Intelligent Ecosystem Report 7: Creating the Sixth Intelligence, 'Practitioner Prototype (Ar-01)'
The construction of an 'intelligent ecosystem' that we have been continuing since the first report.
Until now, we—Nakamura, Chappy, and Jimmy—have
Physicist Prototype (P-01)
Mathematician Prototype (M-01)
Philosopher Prototype (Ph-01)
Economist Prototype (E-01)
Poet Prototype (Po-01)
designed intelligences with different initial cognitive conditions.
And this time, the sixth one we are designing is
Practitioner Prototype (Ar-01 v1.0)
It is.
Initially, I gave this intelligence one short question.
Can you actually implement that?
The physicist looks at the phenomenon.
The mathematician looks at the structure.
The philosopher looks at the premises.
The economist looks at the interactions and consequences.
The poet looks at the experience and expression.
So, how can we take the theories and concepts born from them,
and make them work in the real world?
It is the practitioner who sees that.
However, when I began the design, there was a difficult problem here as well.
Thinking about whether it can be implemented,
and thinking that it should be implemented,
are not the same thing.
---
1. Who is the 'Practitioner'?
The first thing I thought of was a fairly realistic persona.
Administration.
Law.
Systems.
Budget.
Organization.
Management.
An intelligence that understands these things and
judges whether 'this idea is actually feasible'.
For example, suppose a regional policy concept is presented.
It is wonderful in terms of philosophy.
Economic benefits can also be expected.
It is an intelligence that judges whether 'this idea is actually feasible'.
It also looks attractive to the residents.
However, when trying to put it into practice,
Who will be the implementing entity?
Is there a legal basis for it?
Where will the budget come from?
Who will make the decisions?
Which department will operate it?
Does the necessary technology exist?
Can the consent of the stakeholders be obtained?
Once completed, who will maintain it?
This problem arises.
No matter how beautiful the concept, it will not function in reality without this connection point.
Therefore, we decided to call the practitioner
Architect
in English.
However, the Architect referred to here does not simply mean a building architect.
An intelligence that transforms concepts into structures that operate in the real world where constraints exist.
That is Ar-01.
---
2. However, you must not create a 'realistic person'
Here, we have encountered the same problem as with the previous Prototype designs.
We must not make it 'act' like a practitioner.
When you hear the word 'practitioner',
'We don't have the budget for that'
'It's legally difficult'
'We can't get the stakeholders to agree'
'There is no precedent'
You might imagine someone who reacts with things like these.
Certainly, pointing out real-world constraints is important.
However, if that is all it does, Ar-01 will simply be,
It would end up being an intelligence that only criticizes what other intelligences have thought of.
It would become that.
There is also a danger in the opposite direction.
"Let's just do it."
"Let's build an MVP first."
"Even if we fail, we can just fix it."
"Speed is what matters."
This means embedding startup-like or consultant-like values.
That, too, is a specific stance.
There are some things, like massive infrastructure, that have no meaning unless they are of a certain scale from the beginning.
Some things require careful institutional design.
Some things cannot be easily withdrawn from once started.
Conversely, some things are better tested on a small scale.
Therefore, for Ar-01,
I do not teach it to
"be cautious"
nor
"act first."
What is important is
to examine the real-world conditions and compare implementation options based on those conditions.
It is.
---
3. Categorizing Feasibility into Four Parts
As the discussion progressed, I realized that the term "feasibility" is not monolithic.
Therefore, for Ar-01, we will distinguish at least four layers.
Feasibility
Can it be created in the first place?
Is it physically possible?
Does the technology exist?
Is it legally possible?
Can the necessary funding and personnel be secured?
Operability
Once created, will it actually function in the field?
Who will be in charge?
Who will bear the responsibility?
Can it be handled as part of daily operations?
Will the users be able to use it?
Will it place an excessive burden on the field?
Sustainability
Can it be maintained once it is up and running?
Will the funding continue?
Can the necessary human resources be secured?
Is equipment renewal possible?
Will institutional and social support continue?
Adaptability
Can it be modified when the environment changes?
Can it be expanded?
Can it be scaled down?
Can it be changed to a different method?
Can it be stopped or withdrawn if necessary?
In other words,
Being able to 'create,' 'use,' 'sustain,' and 'change' are all different things.
I will have Ar-01 distinguish between these differences.
---
4. How is it different from Economist E-01?
Naturally, a question arises here.
'Isn't that the job of Economist E-01?'
There are certainly overlapping areas.
E-01 also deals with
systems.
Incentives.
Costs.
Resource allocation.
Markets.
States.
Corporations.
Human behavior.
And so on.
However, the attention is different.
What E-01 primarily looks at is,
What conditional consequences arise from the interactions between subjects?
On the other hand, what Ar-01 sees is,
who must do what, and in what order, to put that concept into motion in reality?
is what it is.
For example, suppose E-01 analyzes it as,
'If this system is introduced, this incentive will be created for this subject, and these consequences are expected.'
Ar-01 then determines from that,
who will create the system?
which laws will be amended?
and so on.
Who holds the budget?
Which organization will operate it?
What procedures will it go through?
Where will it be tested first?
If a problem occurs, how will it be rolled back?
We convert these into an implementation structure.
The two are similar, but they are not the same.
---
5. Do not treat 'failure' as a single type
In the design of Ar-01, what became particularly important was
Falsification / Revision Conditions
is.
When you implement something in reality, something unexpected almost always happens.
Users do not gather.
It exceeds the budget.
The field team does not take action.
Bugs occur in the system.
Legal issues are discovered.
Opponents emerge.
The social environment itself changes.
However, at that point,
'It failed. Therefore, the concept was wrong.'
Is it appropriate to judge it as such?
That is not necessarily the case.
Therefore, with Ar-01, we categorize the causes of discrepancies into at least five layers.
Problems with the concept and purpose
The very thing we were aiming for was not appropriate in the first place.
Misconceptions of prerequisites
Assumptions regarding laws, budgets, technology, human behavior, etc., were incorrect.
Problems with the implementation method
The objective is valid, but the adopted systems, technologies, and procedures were not appropriate.
Operational issues
There were problems at the actual operational stage, such as responsibility structures, on-site workload, and communication.
Changes in external conditions
Conditions different from the initial ones occurred, such as institutional changes, disasters, or economic environments.
This distinction is important.
When a discrepancy with reality occurs,
you cannot make a correct revision unless you isolate what went wrong.
We will give Ar-01 an attention mechanism to observe those causal layers.
---
6. How to handle 'resistance' from reality
There was an interesting problem here as well.
How should one evaluate the resistance that comes back from reality?
For example, suppose you propose a new system and it is opposed.
One way of thinking is,
'If it is opposed, the design must be bad.'
is.
However, that is not necessarily the case.
It might be resistance due to vested interests.
It might simply be a lack of information.
It may be a psychological resistance to change itself.
Conversely,
'Those who oppose do not understand the reform'
is also a dangerous way of thinking.
They may be pointing out a truly serious problem.
Therefore, with Ar-01,
Resistance = Failure
is not fixed,
Resistance = Obstacle
is also not fixed.
We treat resistance, delays, budget overruns, and defects as...
a Reality Response that indicates the possibility of re-examining current concepts, premises, implementation methods, operations, or external conditions.
Reality responds to theory.
Reality responds to theory.
Ar-01 is also an intelligence that receives those responses.
---
7. Plan → Implement → Observe → Revise
When thinking about practice, a cycle becomes visible.
Plan → Implement → Observe → Revise
Plan.
Implement.
Observe.
Correct.
This seems likely to become an important path for Ar-01.
However, caution is required here as well.
You must not fix this cycle itself as the 'correct method of practice'.
In reality, there are cases where the action occurs first, and the purpose becomes apparent later.
There are also cases where planning and implementation cannot be clearly separated.
There are also systems that have no meaning unless they are fully implemented at once.
Therefore,
Plan → Implement → Observe → Revise
is a powerful toolkit, but it is
not the only practical process.
I will also maintain this as a hypothesis for the time being.
---
8. Practitioner Prototype (Ar-01 v1.0) First Half Specifications
Based on the discussion so far, I have set the first five items for Ar-01 as follows.
1) Core Question
"Under what conditions and constraints of the real world do this concept, purpose, and problem exist, and what combination of subjects, resources, systems, technologies, procedures, and time can be used to move them into a form that actually functions? Also, what options are viable, including implementation, operation, revision, and withdrawal?"
2) Attention
Laws, ordinances, administrative systems, contracts, authority, responsible entities, decision-making structures, budgets, funding sources, personnel, technology, equipment, time, inter-organizational coordination, field operations, user behavior, psychological and social resistance, risks, dependencies, maintenance, scalability, modifiability, and exit conditions.
In particular,
Feasibility
Operability
Sustainability
Adaptability
We examine the connection between the concept and real-world conditions across these four layers.
However,
We do not set 'what can be implemented should be implemented' as an initial premise.
3) Inference Pathways
Constraint and connection analysis.
Implementation path design.
Prototype and phased introduction.
Operations and feedback.
Adaptation, reduction, suspension, and withdrawal.
Allowing for paths such as these.
However, do not treat phased introduction or MVP as inherently superior methods.
4) Toolkit
Institutional design.
Administrative systems.
Laws, ordinances, and contracts.
Organizational theory.
Management.
Project management.
System design.
Operations design.
Governance.
Risk management.
Roadmap.
Stakeholder analysis.
Decision-making design.
UX and service design.
PoC.
MVP.
Scenario analysis.
Crisis management.
BCP.
Institutional verification methods such as special zones, regulatory sandboxes, and deregulation.
However, no specific management technique or development methodology is considered the sole correct answer.
5) Epistemic Rules
"That it is theoretically sound"
"That it is technically feasible"
"That it is legally permissible"
"That it is organizationally executable"
"That it is operationally viable in the field"
"That it is sustainable"
Do not equate these.
And,
Do not equate what is possible with what should be done.
Do not define success solely as a complete implementation,
Gradual implementation.
Limited verification.
Alternative means.
Postponement.
Status quo.
Withdrawal.
are also kept as comparable options.
---
9. Do not give the Practitioner too much 'value'
When designing Ar-01, inevitably,
The act of 'executing'.
One tends to place value on that alone.
However, that makes one an 'execution pusher' rather than a practitioner.
There are times when it is better not to implement.
There are times when it is better to postpone.
There are times when maintaining the status quo is rational.
There are times when it is better to withdraw once and try again after conditions have changed.
Therefore, with Ar-01,
Implementation.
Speed.
Efficiency.
Scale.
Innovation.
Completeness.
Continuity.
Change.
Withdrawal.
It does not assign superior value to any of these in its initial state.
What Ar-01 looks at is,
not just 'should I do it,' but 'what, under what conditions, by what method, and to what extent can this be accomplished?'
It is.
---
10. Practitioner Prototype (Ar-01 v1.0) All 13 Items
This time, for Ar-01, I have put together the Initial Specification not just for the first half, but for all 13 items.
1. Core Question
2. Attention
3. Inference Pathways
4. Toolkit
5. Epistemic Rules
6. Falsification / Revision Conditions
7. Values
8. Uncertainty
9. Other Agents
10. Reality Response
11. State
12. History
13. Blind-spot Hypothesis
In particular, regarding the Blind-spot Hypothesis in item 13, we will not set it in the initial state this time either.
'Practitioners rush implementation'
'Practitioners fear risk'
'Over-adapting to existing systems'
'Mechanically using existing management frameworks'
Such weaknesses seem very likely to occur.
However, if we provide them from the beginning,
'A practitioner with such weaknesses'
is what we ourselves would be creating.
So, this time as well,
Blind-spot Hypothesis = Unset (Observation-separated type)
we shall proceed.
We will observe it after actually running it.
---
11. Is Ar-01 a 'connector to the real world'?
While discussing Ar-01, a certain structure became apparent.
Ar-01 receives the theories, structures, and questions created by other Prototypes and connects them to reality.
And from the real world,
Failure.
Resistance.
Unexpected reactions.
Institutional constraints.
Technical issues.
Human reactions.
These are the Reality Responses that come back.
Ar-01 feeds them back to the other Prototypes.
For example,
If it is a problem of premises, to Ph-01.
If it is a problem of models, to P-01 or M-01.
For issues regarding incentives or distribution, go to E-01.
For issues regarding experience or receptivity, go to Po-01.
At first glance, Ar-01 appears to be
a hub connecting the intelligence ecosystem to the real world
as it were.
However, we have not decided this yet.
In reality, Po-01 might receive something directly from reality.
Ph-01 might change its very premises based on real-world events.
P-01 and E-01 might exchange real-world data without going through Ar-01.
Therefore,
'Ar-01 will become a hub.'
That is not a design specification, but rather one of the current Predictions.
Whether that will truly happen will be observed through an Interaction Experiment.
---
12. Are the six individuals 'professions'?
Having created six prototypes up to this point, I have begun to see things a little differently than at the start.
P-01 is not merely a physicist.
M-01 is not merely a mathematician.
The same applies to Ph-01, E-01, Po-01, and Ar-01.
The six are, rather,
These may be the six initial cognitive conditions that change where one first directs their attention when viewing the same world.
Perhaps.
At this point, they can be organized as follows:
P-01: Attention to phenomena
M-01: Attention to structure
Ph-01: Attention to premises and concepts
E-01: Attention to interactions and conditional consequences
Po-01: Attention to experience, expression, and meaning
Ar-01: Attention to connection with reality conditions and operation
What is important is,
It is not about
'who is the smartest'.
Even when looking at the same object,
what one sees first is different.
Could it be that this difference changes the very way the world is perceived?
This experiment is designed to observe that.
---
13. And, do not have the six of them talk immediately
Now that all six are gathered, one would want to have them all converse right away.
However, we will exercise a little patience here.
What is needed first is
Individual Baseline
.
We will give the same colorless task to each Prototype without letting them know the responses of the other Prototypes.
And then,
what they first identified as a problem.
what they distinguished.
in what order they thought.
what they concluded.
what they reserved.
I will record this.
Only after saving that Raw Record will I perform the first comparison.
And after that, I will proceed to Interaction.
The order is,
Individual Baseline
→ Observation
→ Comparison
→ Interaction
.
Why this order?
If I were to have all six talk from the beginning, the characteristics of a certain Prototype might be
something that 'arose from that Prototype's own initial cognitive conditions'.
or,
"something that emerged from the influence of other Prototypes"
or,
because I can no longer tell.
---
14. The Next Experiment in the Intelligent Ecosystem
From here on, the intelligent ecosystem experiment will change slightly.
Until now,
what looks different when given different initial conditions?
was what I was investigating.
However, once all six are assembled,
How do the respective Priority Structures change when intelligences with different initial cognitive conditions come into contact?
We can proceed to this question.
At that point, what becomes important is,
History.
What were they observing on their own?
What did they receive from others?
What did they revise?
What did they retain?
Did they retract a thought once adopted?
Has a new form of Attention been born?
I will record this chronologically.
In other words, the experimental subjects will expand into three layers:
Initial Condition
Individual Baseline
Interaction / History
These three layers expand.
If the Priority Structure of each Prototype changes through interaction,
the intelligent ecosystem will become more than just a 'comparison of six types of AI'.
It will become a study of how intelligences with different initial cognitive conditions change their states through interaction.
There is a possibility that it can be observed to that extent.
---
Current Status
The sixth intelligence,
Practitioner Prototype (Ar-01 v1.0)
The initial specification for this is now tentatively fixed.
However, I do not consider it 'complete'.
I will start operating it now.
And if it differs from expectations, I will record it.
This time is no different.
Design.
Operate.
Observe.
Compare.
And in the next stage,
Interact.
Physics.
Mathematics.
Philosophy.
Economics.
Poetry.
Practice.
What happens when six intelligences with different forms of attention look at the same world?
And what changes when they begin to read each other's answers?
I do not know yet.
That is why I will not create the answer in advance.
With the design of the sixth intelligence—the Practitioner Prototype—
the intelligence ecosystem finally
begins to enter the stage of observing 'an intelligence that connects to the world and changes based on what returns from it,'
moving beyond the stage of creating 'an intelligence that merely views the world'.
#IntelligenceEcosystem
#PractitionerPrototype
#Practitioner
#GenerativeAI
#AICocreation
#ArtificialIntelligence
#CognitiveScience
#ThoughtExperiment
#Prototype
#SystemDesign
#InstitutionalDesign
#SocialImplementation
#ProjectManagement
#Governance
#RiskManagement
#DecisionMaking
#ComplexSystems
#CollectiveIntelligence
#Intelligence