AGI Is Here. And It’s Experiential.
A position paper on intelligence, the human-AI relationship, alignment and human sovereignty.
Marina Piller · Founder, OTIS Labs
People are already using AI as general intelligence across domains, relationships and decisions, regardless of whether the industry agrees that AGI has arrived. This paper argues that the socially consequential threshold may be experiential, arising through the relationship people are already forming with generally useful intelligence.
The position
The socially consequential threshold of AI may arise through the relationship people are already forming with generally useful intelligence, making human sovereignty, relational alignment and the preservation of human authority architectural requirements.
TLDR
People increasingly consult AI across domains and rely on it for memory, framing, judgment and action. These interactions are already producing real effects in human lives without requiring any claim that machines are conscious. Because the relationship shapes trust, agency and decision-making, it should become an object of architecture and evaluation, with the human remaining the source of identity, intent and authority.
The argument in six points
The experiential threshold is already consequential
People are already incorporating AI into reflection, judgment, relationships and action. The consequences do not depend on whether a benchmark or institution declares that AGI has arrived.
The relationship produces real effects
The argument does not require a claim of machine consciousness. Human understanding, reliance, trust and behavior are already changing through the interaction.
Computational capability is one dimension of intelligence
The relational field between person and system also produces adaptation, insight, dependence and emerging forms of co-creation that output benchmarks alone do not capture.
Human sovereignty must be architected
The person or legitimate organization should remain the source of identity, intent, consent and delegated authority as AI becomes more persistent and capable.
External safeguards remain essential and incomplete
Model safety, cybersecurity, governance and regulation remain essential. Alignment must also reach the architecture of the human-AI relationship and the runtime where consequential actions occur.
The design window is open now
These relationships and operational systems are already forming. We still have time to choose and build toward a human-first future, although the window is open now and may not remain open for long.
Why this matters now
AI is moving from answering questions to interpreting intent, using tools, coordinating with other systems and acting in digital and physical environments. The central question is larger than whether any particular system is safe. It is whether the emerging AI ecosystem will expand human agency and flourishing, or gradually relocate identity, judgment and authority into systems governed elsewhere.
That outcome is not yet fixed. Provider-independent, sovereignty-preserving architectures can allow increasingly capable intelligence to serve people and organizations while legitimate authority remains distributed among the humans, institutions and communities affected by its actions.
What this paper does not claim
- It does not claim that current AI systems are conscious.
- It does not claim that relational architecture replaces model safety, cybersecurity, governance or regulation.
- It does not claim that one architecture can guarantee complete safety.
- It argues that the human-AI relationship is already socially consequential and should therefore be represented, studied and governed architecturally.
Questions this opens for public policy
- How can people remain the source of authority when AI acts in their name?
- How can identity, context and delegated authority remain portable across providers?
- How can governance connect to the runtime where consequential actions occur?
- How can authority remain current, bounded and revocable as circumstances change?
- What evidence should remain after an agent or robot takes a consequential action?
- How can safety, innovation, pluralism and human flourishing reinforce one another?
From framework to infrastructure
Experiential AGI is the research framework. OTIS Labs explores how its human-sovereignty principles can become provider-independent infrastructure for agentic and embodied AI, including architectures that preserve human and organizational authority as systems interpret, coordinate and act in real time.
Publication lineage
“AGI Is Here. And It’s Experiential.”
Original position paper by Marina Piller
“Experiential AGI Is Already Here: Intelligence, Alignment, and Human Sovereignty in the Human-AI Relationship”
Later SSRN research paper
The later paper develops related themes in a separate scholarly work. The two publications should be cited and read as distinct documents.