danesabo/Obsidian
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Auto sync: 2025-10-20 12:44:27 (5 files changed)

Browse Source 
M  .sessions/Journal.vim

M  Writing/ERLM/main.tex

M  Writing/ERLM/state-of-the-art/v5.tex

A  Writing/Journal/JRNL-20251015-204549.md

A  Writing/Journal/JRNL-20251017-193513.md
This commit is contained in:
Dane Sabo 2025-10-20 12:44:27 -04:00
commit ded8626fa1
5 changed files with 699 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

51
.sessions/Journal.vim
View File

@ -17,10 +17,52 @@ badd +1 ~/Documents/Dane\'s\ Vault/Journal/2025_07_30.md
badd +10 JRNL-20250904-135850.md
badd +96 JRNL-20251003-174601.md
badd +14 journal_config.txt
badd +95 JRNL-20251012-210736.md
argglobal
%argdel
edit journal_config.txt
edit JRNL-20251012-210736.md
let s:save_splitbelow = &splitbelow
let s:save_splitright = &splitright
set splitbelow splitright
wincmd _ | wincmd |
vsplit
1wincmd h
wincmd w
let &splitbelow = s:save_splitbelow
let &splitright = s:save_splitright
wincmd t
let s:save_winminheight = &winminheight
let s:save_winminwidth = &winminwidth
set winminheight=0
set winheight=1
set winminwidth=0
set winwidth=1
exe 'vert 1resize ' . ((&columns * 93 + 93) / 186)
exe 'vert 2resize ' . ((&columns * 92 + 93) / 186)
argglobal
balt journal_config.txt
setlocal foldmethod=manual
setlocal foldexpr=0
setlocal foldmarker={{{,}}}
setlocal foldignore=#
setlocal foldlevel=0
setlocal foldminlines=1
setlocal foldnestmax=20
setlocal foldenable
silent! normal! zE
let &fdl = &fdl
let s:l = 95 - ((21 * winheight(0) + 32) / 64)
if s:l < 1 | let s:l = 1 | endif
keepjumps exe s:l
normal! zt
keepjumps 95
normal! 0
wincmd w
argglobal
if bufexists(fnamemodify("journal_config.txt", ":p")) | buffer journal_config.txt | else | edit journal_config.txt | endif
if &buftype ==# 'terminal'
silent file journal_config.txt
endif
balt JRNL-20251003-174601.md
setlocal foldmethod=manual
setlocal foldexpr=0
@ -37,7 +79,10 @@ if s:l < 1 | let s:l = 1 | endif
keepjumps exe s:l
normal! zt
keepjumps 14
normal! 0
normal! 019|
wincmd w
exe 'vert 1resize ' . ((&columns * 93 + 93) / 186)
exe 'vert 2resize ' . ((&columns * 92 + 93) / 186)
tabnext 1
if exists('s:wipebuf') && len(win_findbuf(s:wipebuf)) == 0 && getbufvar(s:wipebuf, '&buftype') isnot# 'terminal'
silent exe 'bwipe ' . s:wipebuf
@ -45,6 +90,8 @@ endif
unlet! s:wipebuf
set winheight=1 winwidth=20
let &shortmess = s:shortmess_save
let &winminheight = s:save_winminheight
let &winminwidth = s:save_winminwidth
let s:sx = expand("<sfile>:p:r")."x.vim"
if filereadable(s:sx)
exe "source " . fnameescape(s:sx)

2
Writing/ERLM/main.tex
View File

@ -4,7 +4,7 @@
\maketitle
\input{goals-and-outcomes/v6}
\input{state-of-the-art/v4}
\input{state-of-the-art/v5}
\input{research-approach/v3}
\input{broader-impacts/v1}
\input{metrics-of-success/v1}

357
Writing/ERLM/state-of-the-art/v5.tex
View File

@ -1,18 +1,100 @@
\section{State of the Art and Limits of Current Practice}
The principal aim of this research is to create autonomous reactor control
systems that are tractably safe. But, to understand what exactly is being
automated, it is important to understand how nuclear reactors are operated
today. First, the reactor operator themselves is discussed. Then, operating
procedures that we aim to leverage later are examined. Next, limitations of
human-based operation are investigated, while finally we discuss current formal
methods based approaches to building reactor control systems.
\subsection{Current Reactor Procedures and Operation}
%How are operating procedures made and why do they exist
%what are different kinds of operating procedures
Current generation nuclear power plants employ 3,600+ active NRC-licensed
reactor operators in the United States. These operators are divided into Reactor
Operators (ROs) who manipulate reactor controls and Senior Reactor Operators
(SROs) who direct plant operations and serve as shift
supervisors~\cite{10CFR55}. Staffing typically requires 2+ ROs with at least one
SRO for current generation units. To become a reactor operator, an individual
might spend up to six years to pass required training~\cite{princeton}.
%NUREG 0899
The role of human operators is paradoxically both critical and
problematic. Operators hold legal authority under 10 CFR Part 55 to make
critical decisions including departing from normal regulations during
emergencies. The Three Mile Island (TMI) accident demonstrated how
``combination of personnel error, design deficiencies, and component
failures'' led to partial meltdown when operators ``misread confusing
and contradictory readings and shut off the emergency water
system''~\cite{Kemeny1979}. The President's Commission on TMI identified
a fundamental ambiguity: placing ``responsibility and accountability for
safe power plant operations...on the licensee in all circumstances''
without formal verification that operators can fulfill this
responsibility under all conditions~\cite{Kemeny1979}. This tension
between operational flexibility and safety assurance remains unresolved
in current practice.
<<<<<<< HEAD
%how are procedures tested
=======
Nuclear plant procedures exist in a hierarchy: normal operating procedures for
routine operations, abnormal operating procedures for off-normal conditions,
Emergency Operating Procedures (EOPs) for design-basis accidents, Severe
Accident Management Guidelines (SAMGs) for beyond-design-basis events, and
Extensive Damage Mitigation Guidelines (EDMGs) for catastrophic damage
scenarios. These procedures must comply with 10 CFR 50.34(b)(6)(ii) and are
developed using guidance from NUREG-0899~\cite{NUREG-0899}, but their
development process relies fundamentally on expert judgment and simulator
validation rather than formal verification. Procedures undergo technical
evaluation, simulator validation testing, and biennial review as part of
operator requalification under 10 CFR 55.59~\cite{10CFR55}. Despite these
rigorous development processes, procedures fundamentally lack formal
verification of key safety properties. There is no mathematical proof that
procedures cover all possible plant states, that required actions can be
completed within available timeframes under all scenarios, or that transitions
between procedure sets maintain safety invariants.
%Automation already is used for emergency systems
\textbf{LIMITATION:} \textit{Procedures lack formal verification of correctness
and completeness.} Current procedure development relies on expert judgment and
simulator validation. No mathematical proof exists that procedures cover all
possible plant states, that required actions can be completed within available
timeframes, or that transitions between procedure sets maintain safety
invariants. Paper-based procedures cannot ensure correct application, and even
computer-based procedure systems lack the formal guarantees that automated
reasoning could provide.
Nuclear plants operate with multiple control modes: automatic control where the
reactor control system maintains target parameters through continuous rod
adjustment, manual control where operators directly manipulate control rods, and
various intermediate modes. In typical pressurized water reactor operation, the
reactor control system automatically maintains a floating average temperature,
compensating for changes in power demand with reactivity feedback loops alone.
Safety systems instead operate with implemented automation. Reactor
Protection Systems trip automatically on safety signals with millisecond
response times, and engineered safety features actuate automatically on accident
signals without operator action required.
>>>>>>> 568549999a24c6a86f19411cbdf12b642057ade9
The current division between automated and human-controlled functions
reveals the fundamental challenge of hybrid control. Highly
automated systems handle reactor protection like automatic trips on safety
parameters, emergency core cooling actuation, containment isolation,
and basic process control. Human operators, however, retain control of
strategic decision-making such as power level changes, startup/shutdown
sequences, mode transitions, and procedure implementation. %%%NEED MORE
\textbf{LIMITATION:} \textit{Current practice treats continuous plant
dynamics and discrete control logic separately.} No application of
hybrid control theory exists that could provide mathematical guarantees
across mode transitions, verify timing properties formally, or optimize
the automation-human interaction trade-off with provable safety bounds.
\subsection{Human Factors in Nuclear Accidents}
The persistent role of human error in nuclear safety incidents, despite
decades of improvements in training and procedures, provides perhaps the
most compelling motivation for formal automated control with
mathematical safety guarantees.
<<<<<<< HEAD
%Whos in the control room
%how are reactor operators trained
@ -39,3 +121,270 @@
%details of how it worked, and limitations therein
%Digital system ONLY
=======
Multiple independent analyses converge on a striking statistic: \textbf{70--80\%
of all nuclear power plant events are attributed to human error} versus
approximately 20\% to equipment failures~\cite{DOE-HDBK-1028-2009,WNA2020}. More
significantly, the International Atomic Energy Agency concluded that ``human
error was the root cause of all severe accidents at nuclear power plants''---a
categorical statement spanning Three Mile Island, Chernobyl, and Fukushima
Daiichi~\cite{IAEA-severe-accidents}. A detailed analysis of 190 events at
Chinese nuclear power plants from 2007--2020~\cite{Wang2025} found that 53\% of
events involved active errors while 92\% were associated with latent errors
(organizational and systemic weaknesses that create conditions for failure). The
persistence of this 70--80\% human error contribution despite four decades of
continuous improvements in operator training, control room design, procedures,
and human factors engineering. This suggests fundamental cognitive limitations
rather than remediable deficiencies.
The Three Mile Island Unit 2 accident on March 28, 1979 remains the definitive
case study in human factors failures in nuclear operations. The accident began
at 4:00 AM with a routine feedwater pump trip, escalating when a
pressure-operated relief valve (PORV) stuck open---draining reactor
coolant---but control room instrumentation showed only whether the valve had
been commanded to close, not whether it actually closed. When Emergency Core
Cooling System pumps automatically activated as designed, operators made the
fateful decision to shut them down based on their incorrect assessment of plant
conditions. The result was a massive loss of coolant accident and the core
quickly began to overheat. During the emergency, operators faced more than 100
simultaneous alarms, overwhelming their cognitive capacity~\cite{Kemeny1979}.
The core suffered partial meltdown with \textbf{44\% of the fuel melting} before
the situation was stabilized.
Quantitative risk analysis revealed the magnitude of failure in existing
safety assessment methods: the actual core damage probability was
approximately 5\% per year while Probabilistic Risk Assessment
had predicted 0.01\% per year---a \textbf{500-fold underestimation}.
This dramatic failure demonstrated that human reliability could not be
adequately assessed through expert judgment and historical data alone.
%%%SOURCE??? Human Reliability Analysis (HRA) methods developed over four decades
quantify human error probabilities and performance shaping factors. The
SPAR-H method represents current best practice,
providing nominal Human Error Probabilities (HEPs) of \textbf{0.01 (1\%)
for diagnosis tasks} and \textbf{0.001 (0.1\%) for action tasks} under
optimal conditions~\cite{NUREG-CR-6883}.
However, these nominal error rates degrade dramatically under realistic
accident conditions: inadequate available time increases HEP by
\textbf{10-fold}, extreme stress by \textbf{5-fold}, high complexity by
\textbf{5-fold}, missing procedures by \textbf{50-fold}, and poor
ergonomics by \textbf{50-fold}. Under combined adverse conditions
typical of severe accidents, human error probabilities can approach
\textbf{0.1 to 1.0 (10\% to 100\%)}---essentially guaranteed failure for
complex diagnosis tasks~\cite{NUREG-2114}.
Rasmussen's influential 1983 taxonomy~\cite{Rasmussen1983} divides human errors
into skill-based (highly practiced responses, HEP $10^{-3}$ to $10^{-4}$),
rule-based (following procedures, HEP $10^{-2}$ to $10^{-1}$), and
knowledge-based (novel problem solving, HEP $10^{-1}$ to 1). Severe accidents
inherently require knowledge-based responses where human reliability is lowest.
Miller's classic 1956 finding~\cite{Miller1956} that working memory capacity is
limited to 7$\pm$2 chunks explains why Three Mile Island's 100+
%WHAT IS A CHUNK?
simultaneous alarms exceeded operators' processing capacity.
\textbf{LIMITATION:} \textit{Human factors impose fundamental reliability
limits that cannot be overcome through training alone.} Response time
limitations constrain human effectiveness---reactor protection systems
must respond in milliseconds, 100--1000 times faster than human
operators. Cognitive biases systematically distort judgment:
confirmation bias, overconfidence, and anchoring bias are inherent
features of human cognition, not individual failings~\cite{Reason1990}.
The persistent 70--80\% human error contribution despite four decades of
improvements demonstrates that these limitations are fundamental
rather than remediable.
\subsection{HARDENS and Formal Methods}
The High Assurance Rigorous Digital Engineering for Nuclear Safety
(HARDENS) project, completed by Galois, Inc. for the U.S. Nuclear
Regulatory Commission in 2022, represents the most advanced application
of formal methods to nuclear reactor control systems to
date---and simultaneously reveals the critical gaps that remain.
\subsubsection{Rigorous Digital Engineering Demonstrated Feasibility}
HARDENS aimed to address the nuclear industry's fundamental dilemma:
existing U.S. nuclear control rooms rely on analog technologies from the
1950s--60s, making construction costs exceed \$500 million and timelines
stretch to decades. The NRC contracted Galois to demonstrate that
Model-Based Systems Engineering and formal methods could design, verify,
and implement a complex protection system meeting regulatory criteria at
a fraction of typical cost.
The project delivered far beyond its scope, creating what Galois
describes as ``the world's most advanced, high-assurance protection
system demonstrator.'' Completed in \textbf{nine months at a tiny
fraction of typical control system costs}~\cite{Kiniry2022}, the project
produced a complete Reactor Trip System (RTS) implementation with full
traceability from NRC Request for Proposals and IEEE standards through
formal architecture specifications to formally verified binaries and
hardware running on FPGA demonstrator boards.
Principal Investigator Joseph Kiniry led the team in applying Galois's
Rigorous Digital Engineering methodology combining model-based
engineering, digital twins with measurable fidelity, and applied formal
methods. The approach integrates multiple abstraction levels---from
semi-formal natural language requirements through formal specifications
to verified implementations---all maintained as integrated artifacts
rather than separate documentation prone to divergence.
\subsubsection{Comprehensive Formal Methods Toolkit Provided Verification}
HARDENS employed an impressive array of formal methods tools and
techniques across the verification hierarchy. High-level specifications
used Lando, SysMLv2, and FRET (NASA JPL's Formal Requirements
Elicitation Tool) to capture stakeholder requirements, domain
engineering, certification requirements, and safety requirements.
Requirements were formally analyzed for \textbf{consistency,
completeness, and realizability} using SAT and SMT solvers---verification
that current procedure development methods lack.
Executable formal models employed Cryptol to create an executable
behavioral model of the entire RTS including all subsystems, components,
and formal digital twin models of sensors, actuators, and compute
infrastructure. Automatic code synthesis generated formally verifiable C
implementations and System Verilog hardware implementations directly
from Cryptol models---eliminating the traditional gap between
specification and implementation where errors commonly arise.
Formal verification tools included SAW (Software Analysis Workbench) for
proving equivalence between models and implementations, Frama-C for C
code verification, and Yosys for hardware verification. HARDENS verified
both automatically synthesized and hand-written implementations against
their models and against each other, providing redundant assurance
paths.
This multi-layered verification approach represents a quantum leap
beyond current nuclear I\&C verification practices, which rely primarily
on testing and simulation. HARDENS demonstrated that \textbf{complete
formal verification from requirements to implementation is technically
feasible} for safety-critical nuclear control systems.
\subsubsection{Critical Limitation: Discrete Control Logic Only}
Despite its impressive accomplishments, HARDENS has a fundamental
limitation directly relevant to hybrid control synthesis: \textbf{the
project addressed only discrete digital control logic without modeling
or verifying continuous reactor dynamics}. The Reactor Trip System
specification and formal verification covered discrete state transitions
(trip/no-trip decisions), digital sensor input processing through
discrete logic, and discrete actuation outputs (reactor trip commands).
The system correctly implements the digital control logic for reactor
protection with mathematical guarantees.
However, the project did not address continuous dynamics of nuclear
reactor physics including neutron kinetics, thermal-hydraulics, xenon
oscillations, fuel temperature feedback, coolant flow dynamics, and heat
transfer---all governed by continuous differential equations. Real
reactor safety depends on the interaction between continuous processes
(temperature, pressure, neutron flux evolving according to differential
equations) and discrete control decisions (trip/no-trip, valve
open/close, pump on/off). HARDENS verified the discrete controller in
isolation but not the closed-loop hybrid system behavior.
\textbf{LIMITATION:} \textit{HARDENS addressed discrete control logic
without continuous dynamics or hybrid system verification.} Hybrid
automata, differential dynamic logic, or similar hybrid systems
formalisms would be required to specify and verify properties like ``the
controller maintains core temperature below safety limits under all
possible disturbances''---a property that inherently spans continuous and
discrete dynamics. Verifying discrete control logic alone provides no
guarantee that the closed-loop system exhibits desired continuous
behavior such as stability, convergence to setpoints, or maintained
safety margins.
\subsubsection{Experimental Validation Gap Limits Technology Readiness}
The second critical limitation is \textbf{absence of experimental
validation} in actual nuclear facilities or realistic operational
environments. HARDENS produced a demonstrator system at Technology
Readiness Level 3--4 (analytical proof of concept with laboratory
breadboard validation) rather than a deployment-ready system validated
through extended operational testing. The NRC Final Report explicitly
notes~\cite{Kiniry2022}: ``All material is considered in development and
not a finalized product'' and ``The demonstration of its technical
soundness was to be at a level consistent with satisfaction of the
current regulatory criteria, although with no explicit demonstration of
how regulatory requirements are met.''
The project did not include deployment in actual nuclear facilities,
testing with real reactor systems under operational conditions,
side-by-side validation with operational analog RTS systems, systematic
failure mode testing (radiation effects, electromagnetic interference,
temperature extremes), actual NRC licensing review, or human factors
validation with licensed nuclear operators in realistic control room
scenarios.
\textbf{LIMITATION:} \textit{HARDENS achieved TRL 3--4 without experimental
validation.} While formal verification provides mathematical correctness
guarantees for the implemented discrete logic, the gap between formal
verification and actual system deployment involves myriad practical
considerations: integration with legacy systems, long-term reliability
under harsh environments, human-system interaction in realistic
operational contexts, and regulatory acceptance of formal methods as
primary assurance evidence.
\subsection{Research Imperative: Formal Hybrid Control Synthesis}
Three converging lines of evidence establish an urgent research
imperative for formal hybrid control synthesis applied to nuclear
reactor systems.
\textbf{Current reactor control practices} reveal fundamental gaps in
verification. Procedures lack mathematical proofs of completeness or
timing adequacy. Mode transitions preserve safety properties only
informally. Operator decision-making relies on training rather than
verified algorithms. The divide between continuous plant dynamics and
discrete control logic has never been bridged with formal methods.
Despite extensive regulatory frameworks developed over six decades,
\textbf{no mathematical guarantees exist} that current control approaches
maintain safety under all possible scenarios.
\textbf{Human factors in nuclear accidents} demonstrate that human error
contributes to 70--80\% of nuclear incidents despite four decades of
systematic improvements. The IAEA's categorical statement that ``human
error was the root cause of all severe accidents'' reveals fundamental
cognitive limitations: working memory capacity of 7$\pm$2 chunks,
response times of seconds to minutes versus milliseconds required,
cognitive biases immune to training, stress-induced performance
degradation. Human Reliability Analysis methods document error
probabilities of 0.001--0.01 under optimal conditions degrading to
0.1--1.0 under realistic accident conditions. These limitations
\textbf{cannot be overcome through human factors improvements alone}.
\textbf{The HARDENS project} proved that formal verification is
technically feasible and economically viable for nuclear control
systems, achieving complete verification from requirements to
implementation in nine months at a fraction of typical costs. However,
HARDENS addressed only discrete control logic without considering
continuous reactor dynamics or hybrid system verification, and the
demonstrator achieved only TRL 3--4 without experimental validation in
realistic nuclear environments. These limitations directly define the
research frontier: \textbf{formal synthesis of hybrid controllers that
provide mathematical safety guarantees across both continuous plant
dynamics and discrete control logic}.
The research opportunity is clear. Nuclear reactors are quintessential
hybrid cyber-physical systems where continuous neutron kinetics,
thermal-hydraulics, and heat transfer interact with discrete control
mode decisions, trip logic, and valve states. Current practice treats
these domains separately---reactor physics analyzed with simulation,
control logic verified through testing, human operators expected to
integrate everything through procedures. \textbf{Hybrid control
synthesis offers the possibility of unified formal treatment} where
controllers are automatically generated from high-level safety
specifications with mathematical proofs that guarantee safe operation
across all modes, all plant states, and all credible disturbances.
Recent advances in hybrid systems theory---including reachability
analysis, barrier certificates, counterexample-guided inductive
synthesis, and satisfiability modulo theories for hybrid systems---provide
the theoretical foundation. Computational advances enable verification of
systems with continuous state spaces that were intractable a decade ago.
The confluence of mature formal methods, powerful verification tools
demonstrated by HARDENS, urgent safety imperatives documented by
persistent human error statistics, and fundamental gaps in current
hybrid dynamics treatment creates a compelling and timely research
opportunity.
>>>>>>> 568549999a24c6a86f19411cbdf12b642057ade9

138
Writing/Journal/JRNL-20251015-204549.md Normal file
View File

@ -0,0 +1,138 @@
---
id: JRNL-20251015-204549
title: Wednesday, October 15, 2025 - 08:45 PM
type: journal
created: 2025-10-16T00:45:49Z
modified: 2025-10-16T00:45:49Z
tags: [journal]
---
# Wednesday, October 15, 2025 - 08:45 PM
Today is a new journal day, and I have some updates! The
principal topic of this entry is going to be about Matilda,
and a little bit about Sam. Nothing new is really happening
at work recently, other than I've gone to the gym the past
couple of days and that's been fun! I'm quite proud of that.
So here's the deal about Matilda: I need to chill the fuck
out. I'm having conversations back and forth about breaking
up with this chick when *I don't really know her*. I have
known her for *one* month, and we have been dating for
*four* days. How can I possibly really know her?
We went on another two dates. The first one was a bike ride
on Tuesday afternoon, where we rode about 8 miles and
stopped for burgers and beers in the middle. I felt like
things were kind of awkward. We talked about random stuff,
and things were kind of goofy. I think it was somewhat
mutual, but I dunno I feel like maybe I was in a weird
headspace. We chatted about my car some, and I did get a
little bit of a weird reaction when I was like "yeah
completion is probably 5-10 years away". I think that
stunned her a little bit. We wrapped up and things were
okay! A different kind of date than we're used to but it was
a nice ride nonetheless. I think it kinda just goes to show
maybe it's something we don't really share well as a couple.
I like to go fast and I don't think she can keep up :eyes:.
After that date though, I've been in some more turmoil about
what I should do about the relationship. Should I break up
with her? Are there dealbreakers I can't handle? Why did I
move so fast? Why do I escalate based on emotions? And
generally I've been getting a pit in my stomach sometimes
when she says something like "you make me feel special". I
thought this was a gut instinct telling me that I don't want
her to feel that way or that I'm not ready for that
commitment but this most recent date has significantly
changed my perspective.
Today I was planning on seeing her to break up with her. I
had some sandpaper I was going to give her for these glasses
she's making out of wine bottles, and then go on a walk with
her. I had planned it out, I had rehearsed it with Claude
(shoutout Claude btw), and felt pretty prepared for what was
going to go down. But when I saw her, and we started
walking, my gut feeling was very different. She was holding
my hand, and we were just chatting and bs'ing about our days
when I realized that I really don't know her, I really don't
know that this couldn't work, and it's really not so serious
as it feels in my head some times.
I took a step back on my thoughts of her falling for me and
really dug into why I feel that way. I think I'm getting
ahead of myself saying that, because we're still learning
about each other and I think she's totally taking things
more slowly and casually than I am. No one is seeing wedding
bells right now, instead it's really more like we're getting
to know one another with some exclusivity attached. It isn't
so high stakes. Why did I get so anxious about this? Why did
I catastrophize things with problems that don't really exist
yet?
I keep getting tied up on the idea that I need a partner
with whom I can tinker with. Why? I have friends that I do
that with, and honestly, like working alone a lot of the
time. Do I need that in a partner? Maybe, but also maybe
not? Why don't I use this relationship to really actually
find that out? Also, she might actually be interested in
some of that stuff? Who knows? The answer is, not me.
Here's what I do know: when we went on that walk, when we
sat down and talked about our days, when we talked about
random stuff (like data privacy laws, government debt, and
voluntary euthanasia????? Sidebar, that was crazy) I felt
*so* relaxed. It was easy to just chill with her and enjoy
the sunset.
So I've been on this thing where the emotions attached to
her have been sinusoidal. I've gotta be honest, I think
that's my fault, and does not help my clarity in making a
decision in what I actually need in a partner. But, Matilda
is *not* a partner yet. We're right at the beginning of a
relationship, where *both* of us are figuring out if this is
right or not. And that does *not* happen in 4 days. The
things that I've been catastrophizing about are stressing me
out for no reason, and are not real problems yet. It is my
goal for this next week, to just calm down, and try to
journal and relax before making decisions. Ideally, don't
make *any* decisions this next week. What I feel like I
really need is some stability. No escalations, but also I
don't need to break up with her for no good reason either. I
can try and figure my shit out at the same time, knowing
that things aren't super serious right now. I don't need to
put some much pressure on myself.
Matilda is really nice to me, makes me feel calm when I'm
with her, and really stimulates me intellectually just
chatting about random stuff. Why I get so anxious when I'm
not around her is my problem to figure out, and not a reason
to drive decisions. For now, I'm going to keep getting to
know her and try to relax a little. No one is getting
married or having kids next month. To quote what
best-buddy-in-chief Sam said "You know, you're allowed to be
happy." He's right, and I don't need to create new problems.
Then, there is recent details about Samuel! Sam is
struggling some. We went golfing on Monday morning, which
was a blast. I need to make it a priority to show up earlier
than tee times though, as I got there basically 4 minutes
before tee off. Anyways, Sam feels listless. He doesn't love
his company, and Blake has been stressing him out some with
her not really advancing her accounting certifications right
now. I understand. I told him he should set some concrete
goals. If he wants to leave his company, he should set a
target date and write it down (sort of like I'm doing here).
Maybe he should set some goals on his hobbies. The point
was, make it measurable! I think that will help him a lot.
Poker is probably falling through this Friday, with people
being busy. Maybe I'll try to hangout with Sam anyways.
Matilda has her aunt coming into town this weekend, so the
next time we're going to see each other is Sunday. No more
escalating for now, no more "should I blow it all up????"
for a bit, instead, let me just enjoy getting to know this
new person without putting the Atlas-like pressure on it.
That can be enough.

158
Writing/Journal/JRNL-20251017-193513.md Normal file
View File

@ -0,0 +1,158 @@
---
id: JRNL-20251017-193513
title: Friday, October 17, 2025 - 07:35 PM
type: journal
created: 2025-10-17T23:35:13Z
modified: 2025-10-17T23:35:13Z
tags: [journal]
---
# Friday, October 17, 2025 - 07:35 PM
Yeah, so y'know how I said I was going to enjoy getting to
know her, and that that would be enough? Yeah, that didn't
work out. I broke up with Matilda yesterday.
Yesterday I went to my grandparents to do some laundry. In
the morning I felt okay about how things were left with
Matilda, but as the day progressed, bits and pieces of that
anxiety about the relationship and issues I had kept
creeping in. It's really hard because Matilda is such a nice
and good person, but ultimately, there were things about my
values that didn't line up. We were keeping a
Sunday-sleepover up in the air, and this was also kind of a
determining factor for me. I really think that if she would
have stayed over on Sunday that we would have had sex. Or at
least, I would have had to refuse and make things very
awkward.
Matilda told me early on that she didn't want to have sex
unless she was in a committed relationship. I said that's
completely fair! I totally get it. Then, about two weeks
ago, she gave me a heads up that she was getting back on
birth control. It was a weird heads up to get, but we talked
about it, agreed we were going in that direction, and it was
okay. But, recently, there had been some flirtatious action
where we were basically soft sexting that really indicated
to me we would have sex soon. And with me on the fence about
the whole thing, escalating to sex is the exact opposite of
what I wanted to do. I told myself last entry that I would
not escalate, and that I would calm down. Well, I tried to
calm down, but I've arranged things that I will definitely
not be escalating.
I told my grandparents about the situation. How she's such a
nice person, but I can't shake this gut feeling that
something is wrong. They try to be helpful, and are
generally supportive, but they met in high school and have
been together since. Neither of them have really 'dated'. So
to that end, I was kind of on my own.
Then, I'm sitting in the basement while the dryer is running
and I got to a point where I thought to myself "I cannot
keep waffling on this, this anxiety is killing me and I need
to just end it." I texted her asking her how her evening was
going, and she responded that it was going well. She also
sent me a picture of the Christmas tree Downtown and said
'Great news!', and then asked if I'd want to go ice skating
with her or maybe go on a double date with Sam and Blake ice
skating. I never texted back.
I sat in a chair instead, and thought for a long time about
what I was going to do. I thought for a long time if
sticking it out was worth it, but I thought about
conversations with Claude and thought about 'If I don't do
this now, how am I going to feel in a week or two weeks?'.
The answer to that question is still uncomfortable if I stay
this course.
At this point, Krzyszstof calls me randomly. Buddy's
Mercedes broke down in the middle of campus and he was
asking for help. I couldn't help him obviously being way out
at my grandparents, but his interruption broke me of my
stupor and I got the courage to call Matilda afterwards.
I asked her how her afternoon was going. She said it was
going well, and sounded very happy. She was arranging
flowers for her mom's birthday this weekend. She is so sweet
in that way. We had some small talk for a while before I
changed the topic to us. I said that I've been thinking
about us, and that I know this is sudden and is going to
sound like a shock, but I think we should stop seeing each
other. There was a long pause on the phone. When she spoke,
her tone was immediately different and she politely asked
why. I told her that over the past week, I've been dealing
with some personal emotional difficulties and I've come to
the conclusion that it is not a good time for me to be in a
relationship right now. I told her I think she's wonderful,
and that I think she is such an incredibly nice person in
the way that she moves through the world and that I liked
her a lot, but that she deserves someone who is all in for
her and I just don't think I can be that guy for her. I
apologized saying that I wish I knew this beforehand, but I
didn't, and am sorry that I couldn't get there for her. She
started to cry softly, and told me that I didn't need to
hear it from her, but that besides being so smart, that I'm
incredibly compassionate and thoughtful and know how to make
her feel special. She appreciated me being honest and said
that this sucks, but she understands. I told her the one
thing that's most important to me for her to take away from
this conversation is that this is not her fault. I offered
if she wanted to be friends at some point I would be open to
that, but that I know we're both going to need our space
after this jolt. I said I genuinely think she's great and that she
will find her person, just that I don't think it's me. I
apologized again. She told me that she is thankful I was
honest, and that she's rooting for me. She enjoyed our
relationship even though it was so short. We wished each
other good luck.
And that's the last that I talked to her. I started to cry
some too. I wish I could've done this in person on
Wednesday, but for some reason, I just couldn't.
Afterwards I called Sam on the phone to talk to him. He was
supportive in the sense that he knows I've got to do what's
right for me, but cautioned me that he thinks I can make
extremely reactive decisions. He's right. It's a weakness
and a strength. On one hand, I'm able to cut my losses at
times and move on quickly to the next thing, or adjust to
changing circumstances, but on the other, sometimes I can
act too quickly before I have all the cards. I don't think
this situation is like that. Or at least this breakup,
anyways. I've been dealing with this doubt for weeks, while
really the reactive decision was escalating to asking her to
be my girlfriend.
It's over. This sucks, but I know deep down it's the right
choice. My anxiety about the relationship is completely
gone, and instead replaced with a profound loneliness. Part
of me wonders if I blew up something that was good for no
reason, but another part of me knows that's not seeing the
forest for the trees. I feel so bad about this whole
situation and hope these feelings subside soon. It's hard
not to think about. I haven't texted her, and I deleted our
conversation so I can't our chat history. It sucks that when
a relationship like this ends, the whole friendship and
camaraderie ends so quickly too. It's such a brutal cut and
feels like a void has opened up out of nowhere.
I went for a ride today on the bike. I rode over to South
Park to whip around, and then stopped at Brusters. That was
a *blast*. Anyways, I'm still bummed, but I know this was
the right move and things will get better. Luis told me
today that things like this are the price of dating and that
it comes with the territory. He's right.
I redownloaded Hinge. I'm not using it right now, but if
Cinderella likes me I'll at least see it. Up next I'm going
to really write down what I'm looking for in a partner in my
next journal entry. This way, I'll have a list of clear
targets to hit, and who knows, maybe I'll manifest her into
existence. It worked for Lane, at least.
I see Rachel on Tuesday. Yeugh. Time to watch some of The
Pitt tonight and try to relax. I think I might travel into
Pitt tomorrow to go to the gym. I feel like I need it.
I love you, me. We're figuring this shit out.