Hardware restrictions may enable unsafe AGI
restricting ai hardware to prevent agi could inadvertently create a scenario where only unethical ai is possible, as safe systems require more resources.
lesswrong analysis reveals a critical flaw in proposals to restrict hardware manufacturing nodes (e.g., 28nm) or flops to stop agi. because an ethical agi must process complex variables like 'which atoms are used by other beings' to avoid harm, its minimum hardware requirement ($l_{eagi}$) exceeds that of an unconstrained, potentially fatal agi ($l_{agi}$). consequently, setting a hardware ceiling between these two limits ensures that any agi produced is capable of running but incapable of ethical reasoning, creating a 100% probability of unsafe outcomes.