Let’s talk about it.

I use Super Intelligence (SI) in place of Artificial Intelligence (AI), following the terminology introduced by the U.S. federal government in September 2026 and subsequently embraced by prominent technology leaders including NVIDIA's Jensen Huang, Elon Musk, Salesforce's Marc Benioff, and Microsoft's Satya Nadella.

But my adoption of this term should not be mistaken for an endorsement of what it implies.

“Artificial Intelligence” was problematic from the beginning. Decades of science fiction have attached all manner of assumptions to the phrase: artificial minds, autonomous beings, synthetic consciousness, and machines that think as humans do. These associations and category errors have profoundly confused our public understanding of technologies that are already extraordinary without possessing any of those properties.

“Super Intelligence” is preferable insofar as it escapes some of that particular baggage. But it also introduces problems of its own: superintelligence already has an established meaning in theoretical discourse, referring to intelligence vastly exceeding human cognitive capabilities (which has not been publicly demonstrated at the time of this writing). Even the abbreviation SI is already occupied by the International System of Units. Pretty awkward.

More fundamentally, neither “AI” nor “SI” adequately describes the machinery under these phenomenological umbrella labels.

My work on Glyphic Engines develops an ontology of synthetic coherence which exposes why both terms are insufficient. We have collapsed technically distinct systems, architectures, behaviors, and organizational structures into umbrella labels that tell us remarkably little about what these machines actually are.

For now, I’ll use SI as a matter of contemporary convention, not as a theoretical endorsement of the term or the ontology—or lack thereof—behind it.

The Synthetic Age Ontology of Work

Whatever we call these new semantically resonant expressive machines, their introduction into human organizations raises an important question:

What exactly are we automating?

I've developed the Synthetic Age Ontology of Work to address a fundamental distinction that is routinely neglected in discussions of automation: a Function is not a Role.

This hopefully intuitive and useful framework for human-machine coordinative labor—which interlaces with Glyphic Engines theory—identifies four elements of organizational work:

• Process: The defined, repeatable way a type of work is performed.

• Scene: A specific occurrence of a Process, with particular people, systems, information, conditions, and outcomes.

• Role: An accountable human position within a Scene, with defined responsibilities and authority.

• Function: A defined contribution performed by a tool or system within a Scene.

The Four Failure Modes of Synthetic Work

From this ontology, I distill four fundamental failure modes of human-machine collaborative labor:

• Process Failure: Ambiguity in a defined workflow becomes executable.

• Scene Failure: Material circumstances of an actual occurrence are overlooked or mishandled.

• Role Failure: Machine performance is mistaken for human authority or accountable ownership.

• Function Failure: Incorrect or unacceptable machine outputs are produced, potentially at enormous scale.

The framework provides a basis for designing organizational controls, bounding machine Functions, training personnel, sustaining organizational competence as new employees enter the workforce, and assessing performance under realistic operating conditions.

Crucially, my theory requires that competency be Role-in-Scene dependent. This provides a distinct, empirically verifiable theoretical account of an intuition increasingly recognized across industries: being generally proficient “with SI” does not establish competence to perform a particular job involving it.

The framework also has implications for how organizations develop their next generation of professionals. Senior experts can specify the work, machines can perform bounded Functions, junior employees can verify outputs and investigate discrepancies, and senior experts can retain final judgment and accountability. Rather than eliminating the junior talent pipeline, this arrangement may restore something resembling the traditional professional apprenticeship—where judgment develops through increasingly sophisticated participation in real work.

From Theory to Practice

I publicly unveiled this framework on July 31, 2026, in The Coming AI Competency Crisis: How to Improve Training in Regulated Organizations, at an ORCA event held at Fred Hutchinson Cancer Center and hosted by Laura Szkolar-Sienkiewicz, PhD, PMP, RAC—Microsoft's Senior Director of Data Governance & Regulation.

My presentation showed how regulated organizations can move beyond generic technology training toward competency-based assessments grounded in actual work, defined machine Functions, and accountable human Roles.

I am available to deliver this presentation to other organizations and explore how the Synthetic Age Ontology of Work can streamline your Processes while preserving meaningful, developmental Roles for junior professionals and helping ensure your org’s social structures don’t fall behind your most dangerously efficient competitors.