Free The Kids!

Education Bureaucracy, Claude, LA Riots, &tc.

Dear Friends,

Welcome to The Quarter Inch!

As I get older I find myself chilling out. I just don’t hold to some things quite as tightly as I used to. Things I thought of paramount importance now somehow produce a shrug. And yet, on other matters I find myself getting almost maniacal. I think there are actual convictions behind these changes, or at least I hope so. I hope it is “sanctification” that is changing how I view and prioritize things.

As I sit here on a glorious June morning in the golden sunlight and leafy green aura, listening to the robins sing and the chickadees sounding their mating calls (I can’t believe those guys haven’t actually paired up yet—maybe they have, but they still like flirting), it occurs to me that the children of my city are in classrooms. Yes, it is true. Our local school district is keeping children in school until June 14th this year. Now, you all know that I hold the institution of public education is very low esteem, but this travesty makes me almost lose my mind. Only a worthless bureaucracy could make this up. In the interests of making sure the kids are in a classroom for the magical “180” days, we shall deprive them of the precious—more valuable than gold—days of sublime Montana weather.

Nobody is teaching anything at all this week (or even the last two weeks). Nobody is learning anything at all this week. You know it. I know it. Teachers know it. Students know it. Parents know it. Administrators know it. The school board knows it. And nobody cares because we all kowtow to the bureaucracy. I am at the point of thinking, sitting here in the glorious sunshine with a cool breeze from the northwest, that this is immoral.

Free the kids from their prisons.

---

I ran across this fabulous article about Artificial Intelligence and found it quite compelling. It’s responding to an academic paper Apple released on the limitations of LLMs. Gary Marcus is pretty convinced that LLMs (Large Language Models) have basically peaked. What developers are trying to do is create an “LRM,” as I understand it: a Large Reasoning Model. And they are failing. AI platforms like “Claude” are impressive in what they do, but what they do is less impressive than we are giving them credit for. The upshot, I take it, is that we should be quite a bit less worried about AI “taking over” everything and making everyone’s job obsolete. It just isn’t up to a whole lot of tasks. You are not going to trust these things to automate, say, air traffic control or public utilities.