We’re Waiting Too Long to Talk to Kids About Their Futures (Julie Lammers, Britebound) — from thehumanist.xyz by Julie Lammers and Allison Dulin Salisbury
“By eighth grade, kids are already ruling out careers they’ve never had a chance to explore.”
Per Allison:
Julie Lammers, who leads Britebound, makes a compelling case for starting career exploration in middle school for precisely this reason. The goal isn’t to get a 12-year-old to choose a profession, but to help them discover possibilities they might otherwise never consider, while they’re still forming their sense of who they are and what they’re capable of.
I love the implication of this argument. If we care about preserving choice, we should start career exploration earlier, not later. Done well, it could be one of the most powerful ways to expand opportunity for young people whose circumstances might otherwise define the limits of their ambition.
Julie and I talked about what this looks like in practice, why career navigation should be a skill we develop rather than only a service we deliver, and what young people themselves are telling Britebound about the support they want. It’s a conversation that challenged some of my assumptions about when and how we should prepare kids for their futures.
Also from Allison, check out:
[Special Report] AI Could Radically Accelerate How We Learn About Learning — from thehumanist.xyz by Allison Dulin Salisbury and Mandy McLean
AI is making a virtual laboratory for human learning possible, opening the door to testing ideas we could never afford or safely attempt with actual students.
Recent advances in AI raise the possibility of building something education has never really had: a virtual laboratory for learning.
At the center of this possibility are synthetic students—AI models designed to simulate aspects of how real students behave and, eventually, how they learn. Researchers could test hundreds of ways to teach a concept, give feedback, or sequence a curriculum and identify the most promising approaches for different kinds of learners. Developers could try out a new assessment or AI tutor on thousands of simulated students before putting it in front of real ones. Districts could stress-test AI products for safety and begin to understand which tools might work for whom and under what circumstances.
True, this could make running education studies faster and cheaper. But change the cost of experimentation enough and you change what gets experimented on. The promise of synthetic students is that researchers could ask more questions, test more ambitious ideas, explore edge cases, and discard failures before spending years and millions of dollars studying them with actual students.