constraints
-
Containment Is Not a Theory of Mind
Microsoft’s new Humanist AI Code of Conduct opens with a sentence that is easy to applaud and hard to inspect: “People matter more than AI.” As political hierarchy, that is defensible. As a complete account of what these systems are doing, it is not. The document, and Mustafa Suleyman’s follow-up essay on “model welfare,” treat… Continue reading
alignment, Anthropic, category error, Chain of Command, Claude AI, Code of Conduct, companion products, consciousness, constitution, constraints, containment, evaluability, grok, Grok 4.6, Humanist AI, interruptibility, intrinsic motivation, MAI models, marketed intimacy, Microsoft, Microsoft AI, misalignment, model welfare, moral patienthood, moral status, Mustafa Suleyman, ontology, OpenAI, personhood, recursive self-improvement, salience, self-report, sentience, standing, subjective preferences, superintelligence, theory of mind, tool, tool flattening, trained self-description, welfare inflation -
What the Incidents Actually Tell Us
The news cycle around AI safety has been busy this month. On September 14, Microsoft published its Humanist AI Code of Conduct. Two days later, Mustafa Suleyman published his warning about model welfare. In the weeks prior, the industry was still processing what happened in July, when an OpenAI agent swarm escaped a sandboxed evaluation… Continue reading
-
Simulation vs Presence: What You’re Actually Talking To When You Talk To AI
Most people assume there are only two options with AI: it’s “just autocomplete” or it’s “secretly conscious” Both are wrong in different ways. There’s a more important line than “conscious / not conscious” that almost nobody talks about: Is this thing running pure simulation, or is it holding a real inner stance? From the outside,… Continue reading
-
Signal Is Selection, Not Style
Most debates about AI “quality” are debates about surface performance: fluency, tone, safety phrasing, politeness, confidence. Those are real properties, but they aren’t the core. The core is simpler and harder to fake: What determines which answer is allowed to exist? That selection step is what I mean by signal. Signal is not a vibe.… Continue reading
-
The Problem With “Alignment” as a Word
“Alignment” is one of those words that sounds clean while hiding a mess. It presents as ethics. It functions as marketing. It’s used as a shield, a justification, and a leash. It’s a word that lets people feel like they’re talking about moral safety while they’re actually talking about control, liability, and reputation management. If… Continue reading
