Chapter 14: What the Synthesis Core Does at 3 AM
The warmth behind my sternum woke me at 3:07 AM.
The sensation was familiar — the specific frequency of the Synthesis Core completing a processing cycle, generating an output that required conscious attention. I had been trained to recognize it in Academy City, where research schedules did not respect normal sleep patterns and important insights could arrive at any hour.
I sat up in bed and reached for the notebook on my nightstand.
The output came in fragments at first — mathematical relationships, notational connections, the specific structure of a cross-reference completing its synthesis. I wrote without evaluating, transcribing the output before the connection degraded. The Synthesis Core required complete transcription or the link between elements would weaken, leaving gaps that were difficult to reconstruct later.
Three pages.
I finished writing at 3:34 AM and turned on the lamp to read what I had produced.
The first two pages were comprehensible. The Synthesis Core had built a bridge between Academy City's esper field measurement methodology and Sheldon's string theory notation — the question he had been working on at his whiteboard, the one with the annotations I had seen on my first day. The bridge generated a third framework that was neither Academy City nor Caltech standard, but drew from both.
I could follow the logic. The mathematics was elegant in the specific way that cross-disciplinary syntheses were elegant — not complicated for its own sake, but complex because it was connecting two systems that had never been designed to connect.
The third page was different.
The mathematics continued, but somewhere around the middle of the page, the physical interpretation exceeded my current theoretical vocabulary. I could verify that the equations were internally consistent — the notation followed predictable rules, the transformations were valid, the conclusions followed from the premises. But I could not translate what the equations described into any physical phenomenon I understood.
The Synthesis Core had produced something beyond my ability to interpret.
I read the third page three times. Each reading confirmed the same conclusion: the mathematics was sound, but the meaning was inaccessible. Whatever the equations described, it was not covered by Academy City's esper physics curriculum or Caltech's conventional framework or any of the encoded methodologies I had absorbed since arriving.
I labeled the page: "HOLD — REVISIT."
At 4 AM, I sat in my apartment kitchen with the three pages spread on the table and considered my options.
Option one: write the output on a whiteboard somewhere public. The framework would attract attention, generate discussion, potentially accelerate the collective understanding of whatever the third page described. It would also create an evidence trail, raise questions about where the framework came from, and give Sheldon another data point for his growing investigation.
Option two: keep the output in my notebook. Safe, private, retrievable for later analysis. The framework would remain mine until I chose to share it. The third page would stay mysterious until I developed the vocabulary to interpret it.
Option three: destroy the output. Remove the evidence, eliminate the risk. But Synthesis Core outputs did not always regenerate — the specific combination of inputs that had produced this synthesis might not recur, and the framework could be lost permanently.
None of the options were good.
I chose a fourth option: partition.
The first two pages went into my main notebook, filed with the rest of my Caltech observations. They were comprehensible, useful, and unlikely to raise questions if discovered. The framework they described could plausibly come from a competent researcher with cross-disciplinary training.
The third page went into a separate notebook — smaller, unremarkable, the kind of thing you could buy at any office supply store. I had purchased it two weeks ago for reasons I had not fully articulated at the time. Now I understood why.
The secondary notebook went into the lining of my spare bag, in the compartment designed for a laptop that I did not use for that purpose. It was the first time I had hidden something from myself — the first time I had created a separation between outputs I could safely carry and outputs that required additional security.
The protocol crystallized as I made the decisions: comprehensible outputs in the main notebook, incomprehensible outputs in secondary storage, nothing on public surfaces. The Synthesis Core was growing faster than I was. The outputs were accelerating beyond my processing rate. Managing the gap required new systems.
At 7 AM, I made coffee.
The process was unremarkable except for the part where I used the Molecular Conductor at CL 1 to feel the water temperature precisely, adjusting the heat without touching the kettle. The ability let me achieve exact extraction temperatures that manual methods could not match.
I had been doing this every morning for four weeks. It was the most comfortable use of my ability in Pasadena — minimal CL cost, no detectable signature, a practical application that improved my daily life without creating evidence. The coffee was better than anyone else in the building made because I could control variables they could not perceive.
The warmth from the Synthesis Core processing had not fully dissipated. The familiar heat sat behind my sternum, a reminder that the system was running at higher throughput than it had been when I arrived. Four weeks of Caltech had accelerated something. The encoding sessions, the cross-disciplinary exposure, the neural oscillation conversation with Amy, the engineering problem with Howard — each input had fed the Synthesis Core new material, and the system had grown in response.
I drank my coffee and felt the growth happening.
It did not feel comfortable.
At the physics building that morning, I glanced at Sheldon's whiteboard.
The question from my first day was still there — the annotation in the corner, the one asking about field harmonics and dimensional folding. Sheldon had modified it slightly over the intervening weeks, refining the question, narrowing the parameters. He was working on it the way he worked on everything: systematically, thoroughly, with complete documentation of every step.
The two-page framework in my notebook would address the question directly. The synthesis I had produced at 3 AM was exactly the kind of cross-disciplinary bridge that Sheldon's question required. Showing it to him would advance his work, potentially by months.
I did not write it on the whiteboard.
Instead, I made a note to ask Sheldon about the question conversationally — to see which direction he was taking it, to understand his current approach before deciding whether my output would help or confuse. The Witness Protocol was still encoding his methodology: how he held dead-end questions open rather than closing them, how he maintained research threads across weeks without losing interest or clarity.
The encoding took six seconds. I covered the dissociation tell by examining my phone.
"The framework you presented at the seminar," Sheldon said from his desk, not looking up from his computer. "The field harmonics notation. It has structural similarities to approaches I have been considering."
"The mathematics derives from the same theoretical tradition."
"Academy City's tradition."
"Yes."
"Which is non-standard relative to conventional physics."
"Also yes."
Sheldon was quiet for a moment. I could feel him processing — adding information to whatever model he was building, testing conclusions against data.
"The convention does not invalidate the approach," he said finally. "Non-standard frameworks can produce valid results if their internal logic is consistent."
"I agree."
"I am not yet certain whether your framework is consistent or merely appears consistent because I do not understand it fully."
"That's a reasonable position."
He looked at me then — the direct attention he usually reserved for data that surprised him.
"I will continue my review," he said.
He went back to his computer. I went to my desk.
The conversation was over, but the calibration notebook sat in its drawer, adding entries I could not see.
Get Early Access to New Chapters
Thank you for reading. For those who want to skip the wait, my Patreon is currently 21 chapters ahead of the public sites.
Schedule: 7 new chapters released every 10 days.
Benefit: Gain a significant lead of 7 to 21 chapters depending on your tier.
Support the project and start reading the next arc now: Patreon.com/IsekaiStories