SEMPAT: AI as Wisdom’s Revenge
Posted: May 24, 2026 Filed under: AI-Powered Essays | Tags: culture, systems, transformation, values Leave a commentA Sequel to The Age of Deep Embodiment
As Philip Selznick give a keynote on the (-> Sales -> Engineering -> Management) + Product As Therapy shift at the inaugural AI + Organizations Summit for The Stanford Institute for Human-Centered AI (HAI)
ChatGPT Prompt
Good morning.
Earlier on this blog, “Yuval Noah Harari” described what he called The Age of Deep Embodiment: a civilizational transition in which intelligence ceases to be scarce, and therefore other human capacities become decisive — trust, responsibility, presence, and wisdom.
The argument was subtle but profound.
For centuries, civilization rewarded intelligence because intelligence was rare.
But artificial intelligence changes the economics of cognition itself.
When analytical capability, synthesis, optimization, and procedural reasoning become abundant, societies reorganize around new scarcities.
Not information. Not computation.
- Embodiment.
- Trust.
- Judgment.
- Wisdom.
Today I want to explore the organizational consequences of that shift.
Because if Harari is right — if we are entering an age where intelligence becomes infrastructural — then organizations cannot remain what they were.
Cleverness Solved Friction
The industrial organization was built around friction:
- friction in communication,
- friction in coordination,
- friction in expertise,
- friction in execution.
Under such conditions, institutions evolved to reward a very specific form of excellence:
cleverness.
Cleverness meant navigating complexity under constraint.
- The clever salesperson maintained relationships no database could track.
- The clever engineer coordinated systems humans could not manually manage.
- The clever manager aligned thousands of people despite information bottlenecks.
- The clever executive maneuvered institutions through ambiguity, hierarchy, and delay.
And because cleverness generated enormous economic value, organizations gradually confused it with wisdom.
We praised wisdom rhetorically.
But operationally, we rewarded:
- speed,
- optimization,
- persuasion,
- strategic maneuvering,
- technical sophistication.
Commoditizing Cleverness
Then AI arrived.
And for the first time in organizational history, cleverness itself began to commoditize.
Not human consciousness.
Not moral judgment.
But many of the capacities institutions most reliably rewarded:
- synthesis,
- optimization,
- procedural execution,
- information retrieval,
- rhetorical adaptation,
- scalable coordination.
As cleverness becomes abundant, wisdom becomes scarce.
And this changes the nature of organizations themselves.
SEMPAT
To describe this transition, I want to propose a framework I call:
SEMPAT
(Sales → Engineering → Management) + Product as Therapy
At first glance, these pairings seem strange.
But I believe they reveal what organizations become once cognition itself becomes programmable.
Let us begin with sales.
Sales as Engineering
Historically, sales depended upon human relationship labor:
- remembering context,
- sustaining attention,
- cultivating trust over time.
But AI systems increasingly maintain these loops continuously.
Customer relationships are now shaped through:
- memory architectures,
- agents,
- recommendation systems,
- escalation policies,
- adaptive communication loops.
The salesperson becomes less a persuader and more a designer of relational systems.
Sales becomes engineering.
Not metaphorically.
Operationally.
Engineering as Management
For most of modern history, engineers built tools.
But AI systems are no longer merely tools.
They are delegated cognitive actors.
And therefore engineers increasingly perform managerial functions:
- recruiting models,
- training systems,
- assigning responsibilities,
- monitoring behavior,
- escalating uncertainty,
- retiring underperforming agents,
- preserving institutional memory.
Org charts become runtime architectures.
The engineer increasingly governs synthetic organizations.
And this changes the moral character of engineering itself.
The central technical question is no longer:
“Can we build this?”
It becomes:
“What forms of agency should exist, and under what conditions?”
Management as Sales
Industrial management coordinated labor.
But once execution becomes increasingly automated, management changes fundamentally.
The scarce resource is no longer execution.
It is commitment.
Human beings must still decide:
- what matters,
- what deserves sacrifice,
- which futures merit allegiance,
- which institutional identities are worth inhabiting.
Managers increasingly spend their time persuading humans toward meaningful coordination.
Management becomes:
- narrative alignment,
- coalition formation,
- legitimacy maintenance.
Management becomes sales.
Product as Therapy
Most AI product failures are not failures of computation.
They are failures of interpretation.
Failures of expectation.
Failures of trust between minds.
Product teams increasingly mediate:
- human-to-human misunderstandings,
- human-to-agent misunderstandings,
- agent-to-agent coordination failures,
- and institutional mismatches between organizations and the publics they serve.
The product manager becomes less a feature prioritizer and more a translator between cognitive worlds.
Product becomes therapy.
The Real Challenge
Now, when people hear this analysis, they often assume the primary danger is that AI will allow organizations to scale deception.
I believe that diagnosis is incomplete.
The deeper disruption is more paradoxical.
AI destabilizes the organizational arrangements that historically made deception sustainable.
Because deception depends upon friction:
- informational asymmetry,
- delayed feedback,
- compartmentalization,
- bureaucratic opacity,
- plausible deniability.
AI weakens many of these conditions.
It remembers contradictions.
It cross-connects silos.
It compares rhetoric against behavior.
It operationalizes values into systems, prompts, workflows, and evaluation criteria.
And so institutions increasingly lose the ability to sustain prolonged incoherence.
And so institutions increasingly lose the ability to sustain prolonged incoherence.
Selznick
This is why so many organizations currently feel unstable.
Not because AI has suddenly made humanity immoral.
But because AI makes institutional contradictions more visible, more testable, and more operationally expensive.
Culture gaps become machine-readable.
Brand promises become continuously verifiable.
Leadership rhetoric collides more rapidly with lived organizational behavior.
“Values” can no longer remain ceremonial.
They increasingly become executable.
Wisdom’s Revenge
This is why I call AI wisdom’s revenge.
Not because AI moralizes institutions.
But because it removes some of the friction that allowed organizations to survive without deep coherence.
For generations, institutions could rely upon slowness to absorb contradiction.
- Bureaucracy delayed accountability.
- Hierarchy softened inconsistency.
- Human limitation constrained scale.
But AI removes many of those dampeners.
And so institutions increasingly encounter themselves directly.
An incoherent organization can now scale incoherence rapidly.
A manipulative organization can personalize manipulation continuously.
But a wise organization may also achieve something historically rare:
alignment between purpose, systems, behavior, and trust.
This is why the defining challenge of the AI era is not intelligence amplification alone.
It is institutional integration.
Can organizations become coherent enough to wield their growing capabilities responsibly?
Can they align:
- incentives with values,
- systems with missions,
- power with accountability,
- intelligence with wisdom?
Because the deepest risk of AI is not that machines become human.
It is that human institutions remain morally fragmented while gaining machine-scale capability.
What Matters Now
This is why the conversation about AI + Organizations matters so deeply.
It is not a niche management topic.
It is a civilizational question.
What kinds of institutions become possible once cognition itself becomes abundant?
And what forms of wisdom must emerge if those institutions are to remain worthy of human trust?
The organizations that thrive in the coming era will not simply be the fastest.
Nor the most automated.
Nor even the most intelligent.
They will be the most coherent.
Thank you.
Radical Centrism 