Teaching in the rubble

Most of the corpus is builders talking shop. The education thread is the exception. It's the group's conscience.

It runs from casual 2024 experiments to a hard 2026 question: if the tools can do the work, what is school for?

The thread is anchored by Ying, an educator who keeps pulling the conversation back to a problem the builders can wave away and she can't — her students.

---

2024: AI as a thing to learn with

Education shows up first as personal exploration. James works a machine-learning challenge on his own. David talks about AI-assisted learning around his trading research. Optimistic, individual, no alarm. AI is a faster way to learn whatever you're curious about.

---

2025: the channel problem, and the franchise idea

Two concrete ideas surface.

Where do you keep up? The group maps the real channels — Reddit's Local Llama, specific YouTubers, newsletters — and names the uncomfortable fact: academia is the worst channel for timely information. Anything in a curriculum is accepted dogma, not the edge.

8am-ai.com as a pattern. David floats the peer group itself as an open-source, franchisable model — a way for others to run their own learning loop outside any institution. The meeting becomes the method.

By autumn the policy layer enters. David reports back from a conference on government, research, and industry in Canadian AI. The thread widens from "how do I keep up" to "how does a whole system keep up."

---

2026: the foundational-knowledge problem

The question stops being logistics and becomes whether the next generation can think at all.

Students use AI to produce acceptable output without the knowledge to evaluate it. A false confidence that they know things they don't. David's image: his daughter using Google Lens to clear an exam, skipping the struggle that builds the skill. Juan's requirement: you have to be able to combat an AI suggestion, to critique an approach — which you can only do if you understand the domain.

The working answers, none of them tidy:

Raise the complexity until AI alone can't clear the bar, so students use it as a learning aid, not a substitute.

Go direct. Bypass the slow institution — direct to student, direct to teacher — because the five-year curriculum cycle can't track a field that moves weekly.

Teach the meta-skill. Critical thinking, design thinking, first principles, airdropped into the curriculum as the durable thing. The specific tools won't survive the semester.

Use the students as compute. Assign them to become domain-AI experts. Have them do bad research and then validate it. Learn evaluation by doing evaluation.

---

Strip away the specifics and the education thread asks the same question as the rest of the corpus, where it hurts most. When the machine produces the answer, the scarce skill is judging whether the answer is true. The builders call it "how do you know." Ying calls it the thing her students can't yet do. Same problem.

The optimistic reading the group keeps returning to: this is a chance to teach the thing schools always claimed to teach and rarely did. There's no longer much point teaching anything else.