Intelligence) turning up in all kinds of places unannounced. For example, sometimes as I’m going about my day, my phone’s AI will randomly start answering something. Or looking something up. It’s called Gemini and I still haven’t figured out how to delete it. I feel like that’s not even possible. Gemini, it seems, is here to stay.

Every time I do an internet search, some AI-generated “overview” is the first thing to show up. I sometimes find this even more dubious and suspect than scrolling through regular results. First of all, I have no idea where the “overview” pulled the content. It could have been from random reddit-chats, poorly constructed web pages, rampant bot-generated disinformation, and something legit. There’s no way to know. At least scrolling through results provides me with the option to see the source. Or to review the information in its original context. The AI-overview likely mixes little bits of information from lots of different places.

What feels crummy about the whole situation is I didn’t ask for the “convenience” of AI to help me on my phone or with searching. It was just there one day. When Gemini appeared on my phone, it was automatically the new default. I use Google-assist when I’m driving and need to do something hands-free. However, I never use it during other times. I thought it was shut off. Then one day, while having a conversation, Gemini started looking up some stuff related to what I was saying. It felt creepy, intrusive, and not helpful at all. In fact, I had to spend a lot of time searching for non-AI generated responses to find out how to disable the feature. Even now, it will still occasionally show up, despite my efforts to turn it off in settings.

When I think of all the amazing ways that AI can be helpful, it’s not in these overt and unwanted ways. Rather, I imagine AI taking care of mundane, time consuming, and repetitive tasks. Or using it’s huge power to do useful things such as blur my image from photos that others take and may post publicly. Mostly, I think about having a choice about whether or not I want to use AI on my personal devices. Making it the default automatic option feels intrusive to me.

As you can see, my finished plate was definitely “Worth the Wait.”

Towards the end of 2023 I swapped my gas car for a hybrid right before winter started. I did feel a nervous wondering how the car would react to the cold weather. The 2023-2024 winter was extremely mild and the car was absolutely fine. This winter, however, is a different story and I’ve had a number of weird things happen.

My car couldn’t handle the frigid, polar-vortex temperatures one morning, around -18 degrees Celsius. I blogged about this before in “Such a Thing as Too Much Technology.” Naturally, it was a morning I had to commute to the office. Despite warming the car for several minutes many warnings flashed on the dashboard. The car is still pretty new, so I thought this was odd. It was rotating through 4-5 different warning messages about things that may have been wrong. I had a hunch it was related to the cold weather. As the car warmed up, the warnings gradually disappeared except for one that the brakes may not be working properly. They seemed to be working fine so I drove to the office. I left early that day in case my car wouldn’t start, however, by that time the message had mysteriously disappeared.

Last Friday I drove home from the office in a snow storm. It was windy and snow was blowing directly at my car. Somehow this impacted some of the sensors. I had a visibility warning on my dashboard and some of the side signals that let me know if another car is in my blind spot also weren’t working. Rather, they worked intermittently, which almost felt scarier because I couldn’t trust it. Luckily, I learned how to drive without any of these things. I also retain some “old-school” driving habits like turning my head to physically look if anybody is in my blind spot, despite the side mirror sensors.

I feel good that I learned how to drive when a car was still just a car. They didn’t have computers or fancy features displayed on large screens. While I do enjoy many of the conveniences, esp. the GPS, ability to play podcasts, and dictate messages, none of things matter as much as my safety and my passengers. It’s scary to experience the many strange ways that car technology fails as soon as the weather gets bad.