Friends hear the worry in your voice and answer the worry. Fulcra takes the question apart instead: it matches the decision’s structure against a latticework of 167 mental models and returns counsel with every model named and sourced. Below, one session runs end to end, dissected as it goes.
I’ve been offered double my salary at a startup that might not survive two years. I love the team I’d be leaving, and I’m scared I’ll regret it either way. Do I go?
Fulcra doesn't weigh how you feel about the question. By construction, it can't. What gets mapped against the lattice is the question's structure: problem type, complexity, time horizon, reversibility. Only the models that load-bear are returned, named, and applied to your specifics. The rest of the lattice stays out of it.
The love for your team is real, and so is the fear. But the regret you’re bracing for waits on both branches, so it cancels. What’s left is structure: immediate gains against future resilience.
Go, but only if you can front-load enough savings to absorb the instability, and the role meaningfully widens the network and skills you’d keep if it ended tomorrow.
At 2× (the gap you named) the move carries convexity1: the instability is your cost of admission for an upside you can survive.
Two models temper it. Leverage2: the skills and network you’d keep if the startup dies are the payout that can’t be repossessed. And circle of competence3: you can’t judge their survival odds, so don’t bet on those. Bet on what you keep either way.
Counsel, reasoned: every model named, sourced at the foot. That, end to end, is the product.
Request an invitation →The strongest thing we make is the latticework itself, so here is its opening, A through C, exactly as it reads in the library: every entry named, sourced, and yours to curate. Open any entry and its summary, origin, and source come with it; mark a row and Fulcra reasons from your working set.
You’ve got to have models in your head. And you’ve got to array your experience, both vicarious and direct, on this latticework of models.C. Munger · 1994