Discussion about this post

User's avatar
Don Crawford's avatar

Having managed charter schools where all children learned to read and do math at or above grade level, I know we can do better. It doesn’t require more money. It requires that teachers teach differently with effective instructional methods and curriculum. Competition for customers will incentivize the adoption of more effective methods and curriculum. The details of education are critical to success. A great school is like a great restaurant. The people in charge have to know a lot about how to achieve excellence and must be highly motivated to do so. They must fear the consequences of doing a poor job enough to be willing to change to what is more effective. Our current system, that is insulated from the consequences of doing a poor job,cannot reform itself.

Jefferson Kim's avatar

I used to think university was worthless except for STEM fields. Now, with AI reasoning models averaging around 115 IQ, I'm questioning even the value of traditional high school education.

The article documents how two-thirds of Idaho students can't identify arguments in text or form opinions from what they read. Meanwhile, we're debating phonics instruction while AI is fundamentally changing what literacy means. A middle schooler who fully understands how to navigate AI across all subjects may be better prepared for the future than a high school graduate who's mastered traditional academics.

This represents a meta-shift in education. It's no longer primarily about rote memorization or even comprehension in the traditional sense—it's about the ability to leverage existing technologies and become effective AI "handlers." When AI reasoning likely surpasses 130 IQ in the near future, that skill becomes more valuable than conventional subject mastery.

The core competencies students need are evolving: they must learn to filter out AI-generated slop, communicate effectively with AI systems, and maintain discernment about outputs. The phonics debate, while important for baseline literacy, may be addressing yesterday's problem. The real question is: what's the minimum literacy threshold needed to effectively direct and evaluate AI, rather than be led astray by it?

We're spending $78 million on reading instruction while the fundamental nature of reading's role in society is shifting beneath our feet.

No posts

Ready for more?