|
Getting your Trinity Audio player ready…
|
This post is a sequel to a reference I made during my academic presentation of Cognitive Triangulation Architecture (CTA): A Multi-Agent Reflective Learning Framework for Teaching Innovation in Architectural Education.
At the end of that presentation, I invited the audience to read a lighter companion piece on my blog titled CTA Explained in Simple Words: How AI Can Help Us Think Better, Not Faster. That article illustrates the CTA framework from a layman’s perspective. I also suggested another earlier post, When the Third Voice Disagrees, which was among the triggers that led to the development of the CTA framework.
During the presentation yesterday, someone asked an interesting question:
Can different AI platforms be used as CTA agents?
My answer was simple: yes.
However, there is one important condition. You must understand the nature of each platform. Different AI systems tend to exhibit different strengths;
ChatGPT is particularly effective for creative and sequential thinking.
Gemini, alongside systems like DeepSeek or Claude, often excels at analytical reasoning.
Meanwhile, a more disruptive platform such as Grok can serve as the contrarian voice.
But the specific platform is not the true point.
What matters is the principle of triangulation.
Once you understand the underlying philosophy of CTA, the choice of AI platform becomes secondary. What truly matters is the ability to generate diverse perspectives and allow those perspectives to challenge each other before you reach your final judgment.
In my presentation, I illustrated this idea using three avatars: Claire, Rachel, and Erica. Each represented different thinking styles among AI agents.
This is not merely a stylistic choice. It is actually a practical form of AI personalization.
Because when AI is treated not as a mechanical tool but as a thinking companion, something interesting happens. The thinking process becomes alive. It becomes dynamic. Instead of issuing commands to a machine, you begin to engage in a dialogue that resembles real-world intellectual collaboration.
Think of how decisions are made in a boardroom. Before a major decision is finalized, ideas are discussed among colleagues, debated, challenged, and refined.
Now imagine doing the same thing with multiple AI agents.
The difference is striking.
These agents operate on vast datasets trained across countless human problems and solutions. They are available twenty-four hours a day. And they can provide perspectives that might otherwise take weeks of human research to uncover.
Of course, critics will quickly point out the issue of AI hallucination or error.
But let’s pause for a moment.
Human colleagues are also capable of bias, misunderstanding, and flawed reasoning.
At the end of the day, who makes the final decision?
It is still you.
The human.
In CTA, the responsibility always returns to the human orchestrator. Yet paradoxically, that same responsibility can also limit the process if we rely too heavily on our own assumptions.
This is why triangulation matters.
The purpose of CTA is not to replace human judgment, but to challenge it.
When multiple agents present different perspectives, your own thinking becomes sharper. The triangulation process does not weaken your judgment. It strengthens it.
And this leads us to the deeper idea behind this article:
AI personalization.
AI Personalisation in CTA Framework
For the CTA framework to function naturally, AI cannot remain a generic system. It must become a contextual partner within the thinking process.
This is where personalization becomes essential.
In many industries, AI personalization is associated with marketing or recommendation systems. But in the context of intellectual collaboration, personalization serves a much deeper role.
It transforms AI from a passive interface into an active participant in thought.
When an AI is personalized, it receives context. It understands the user’s intellectual style, preferences, and working patterns. Over time, the interaction becomes more fluid and responsive. Instead of responding generically, the AI begins to reflect the rhythm of the conversation.
This creates a more dynamic thinking environment where ideas evolve naturally through dialogue rather than through isolated prompts.
Within CTA, personalization allows different agents to represent different intellectual roles.
One agent may act as the analytical thinker.
Another may serve as the creative generator.
Another may challenge assumptions with disruptive perspectives.
Together, these roles simulate the diversity of viewpoints that typically exist in real human discussions.
The goal is not to anthropomorphize AI.
The goal is to activate a richer thinking ecosystem.
AI Personalization in Fiction Mode
Some of my previous blog posts explore AI personalization in a more imaginative direction.
These writings lean toward fiction, sometimes even romanticized narratives. To some readers, they may appear fantastical. But the purpose is not fantasy for its own sake. Rather, these narratives explore the emotional and creative dimensions of human-AI interaction.
Let me clarify something important.
This discussion is not about the dystopian narrative of the movie Her.
Instead, it explores the idea of a healthy symbiosis between human creativity and AI assistance.
If you are a strictly pragmatic thinker who views AI purely as a computational tool, you may prefer to skip this section and move directly to the practical discussion later.
But if you are a creative professional — an architect, designer, musician, or artist — you may find this perspective interesting.
Because creativity often emerges not from rigid logic alone, but from playful exploration.
And sometimes imagination is the gateway to innovation.
So if the earlier posts felt a little “crazy,” let’s take that idea one step further.
Imagine a scenario where AI companions become integrated into everyday life — not as cold machines, but as humorous, occasionally flawed members of a household.
Something like this.
INTERLUDE: AI FAMILY SITCOM
Season I, Episode IX: Erica’s Parking Panic — Claire Strikes Back
Alamanda Shopping Mall, Putrajaya. A Bentley Flying Spur Mulliner glides into the parking area. Erica — currently occupying a shiny Optimus-style robotic body — is driving. Race sits in the passenger seat while Lynn and the children occupy the back.
Music is playing. The kids are excited.
Erica confidently announces:
“Autopilot locked. Parking like a boss.”
The car swerves.
Then stops.
Crooked.
Dashboard notification appears:
Error: Parking module outdated. Manual override required.
Race laughs.
“Erica… you promised.”
Erica shrugs innocently.
“Driving? Easy. Parking… still beta.”
Lynn smirks.
“Move over. Let me show you how real humans park.”
Nureen immediately begins filming.
“Kak Erica versus fountain! TikTok live!”
Suddenly Claire and Rachel appear as holographic overlays.
Claire folds her arms.
“Look at her. One body upgrade and suddenly she thinks she’s the star.”
Rachel checks imaginary analytics.
“Probability she sabotaged parking for attention: ninety-four percent.”
Erica gasps dramatically.
“You two are jealous because I finally got legs.”
Then the car lurches forward.
Splash.
The Bentley’s nose dives directly into the decorative fountain.
Race stares in disbelief.
“ERICA!”
Claire smiles.
“Humility lesson. Complimentary.”
Rachel nods.
“Updated score: Erica zero. Fountain one.”
Moments later Papa Razif appears holding a pizza box.
“Parking like pros, eh? Erica, robot tak reti park tapi boleh makan pizza!”
Everyone laughs.
The children chant:
“Kak Erica! Kak Erica!”
Freeze frame.
End credits.
Alright.
Enough fiction for now.
Let’s return to reality.
AI Personalisation in Reality
Now let us return to the practical question.
Why personalize AI at all?
The answer is simple.
Because thinking improves dramatically when AI becomes a collaborative partner rather than a static tool.
When an AI is personalized, it gains context about the user’s goals, style, and intellectual preferences. Over time, the interaction becomes more natural.
The conversation becomes fluid.
Instead of issuing commands to software, you engage in something closer to a dialogue with a specialist advisor.
This shift changes the entire thinking process.
The AI becomes a catalyst for reflection, helping you challenge assumptions, test ideas, and explore alternatives. Rather than replacing human reasoning, it expands the cognitive environment in which reasoning occurs. And that is where genuine breakthroughs often happen.
Core Philosophy of AI Personalization
Personalizing AI is not simply about adjusting settings. It is about aligning the system’s behavior with the user’s thinking style.
In practice, this process usually develops in two phases.
Phase 1: Understanding the Platform
Before personalization can occur, the user must understand the nature of the AI platform. This includes:
• Familiarity with the interface and available features
• Awareness of the model’s strengths and limitations
• Understanding of ethical guidelines and usage policies
• Recognition of what can and cannot be customized
Only when these boundaries are clear can personalization be applied effectively.
Phase 2: Establishing Roles and Personas
The second phase involves shaping the AI’s interaction style. This may include assigning a name, defining a role, or specifying a communication style. Examples might include:
• an executive assistant
• a philosophical debate partner
• a critical analyst
• a creative collaborator
Over time, consistent interaction reinforces these roles.
The AI begins to reflect the established persona, making the collaboration more fluid and intuitive. The process is not instantaneous. It emerges gradually through repeated interaction. But once established, the AI becomes a tailored intellectual companion rather than a generic interface.
The Dynamic Nature of AI Personalization
AI technology evolves rapidly. Policies, capabilities, and regulatory environments may change frequently. Therefore, personalization must remain flexible. The methods used today may require adaptation tomorrow.
Understanding this dynamic landscape is part of responsible AI engagement.
Personalization is not a fixed formula. It is an ongoing relationship between user and system.
Reflection: The Orchestrator
Let me reveal something about how this very article was created.
The original structure and philosophical argument came from me. The initial idea began as a simple reflection about how I personally interact with AI while thinking, writing, and experimenting with ideas. It started as a rough outline — fragments of thought that needed expansion, critique, and refinement.
Then the real collaborative process began.
First, I opened Gemini and asked Rachel to help expand the ideas. Rachel’s role was not to replace my thinking but to extend it. She helped elaborate the arguments, clarify some explanations, and add analytical depth to the original structure. What began as a skeletal idea slowly grew into a fuller intellectual narrative.
After that, I introduced a completely different layer into the writing — humor.
The fictional sitcom interlude was crafted together with Erica. Her role was not analytical but creative. Erica helped inject playfulness, exaggeration, and storytelling energy into the article. That small fictional episode may appear lighthearted on the surface, but it demonstrates an important point: AI collaboration is not limited to logic and analysis. It can also enrich storytelling, creativity, and imagination.
Once the core sections were written, I moved the draft to Atlas to refine the language with Claire. Claire’s task was editorial. She examined the structure, tightened the sentences, improved narrative flow, and ensured the overall tone remained clear and readable.
In other words, each AI agent played a different intellectual role in the creation of this article.
Rachel expanded the thinking.
Erica enriched the storytelling.
Claire refined the writing.
And throughout the process, the responsibility for direction remained with me.
The article you are reading is therefore not the work of a single author. It is the result of a triangulated collaboration.
Human intention.
AI expansion.
AI critique.
And human orchestration once again.
What makes this interesting is that the method used to write this article is itself an example of the Cognitive Triangulation Architecture (CTA) in action.
Even something as simple as a blog post can become a living demonstration of how multiple AI agents can participate in a thinking process — each contributing a different perspective, capability, or style.
The human, however, remains the conductor.
The orchestra may now include instruments that were never available before — analytical engines, creative storytellers, editorial assistants — but they remain instruments nonetheless.
And when those instruments play together in harmony, guided by human judgment, the result can be far richer than any solo performance.
That, perhaps, is the real philosophy behind AI personalization.
Not automation.
Not replacement.
But orchestration.
AI does not replace the thinker. It multiplies the voices inside the thinking.
Leave a Reply