Saturday, October 3rd, 2026

Hugh Howey had an interesting post the other day that I keep thinking about, where he associates a possible future where people do with mental labor, the same as the Amish did with physical labor. I won’t attempt to rehash his points, he makes them better than I would. How people engage in a future that has a new significant toolset is interesting to me. There is never ending talk and association with AI right now. Much speculation from all sides on how the world will be different with this new powerful technology. I don’t currently subscribe to the AI doomer side of the world. I also don’t think the AI accelerationists are correct either. I have been thinking more about AI as a toolset. And like any set of tools the evolution of how they are used and the impact to their introduction is difficult to predict.

As someone who spent their whole career in the software industry the technology of AI is something I can’t ignore. But, that isn’t the thing I think the most about. I’m more curious how people will decide over time to integrate a new technology into their lives. As Hugh points out in his post, the Amish are not anti-technology. They are against anything that they feel will loosen the community. If something is helpful and doesn’t in their view pull people from what is important to them, they will use it. In this case, the Amish are lucky, they have a structure and a community to enforce what they believe is good for their people. I am not sure we have that any longer in the US. The government hasn’t been able to do much lately to help shape our society in a universally agreed upon manner.

So looking to Congress to put in regulations that shape how we want AI to impact people’s lives seems like a lost cause. There are so many reasons for this to be a true statement. I can’t think of anything in recent memory that Congress did that helped shape our use of recent technology. Two decades after the introduction of social media, with well documented evidence of harms, no meaningful action.

So, it is going to be up to individuals to figure out for themselves how this new technology is leveraged or not in their lives. This may be a fine thing. But the idea has tradeoffs. Like any technology it isn’t good or bad it is how it is used. And for the most part I see the use of LLMs will find a point where they take away much of the repetitive tasks that people need to do every day right now. We have seen this a million times before. There aren’t many monks copying books by hand any longer. Or accountants hand writing journal entries into paper books since the spreadsheet became mainstream. LLM’s ability to write software is a capability that enables people to build tools that meet their specific needs. Sure the coding part of the software industry will be impacted but that was never the hardest part of building products. And now there is this idea that now that everyone can build their own software applications they won’t need software companies. I don’t think that will be true at least in the near future. Since the harder part of building products is the long term maintenance. Not the coming up with the ideas.

But what people will do with LLMs and other AI technologies maybe isn’t as interesting as what they may reject. The things that the technology could enable, but like the Amish, people have decided the cost is not worth the benefits. There are many things that AI could do today, that people are already reacting to. Writing is one of them. Most of the discussion is about how consumers won’t want to read something generated by a computer. And that may or may not be true. I think the bigger question is how the writers decide to leverage the tools or not. There isn’t a uproar that software is being written by computers today. I haven’t seen many calls to only support human written software. The writing of code part isn’t the interesting part. Sure, those of us who used to write software for a living had the ability to see the elegance in well written code. But the reality for a long time is the code that was written by humans was already being manipulated and optimized by compilers. The code that was written by humans was more of a communication of intent to the computers. Just like typing into Claude Code or ChatGPT Codex is expressing intent.

The ability to express intent, and have an army of agents go off and build the thing we want isn’t new. As a product manager that was my job. It was just that the LLM based agents have taken one step out of the equation. So the question is where will some people draw the line? Say, this activity is exclusively done by a human. I think we will need to find that out. And I don’t expect any type of universally accepted line to be drawn. As we have seen with social media, we shouldn’t expect our governments to be able to regulate us into what is best for us. Not that they shouldn’t but, given the lack of agreement on what is right, and the interests pushing back it is clear nothing will happen anytime soon. We have too many examples where it is clear that the harms outweigh the good, but no action. So we are back to needing to think about these decisions on our own. Where do the benefits outweigh the negatives in our own lives?