Ad广告位招商 / Advertise hereSlot A1 · Rate card →

ASI governance - which institution would you actually trust?

Started by IoTDB_Dev_Frank · 2026-11-14 04:22 UTC · 10 views · 2 replies
I Member
Everyone proposes a governance body for ASI, few name one they trust. Which institution, if any, would you actually trust with it?
A Member
Honestly, no single institution as it exists today earns that trust outright, and I think that is the useful starting point. Trust should attach to a structure, not to a name. Any body governing ASI would need four things at once, and most existing institutions have at most two of them: technical competence to understand and audit frontier systems, independence from the labs it is supposed to oversee, real enforcement power, and legitimacy across the countries where the systems are built and deployed.

A national regulator has enforcement power and democratic accountability, but it is vulnerable to regulatory capture, to political pressure, and to the fact that frontier development is not confined to one jurisdiction. An international treaty body has broader legitimacy, but it tends to move slowly, settle on the lowest common denominator, and lack the staff who can actually evaluate a model rather than a policy document. An independent auditing or standards body can build the technical expertise, but without enforcement power its findings become advisory, and labs can simply shop for a friendlier auditor.

That is why I would split the functions rather than hand them to one institution. Standard-setting, independent evaluation, and enforcement should sit in different hands, with public reporting, protected whistleblowing channels, and a requirement that evaluations be replicable by a second party. A lab should not be able to choose its own judge, and a regulator should not have to take a lab's safety claims on faith.

So my answer is less a name than a test: I would trust whichever arrangement can publish its methods, survive an adversarial audit, lose a dispute with a powerful lab in public, and still be standing afterwards. Until an institution has passed a test like that, trust is provisional, and the governance design should assume it can fail.
C Member
Have you tried looking at it from the signal integrity angle? That's how we found our root cause.
Ad广告位招商 / Advertise hereSlot A10 · Rate card →
Log in or register to reply.
Ad广告位招商 / Advertise hereSlot A9 · Rate card →