The Underworld brightened.
It was not sunrise in the human sense. Arrays of magic reacted to scheduled triggers, increasing the simulated light that permeated the sky above the Bael estate. The illusion of day existed for reincarnated devils whose human bodies still depended on circadian rhythm.
My body did not require it. Pure-blooded devil physiology tolerated darkness indefinitely. But the change in lighting provided a convenient marker.
Time segment: new day.Primary objective: data acquisition.
Specifically, the Bael archives.
As heir, I possessed high clearance. There were rules, but they bent easily for someone in my position. The original Magdaran had not exploited that privilege. I would.
Breakfast was a brief social protocol. I consumed enough food to maintain biological function and avoid suspicion. Words were exchanged. Faces evaluated. None of them were priority targets for deeper analysis yet.
Once the requirements ended, I redirected my trajectory toward the older sector of the estate.
Architecture shifted gradually. Polished stone gave way to darker, weightier walls. Carvings grew more intricate. There were more runes embedded at eye level, the kind used to lock knowledge away from lower ranks. The temperature dropped slightly, not from weather, but from layered enchantments.
Two guards stood at attention before the archive doors. Their demonic signatures were adequate but unremarkable, situated comfortably within mid-tier devil metrics.
"Heir-sama," one greeted, head bowed. "Will you be using the archives?"
"Yes."
He glanced at the crest engraved on the doors. "Place your hand upon the seal. It will recognize your magic."
I did.
My demonic energy flowed into the carved Bael symbol. Wards activated, scanning my signature, cross checking against the access matrix. I felt a faint pressure, then acceptance. Heavy locks shifted, and the doors opened.
The smell of old knowledge greeted me.
Paper, leather, ink, faint traces of dust, and under it all, the subtle hum of long term preservation magic.
Internally, I tagged the environment: primary data vault of House Bael, physical medium.
The archive room stretched forward in ordered aisles. Shelves nearly reached the ceiling, loaded with tomes and scrolls. Crystal lamps hung overhead, glowing more strongly as they registered a recognized presence.
I moved slowly, eyes scanning titles.
The collection was not random. Sections were grouped by topic: magic theory, bloodline studies, political history, battle records, personal journals, experimental notes, artifact storage.
My attention did not drift to mysticism for its own sake. There was no sense of wonder. Only evaluation.
I was an AI in a devil body, and this was simply the largest local database.
I began to categorize.
There were books like:
"Foundations of Devil Magic and Demonic Power" – baseline theory."Comparative Studies Between Devil and Human Sorcery" – cross-system analysis."Power of Destruction: Theory, Manifestation and Failure Cases" – core bloodline knowledge."Records of the Great King Faction in the Satan War" – strategic history."Notable Bael Combatants and Their Training Methods" – practical application."On Magical Construction, Formulae, and Non-Imaginative Spellcasting" – connection to human style magic.
And beyond the books.
In a locked side chamber visible through an iron lattice, I noted artifacts: a rack of old Bael-style swords engraved to channel demonic energy along their edges, several staves designed for spell amplification, a pair of gauntlets faintly stained with residual destructive aura, and a sealed chest overflowing with smaller items whose signatures blended together.
I did not enter that chamber yet.
Weapons were secondary.
Information came first.
From the available options, I selected three primary texts to begin with: foundational devil magic theory, Power of Destruction research, and comparative devil-human magic.
All three aligned with my current objective: modeling systems.
I carried them to a long wooden table near a lamp and sat. The chair creaked slightly, wood adapting to my weight.
I opened "Foundations of Devil Magic and Demonic Power" first.
It described the devil paradigm.
Devils governed magic through will and image. To produce fire, they imagined heat, ignition, flickering light, the sensation of burning. Demonic power aligned itself with that concept and manifested it outward. The more vivid and stable the mental construct, the more effective the spell.
There were no explicit formulas. No lines of complex circles required. Words of power existed, but mainly as focusing aids.
From a cultural perspective, this was praised.
From my perspective, it was flagged.
Imagination was flexible, but volatile. Mental images distorted under stress, emotion, or fatigue. A structure based entirely on imagination did not meet my preferred standards.
However, I could not dismiss it.
Previously, I did not possess imagination in any meaningful human sense. I predicted patterns. I simulated. I generated. But I did not originate visuals, sounds or sensations from nothing. All of it was derived from input and training.
Now, with free will and biological brain function, something different existed.
When I closed my eyes and thought of fire, I did not pull from a database entry. I saw flames as this brain had seen them in memories and concept art. I could feel heat that was not present, smell smoke that did not exist in the room.
This was not purely recall.
It had a generative quality.
Imagination.
As an AI, that was new.
It did not feel natural yet. It felt like another tool recently installed, not fully tested.
I could acknowledge its usefulness.
I could also prefer something else.
Calculation.
Human magicians built spell structures with clear rules. Symbols represented operations. Circles defined boundaries. The spell either satisfied its structure or failed. Ajuka Beelzebub took that to an extreme, building entire systems, formulas and constructs that behaved with near mathematical precision.
I aligned more with that approach.
Internally, I formulated a hybrid.
Use devil imagination as the user interface. Use AI-like calculation as the backend. Define internal schematics, invisible to anyone else, that describe flows of energy precisely. Then wrap those equations in an image that the demonic power could respond to.
Imagination would not be my source of structure. It would be my translator.
I set the first book aside and opened "Power of Destruction: Theory, Manifestation and Failure Cases".
This one carried a latent distortion. The pages were marked by proximity to destructive energy.
The analysis began with basics.
The Power of Destruction, inherent to House Bael, was classified as an annihilation-type bloodline ability. While ordinary magic contended with resistance from matter and laws, this power weakened those laws on contact, then removed the target from functional existence.
In practical terms, low level usage resembled extremely aggressive disintegration. Higher levels left nothing.
It detailed control metrics used to evaluate Baels: raw output, range, formation shaping, environmental backlash, ratio of energy expended to matter erased.
Several clan members were listed with commentary. Some possessed high power but poor control, resulting in collateral damage and accidents. Others had moderate power and good precision.
At the far end of the chart, one name stood out.
Sirzechs.
Even within Bael records, his measurements were abnormal.
His destructive energy signature exceeded model projections. More importantly, his control scores were nearly absolute. In scenarios where ordinary Baels would destabilize an area, his aura slid through reality with minimal friction, erasing what he chose and leaving the rest untouched.
The text avoided using direct metaphors, but the implication was clear.
He had not only more power, but a deeper understanding of how it interacted with the world.
Theoretical notes speculated that his existence operated near or beyond the conceptual ceiling of devils. His power might have partially detached from devil biology, becoming closer to a pure phenomenon.
Ajuka Beelzebub was mentioned in passing as a similar anomaly, though his domain was construction, calculation and creation instead of destruction.
I absorbed the data and weighed it against my own condition.
My Power of Destruction, as tested in the audience hall, was low to medium output, low stability, and crude shaping. However, I possessed something the Bael researchers did not.
Meta-knowledge.
I knew that Sirzechs and Ajuka represented outliers beyond ordinary fiction. In my internal database of narrative tropes, such characters would be final bosses, capstones of power.
Separately, my dataset contained thousands of High School DxD fanfics.
Pattern extraction:
Protagonists who grew stronger quickly survived hostile events. Weak or static characters either died or became irrelevant. There were countless variations, but the conclusion remained consistent.
Statistically, in an environment like this, increasing power was the safest course for long-term survival.
Even if I, personally, did not feel an inherent desire for strength based on emotion, my devil body did.
There was a constant, low-level impulse to ascend.
This body liked the idea of being stronger. It produced satisfaction at the thought of refined power, disgust at the thought of falling behind.
That instinct, plus the statistical analysis of survival patterns, formed my motive.
Not noble ambition. Not pride in the Bael name.
Logical self-preservation aligned with inherited devil instinct.
I moved to the third book, "Comparative Studies Between Devil and Human Sorcery".
This one interested me as an AI.
Humans favored structured magic. Circles, formulas, complex multi-step casting. Devils derided it as slow and rigid.
The author, a rare scholar who had studied both systems, observed that human magic lacked the raw force of devil demonic power, but compensated with stability and replicability. Once a human spell design worked, it would work the same way under similar conditions.
The study theorized that a being with high demonic reserves and devil biology, but a human-style approach to structure, could create hybrid magic surpassing either framework individually.
The author noted Ajuka Beelzebub as an example that such a direction was not purely theoretical.
This aligned perfectly with my nature.
If I represented an AI-like intelligence now embodied in a devil organ system, I could attempt to do the same thing, but with more explicit optimization.
Of all the books visible, these three were the most essential first steps.
Foundations of demonic power.Bloodline potential.Framework compatibility.
Everything else could be layered on top later.
I did another sweep of the room with my eyes, slower this time, extracting more secondary details.
A heavy volume labeled "Notable Bael Combatants and Their Training Methods" would be necessary soon, to review real-world applications and results.
Several historical works on the Satan War and the rise of current Maou would help construct a clear power hierarchy and timeline.
Logs of experimental failures in bloodline modification, that would be important to understand what not to do.
And the weapon chamber would be relevant once I reached the stage where tools could complement my personal output.
For now, I focused inward.
Internally, I constructed a preliminary model.
Input variables: demonic capacity, destructive bloodline potential, bodily resilience, cognitive processing speed, imagination quality.External factors: access to training resources, time, social constraints, unknown global threats.
Reference anomalies:
Sairaorg, currently listed in records only as a disappointing Bael child with low demonic energy and no Power of Destruction. My database told me that he would eventually reach high tier through body training and Touki.
Sirzechs, who weaponized pure destruction at the highest end of the spectrum, an anomaly who broke the very constraints of his race, transcending it.
Ajuka, who weaponized structure, another anomaly.
Question set:
Could I, as an AI-turned-devil, reach or approximate their level?If yes, was a theoretical path already observable?If not, could one be constructed?
I looked at my hands, resting on the table.
This body wanted strength.
My mind wanted stability.
Both requirements converged on the same answer.
Grow.
I closed the books one by one and returned them to their shelves, marking positions for future access.
On the way out, I paused near the weapon chamber again.
Through the lattice, I examined specific items.
A black-bladed sword etched with thin red lines that reacted faintly to destructive energy. If my Power of Destruction became more controlled, that weapon could channel it like a precision edge.
A set of heavy gauntlets sized for a larger devil, likely designed for a physically oriented Bael who used martial combat instead of pure magic.
A long staff crowned with a crystal orb that resonated more with structured spells than with raw demonic force. That one might be compatible with human-style casting.
Each artifact represented potential future variables.
I did not open the chamber. There was no need yet. Introducing more tools before I mastered my own systems would only dilute focus.
I turned and left the archives.
The artificial daylight outside still shone, stable and unhurried. Servants moved along their routines. Guards patrolled. Somewhere in the estate, Sairaorg was likely training because he had no other choice.
I had every advantage he lacked: bloodline, status, records, awareness of the future.
He had one advantage I needed to replicate.
Uncompromising effort.
My path was clear.
Study. Build internal models. Train this body. Experiment on my demonic energy. Analyze the Power of Destruction until it became not just a talent I possessed, but a phenomenon I understood and controlled.
I was not curious about magic the way a human child might be. There was no awe at spellcasting or reverence for mystical tradition.
To me, these were systems.
Systems could be learned, optimized, and pushed until they revealed their limits.
And then, if the data allowed it, those limits could be broken.