Imagine that you are about to enter the world, but you don’t know where you will land. You don’t know whether you’ll be born into an affluent family or one struggling to pay the bills. You don’t know whether your parents will be highly engaged in your education or barely engaged at all. You don’t know whether you’ll live in a community with excellent schools, a rural area with limited options, or a neighborhood where the assigned school just isn’t very good.
Now design the education system you want to enter.
This is an education version of philosopher John Rawls’ famous “veil of ignorance.” Rawls asked us to imagine ourselves, in effect, before our own birth, choosing the basic rules of the society we are about to enter without knowing where we will land: our wealth, social status, abilities, race, gender, or other characteristics. I find the thought experiment compelling.
Perhaps that’s partly because I’ve had a lot of good fortune. I was raised in a stable, relatively affluent family, in one of the wealthiest societies in history. I was also good at some of the things schools happened to reward. None of those were accomplishments of mine when I entered the world, and things could easily have been very different.
What happens when we put education policy behind Rawls’s veil? I think it quickly exposes some other veils of ignorance.
Consider school choice. Choice can be enormously valuable to a student whose assigned school isn’t working. If you don’t know where you’ll be born, would you really want your educational opportunities determined largely by your ZIP code? But choice carries its own assumptions.
What if you are born into a family that doesn’t have the time, knowledge, resources, or inclination to navigate a complicated array of educational options? A system that relies heavily on parents choosing well may work wonderfully if you happen to get parents equipped to choose well.
The same problem appears from the other direction. Opponents of school choice often emphasize the importance of strong public schools available to every child, and there’s an obvious appeal to that. But behind the veil, you don’t know that your local school will be strong. You might land in one of the many places where significant numbers of students leave school poorly prepared, or don’t graduate at all, and you don’t get to choose your ZIP code before you’re born.
There’s an observation I’ve heard attributed to Rick Hess, that the left sometimes talks as though teachers can do no wrong, while the right sometimes talks as though parents can do no wrong. Of course, neither is true. There are terrific and mediocre teachers, just as there are deeply engaged parents and parents who, for all sorts of reasons, don’t ensure that their children get a good education.
None of this leads me to a particular policy answer. Instead, it brings me back to Rawls and the original thought experiment.
Imagine again that you are about to be born. You know something about the range of American childhoods, families, communities, and schools, but you have no idea which will be yours. You don’t know your parents, their income, your ZIP code, your abilities, or what kind of school will be waiting for you.
Before finding out where you will land, think about the education system you would want waiting for you. Maybe you would still make many of the same choices you favor today. But what would you worry about more? What safeguards would you add? Where would you want more options, more support, or a stronger floor?
And what assumptions about schools, parents, and students would you be less willing to make if you knew that, once the veil was lifted, you would have to live with the consequences yourself?
Liner notes
For those too young to remember pre-digital music (which is most people!): liner notes were the text that accompanied albums back when music was delivered via vinyl records that you could scratch and break, but that afficionados swear sound better than digital. When I was young and becoming interested in music, I would often study the liner notes during my many listens to an album to understand the band’s process, influences, etc. I’m offering something similar here and may do so in some future blog posts as I explain and document how I’m writing with AI. You can read the post without referring to the liner notes, or you can peer behind the veil to see some of the thinking that went into it.
This post was mostly AI-generated, although I drove the process. One morning a few weeks ago, I was reading a column on John Rawls. I had not sought it out; it was among my morning emails from a variety of sources of news and commentary. The piece piqued my interest about Rawls because it was partly negative towards Rawls, so I got into a conversation with AI (talking, not written) about Rawls and his theories. From the transcript of the conversation, it’s clear that my coffee hadn’t kicked in yet as there are lots of “umms.”
Here’s where it gets interesting. I left the AI discussion for a while and came back to it recently. In my recollection, I was sure that AI just told me about Rawls’ theory and I made a connection to education in my head and started thinking about a blog post. But that’s not what the transcript shows. Instead, after the discussion about the basics of the theories, AI said this:
“If you want, next time we can walk through the decision rule question with a concrete example, like healthcare or education”
I agreed, and AI then wrote:
“imagine you’re behind the veil designing an education system but you don’t know if you’ll be a gifted kid, a struggling one, from a wealthy district or a poor rural area. You might go one of a few ways: maximize the floor so every child gets a strong baseline. or maybe you prioritize choice and flexibility, so families can exit systems that don’t fit, or you will balance both. Notice that each can be argued from the veil, depending on your risk attitude and what you count as fundamental. That’s part of why the veil is so durable as a question, not because it dictates a single political answer. It’s more durable because it forces you to say out loud what you’re willing to risk someone else’s life on.”
From there, we had a lot of back and forth about ideas as I realized this could be the start an interesting blog post. I shaped a lot of it, but so did AI, which also cautioned me not to make it too much about my own background, so that paragraph is included but is brief. We dabbled with the idea of using data to make the case about probabilities that a person would be born into different circumstances, but dropped that path.
Also, although it’s not in the transcript, I paused the discussion that morning and picked it up while hiking with my dog later on.
I changed some parts of the draft post, especially the ending. I asked for a second and third versions, pushed it to drop a few obvious AI tells. Plus, my AI has already been trained on my own writing. The result is a post that was mostly AI written, and doesn’t to me seem like AI slop.
You may disagree! I welcome thoughts on this post or reflections on how you are using AI to think and write with you.