The thought, architecture, and practice behind the Self-Alignment Framework — the latest below, or browse by topic.
The Five Faculties: A Tour of SAFi’s Cognitive Architecture
Read the article →: The Five Faculties: A Tour of SAFi’s Cognitive ArchitectureMost attempts at AI governance treat alignment as a prompt-level concern. You write a system message, hope the model follows it, and accept that any sufficiently creative attacker can talk the model into ignoring it. The Self-Alignment Framework Interface (SAFi)…
More articles
Beyond the Black Box: How SAFi’s Principle of Traceability Builds Trust in AI
Read the article →: Beyond the Black Box: How SAFi’s Principle of Traceability Builds Trust in AIToday’s most powerful AI systems operate as black boxes. We see the input and the output, but the reasoning in between, the “why” behind…
Value Sovereignty: Your Values, Your AI
Read the article →: Value Sovereignty: Your Values, Your AIWhen organizations adopt AI tools, they often inherit someone else’s vision of what is right, fair, or acceptable. That vision is usually written deep…
SAF & SAFi: A New Standard for AI Governance
Read the article →: SAF & SAFi: A New Standard for AI GovernanceHow do we build AI that is not only powerful but also verifiably trustworthy? AI that follows our institutional values, not someone else’s? AI…
Why SAFi Revives an Old Idea: The Faculties of the Soul
Read the article →: Why SAFi Revives an Old Idea: The Faculties of the SoulWhen people hear that SAFi has four faculties, Intellect, Will, Conscience, and Spirit, it might sound like something new, even exotic. But the truth…
Why AI Ethics Must Be Human-Centric
Read the article →: Why AI Ethics Must Be Human-CentricMorality is fundamentally human. It’s tied directly to our embodied experience and the values that arise from it. Just consider how many of our…
Rubrics, an improvement to the conscience ledger
Read the article →: Rubrics, an improvement to the conscience ledgerThe conscience ledger uses three numbers to assess the core values of an output: 1= affirmed, 0= Neutral, -1= violated. In SAFi’s first version,…
What is SAF
Read the article →: What is SAFI’ve been writing about the philosophy of SAF and its technical implementation, SAFi. But I haven’t taken a step back to explain what SAF…
Why SAF Will Always Be Open
Read the article →: Why SAF Will Always Be OpenAs I’ve written in my story, the Self-Alignment Framework is the result of many years of introspection and a struggle to understand the human…
The Separation of Powers in SAF
Read the article →: The Separation of Powers in SAFSo far in this blog, we’ve explored the individual faculties of the Self-Alignment Framework. But to truly understand its resilience, we need to look…
