The Ascendant Wizard

"Remove route four from consideration."

[Understood.]

The option disappeared, and Morena continued looking at the remaining three. Connection reinforcement, compression, and return filtration.

They were all useful, but they could not simply be added together without consequence. A spell was not a pile of separate parts; it was a structure. If one part changed, the others would need to adjust around it.

"Can all three be combined?"

[Possible.]

[Risk: Medium to High due to interaction between compression and return filtration.]

"Explain."

[Compression increases speed and density of returning feedback, while return filtration requires a stable intake rate. Due to this, there is an increased risk of a subpar outcome by attempting to adjust in both directions at the same time.]

So she could do it, but it would weaken the outcome. While it would still probably be better than the base spell itself, she would have to give up on specialization for that.

That was the issue with having too little information. The AI could see flaws, but seeing flaws was not the same as having the answer to them; it could only work with the data it had, and right now, the database was too shallow.

She needed more spells, models, theories, and anything she could get. The best would be proper first-hand experience and knowledge from Wizards on how they built their spells and foundations. But she didn’t have that now.

"Then create several models. Prioritize stability first, mana efficiency second, sensory clarity third. Do not increase risk beyond medium."

[Understood. Beginning model construction.]

The AI began to run countless calculations and simulations in the background, and it seemed to be taking quite a while this time around.

Morena sat in silence, watching the spell model shift again and again in her mind. Most of the early attempts were rejected, some improved range but worsened detail, while some reduced mana loss but weakened the spell so much that the pulse would barely leave the caster.

Time ticked by quicker than she expected; an hour, then another, and soon more. Morena did not rush it; she only gave instructions when needed and allowed it to work.

By the time her eyes began to ache from staring at the floating spell circle, the AI finally stopped changing the model.

[Model Three complete.]

[Changes Applied:]

[Connection Thread Reinforcement: Minor.]

[Pulse Channel Compression: Minor.]

[Return Filtration: Simplified adjustment.]

[Projected Result:]

[Mana loss reduced by an estimated 18-24%.]

[Range improved slightly.]

[Feedback clarity improved slightly.]

[Engraving difficulty increased by 12%.]

[Risk Level: Low.]

’Hmmm. It’s not bad.’

It wasn’t bad, but it wasn’t really good either. That was the first thing she thought; after several hours of work, it was pretty negligible.

But after thinking about it, that was already impressive.

The AI had taken one complete spell, a few fragments, and some theory, then improved the structure without breaking it. That alone was beyond what most apprentices could even dream of doing.

"Can it be improved further without more data?"

[Not recommended. Further changes will increase structural uncertainty; the current model is the safest improved version based on available information.]

Morena nodded slowly.

"Save it as Sensory Pulse Model Three."

[Saved.]

The original circle remained above the booklet. Beside it, inside her mind, the altered version slowly rotated.

They were similar enough that even Morena could recognize they were the same spell, but different enough that the improved one could be considered an alteration of the original.

If she showed it to others, they would probably just assume someone made a new spell that was similar, rather than upgrading the existing one. That was how complicated spell-making and modification were.

But with the spell model finished, the next problem arose. Changing the model was only the first step; learning it was something else entirely.

Even with the AI guiding her, carving a spell into the Matrix would not be simple. The structure had to be memorized until her mind could hold it without the booklet. Then she had to shape it with mana, not as a picture, but as an actual structure within her mental space; after that came the engraving itself.

It wasn’t uncommon for someone to take weeks, months, or even years if it was complicated. Morena did not have that much time, but she would not be able to finish it in one night either.

’Run a simulation, how long would carving the improved model be without the AI existing?’

[Without the assistance of the AI the estimated time of engraving is within 2-4 months, depending on external factors.]

’What about with the assistance of the AI?’

[Estimated time to prepare engraving: 3-5 days.]

Morena tapped her fingers against the desk again.

’Is my talent that poor? Or perhaps the model is just complicated?’

She knew it would take a while to do since it was her first spell, but she didn’t expect it to take months just for this. Thankfully, she had the use of the AI. With its use, it was only three to five days.

"Begin training plan and run various simulations to find the perfect way to engrave it. First stage only, memorization and shaping practice."

[Understood.]

[Stage One: Structural Memorization.]

[Goal: User must recreate Sensory Pulse Model Three internally with less than 3% deviation.]

[Estimated time: 1-2 days.]

The AI formed a virtual model in her mind as she closed her eyes, allowing her to visualize the different layers in great depth and detail. She could focus on each specific part until she had memorized it, instead of the entire thing in one go.

This would allow her to carefully piece it together like a puzzle, instead of overwhelming her mind at once.

"Start with the outer circle."

[Displaying outer circle.]

As she gave the command, the AI made only the outer part of the circle visible. With this advantage, Morena could focus solely on meditation, where her entire mind could shape her mana into the form it needed to be.

At first, it was hard; she had never used mana in such a way before, and she couldn’t properly get it to move the way she wanted it. It was like trying to move ink through water; it just kept leaking out and messing up.

But eventually, she got the hang of it.