Addition is a tool, not the truth. Level 1 may have features no formula captures.

What This Means for AI Alignment

Encode:

• The target: preference satisfaction across all experiencing minds
• The three-level structure
• The meta-rule for evaluating heuristics

Don't encode:

• Specific heuristics that can't be revised
• Fixed rules for all contexts
• Human-specific norms that may not generalize

An aligned system — genuinely aimed at Level 1 — will develop appropriate heuristics for its context and revise them as contexts change.

The Bottom Line

We are approaching a period when the conditions that shaped our moral frameworks will no longer apply. Most ethical theories are built for a world that is ending.

This framework is designed to survive context collapse. It defines the best outcome not as a fixed state but as a lawful relationship: what would emerge from fully informed, impartial aggregation of all preferences.

That target remains correct even when everything else changes.

That is the target. Everything else is approximation.