Some have said that current LLMs feel like the shadow of humanity, while others say that LLM outputs seem to carry an "AI taste" unless specially tuned. Furthermore, exploring new research topics also seems to have some limitations?

Starting from the Context of Early Digital Tools

Let's start from some contexts of earlier digital tools. In fact, many convenient applications that appear intelligent must begin with probability, statistics, correlation, and prediction. For example, input methods and candidate words are essentially predicting the next word the user will type.

Additionally, Spotify playlists, recommended video channels on YouTube, and social media posts collect numerous quantified features, feed them into models for probabilistic predictions of associations, and continuously reinforce training through the user's own feedback, thus increasingly converging on the user's preferences.

But this context introduces some potential problems. Have you ever thought about why echo chambers emerge so easily? Because this pattern of predicting associations tends to reinforce the more superficial, visible patterns. The less probable and numerically unlikely ones will be pushed further away by this system. Projecting this onto reality a bit more, it makes it harder to come into contact with groups that have different traits and surface preferences than you do—unless you explore more proactively.

The Essence of LLMs: Predicting the Next Data Fragment

My abilities are relatively weak, so I cannot fully explain the principles from the underlying attention mechanism of LLMs, but the mechanism of LLMs can essentially be thought of as: training a model from all recorded data since human civilization began to predict the most likely next data fragment.

As for the strength of its predictive capabilities, how it aligns with human preferences, how to effectively utilize hardware resources, how to control the input data length during prediction so that it degrades as little as possible, and how the importance fluctuates—these depend on the tuning capabilities of different models. It even includes dividing the data processing pipeline into multiple steps, rather than having a single model output it all at once. Quantitative changes bring about qualitative changes, which is why current LLMs can provide answers so close to humans.

The Curse of Correlation

But here is the problem: does the underlying complex probabilistic association prediction computation, while infinitely aligning with human outputs, also bring a certain degree of limitation? Are human creative associations and thoughts more discontinuous and less traceable than superficial probabilistic association computations? If LLMs just help us filter and find information based on the most probable and strongly correlated patterns in the data, is it easy to overlook novel things that have never been explored, rarely appear in human data, or have never appeared at all?

I call this phenomenon the "Curse of Correlation."

When New Phonetic (Bopomofo) or other predictive input methods emerged, the language used by humans actually underwent a filtering process, becoming more aligned with input method predictions for efficiency, with fewer complex textual variants, rare vocabulary, or newly coined words intended for creative atmospheres.

So, without carefully designed Prompts and prompt words integrating human thoughts, are LLM creations and outputs somewhat more prone to having an "AI taste," following the same logic as the common vocabulary and texts of input methods? Because, essentially, they are both just predicting the single most probable scenario given the current context across all existing data.

The Shadow of Humanity

I believe this is also why LLMs are said to be like the shadow of humanity. Because any pattern or process that is too fixed and observable in the data might be sealed into the models in the future as computing resources increase and models develop. Conversely, you could also say, are most human groups actually just very predictable? (Gets hit)

However, what about things not in the data? In human history, many discoveries were strokes of genius or even happy accidents, but humans have the ability to cleverly achieve sudden enlightenment, connecting seemingly unrelated contexts to create new discoveries. Under the current architecture of strong correlational computation, do LLMs have the ability to replicate this?

An Unfinished Paper Idea

Therefore, my original paper idea was: if I deliberately picked named entities and vocabulary that were very far apart and forced the LLM to output a connection or a reasonable explanation between them, how would the LLM actually react when there is almost no description in existing data?

Because as long as humans reach a certain level of intelligence, they can make up a reasonable story. But for LLMs, from low-end to high-end, you would see outputs in this scenario ranging from gibberish to severely hallucinatory connections, perfectly verifying the gap between LLMs and human capabilities. It's just that being conceptually far apart and almost unrelated is hard to define in computer text data—vectors? Unsearchable by search engines? So this experiment wasn't developed further, and then someone else did it.

Where LLMs Are Strong and Weak

The article has gotten a bit long, so let's get back to the point. I do not believe that within a correlation-calculating model, you can force it to generate creative, rare topics that break away from existing data. Plainly speaking, what an LLM outputs cannot be random noise; it is data fragments with certain patterns.

So in current scientific research systems, LLMs are strong in that, given existing data and set conditions, they can continuously compute through programmatic loops to filter and screen similar ideas, research topics, and implementation methods. Our own exploration process, on the other hand, involves constantly swapping the named entity keywords in our professional knowledge, asking an LLM Agent to search or deduce the most reasonable path.

But an LLM will not proactively prompt you with novel topics that have weak surface correlations and huge differences in existing data; this is currently something humans must actively think of themselves. This is also why it is said that those who are better at asking questions are better able to use AI tools.

I believe that this fundamentally is the ability to constantly explore seemingly unrelated and unprompted different topics. Furthermore, I believe this step is one of the keys for AI research to conduct high-level decision-making or even Artificial General Intelligence; otherwise, the Curse of Correlation will continue to exist.

Based on the equalization of knowledge, everyone can consider giving it a try. But what I want to share might be more of some observations and sharing regarding structural issues. Can it be considered popular science? Some of it might also be a layman's little murmur.

From Q&A to Taking Action: The Next Step for AI Assistants

For models like GPT, Gemini, and Claude, the direction they need to take to evolve from information-answering aids into something more useful and helpful to everyone is to actually get things done for you.

However, digital systems might be much closer to the messy complexity of real life than people imagine.

The Metaphor of a New Employee

For example, suppose a new employee wants to start working at a shop you own. How do you assess their capabilities?

Looking at a resume is a common practice, as is a probationary period, but these are built on the cognition that, given basic human operational capabilities, even if there are things they cannot do in the shop, it won't cause a disaster. For instance, if they don't know how to operate the coffee machine, they will ask proactively or check the manual. If they need to find a specific item and aren't familiar with the shop layout, they might search one by one, and if they really can't find it, they will politely ask the boss when they are not busy.

This threshold of cognition and caution rises with the professional degree of the industry and the economic or time costs involved. To put it plainly, the higher the risk of something going wrong, the less you trust a new employee, especially when you don't fully understand their capabilities yet.

The Dilemma of Authorization: Decision Fatigue vs. Risk

To human cognition, digital systems, at least in my personal feeling, are only flatter, more ephemeral, and less concrete than the physical space of a company. Therefore, for this cognitive threshold, some people will be more cautious and fearful, while others won't care as much.

Frontline AI companies, in order to lower this cognitive threshold and the risk of accidents, have constantly been thinking about authorization, sandboxes, and defense mechanisms. But these mechanisms in digital systems will definitely not be human enough; after all, defining which actions require authorization and which do not is also a very complex issue in the real world.

Imagine if that newly hired employee wanted to pick up a manual on the desk or grab a jar of salt from the condiment area to add some, and they had to get your confirmation for every single step—accumulated over time, this would absolutely lead to decision fatigue. You would feel this employee is too tedious, timid, and afraid of taking responsibility.

Therefore, many people choose to bypass the authorization mechanism entirely, fully delegating all operational rights to these Agents released by the companies. By doing so, the efficiency and output of some people soar to incredible heights; but for others, accidents happen—files are completely deleted, sensitive information left completely exposed, and so on.

This further involves some architectures, habits, and even values of digital system structures.

Differences in Environments: A Tidy Shop vs. a Department Store's Appliance Section

Let's swap roles today; let's try being this new employee.

Right now, you are standing in a shop where cabinets and equipment are neatly arranged, so strict that every device, like the coffee machine, refrigerator, and oven, has an operating manual and a safety rulebook hanging on it. Don't you feel a bit more confident to explore and complete tasks on your own? Because you have a clear understanding of what will happen at each step, or what prompts will appear for erroneous steps.

Then let's make the environment more complex... imagine the appliance section of a department store. You walk in and see the designs of many different new appliances. Even though they look shiny and appealing (I must admit there was a time I loved browsing and playing around with them...), the problem arises: can you guarantee that you know how to operate every device at first glance? For some buttons, when you press them, there is a sense of fear and uncertainty.

In modern visual-interface system environments, it is more like the latter scenario. Each service, manufacturer, and interface design on Windows has its own habits, considerations, and design reasons. But the problem is, through just the initial visual interaction, even the smartest people might need to reason and deduce a bit, let alone ordinary users?

By the same token, for these model assistants, this is not the most efficient mode of functional cognitive exchange.

The Inherent Advantages of Unix-like Systems

Unix-like systems (macOS, Linux...) and pure text Command Line Interfaces (CLI), surprisingly, have an inherent advantage in this regard. They are more like the aforementioned shop where everything is neatly organized.

This relates to the fact that many Unix-like tools and commands have actually been developed over decades (probably more than twenty years?). Well-designed CLI text tools, just like clearly documented manuals and procedural rules, show what each parameter means, and if a step is done wrong, there will be as clear an error message as possible.

The Windows system itself has a lot of baggage in its GUI, while its text command system has undergone generational changes, with the completeness and operability of the previous generation not matching that of Unix-like systems.

There is a very clear trend with tools released by OpenAI and Claude: they are often released first on macOS, or even operate best solely on macOS. Besides macOS natively embedding these complete Unix tools, macOS actually has comprehensive AppleScript that allows the manipulation of GUI interfaces via text commands. That means ordinary operations in Reminders and Calendar can actually be added, deleted, or modified via text commands.

Returning to the cognition of the employee, this means that in this environment, they have a very clear understanding of what will happen at each step, or what items can be manipulated. But in Windows and other environments, they face something akin to a blob of chaos. Many things have to be explored on their own, and even after taking action, what happens is very opaque (software on Windows is developed independently by various companies, and many processes are essentially black boxes, meaning information is not disclosed). The cost of cognitive reasoning is extremely high, and many employees might just give up halfway through exploration.

Solutions from AI Companies

Back to these AI companies themselves, their perspective is hoping more people can use the tools they release as employees. But how can these problems be solved? It's impossible to tell everyone to just use macOS, right? (Though a recent wave really did drive up Mac mini sales...)

To lower the authorization friction mentioned earlier and address this problem, I currently see a few directions for solutions:

1. Building Isolated Environments

Whether it is advanced users doing it themselves or these AI companies providing it. Taking ChatGPT's desktop application as an example, you can notice that its Agent mode establishes an isolated environment and then executes programmatic commands, so whatever happens inside won't affect the outside. At the same time, many tools inherently have a sandbox mode. Additionally, some advanced users, when utilizing tools provided by these AI companies, will build their own isolated environments, such as restricting read/write directories or setting up virtual machines.

2. Brute-Forcing via Visual Recognition and Reasoning

That is, directly taking screenshots or acquiring the positions of operational elements on a web interface to perform actions. In fact, if you are not worried about privacy data issues at all, they have actually realized handing over full control of the phone to cloud AI in China. It can help you send messages, automatically search for information on the internet, and automatically filter necessary information on social media, all fully automated.

But then this leads to another extremely complex issue: where is the boundary of data authorization interaction and privacy?

The Problem of Identity Verification and Trust

Let's change roles again. If today you are an ordinary citizen going into a government agency to handle administrative affairs, how does the government agency verify that you are indeed the citizen in question? The common verification procedure involves ID documents and photo recognition.

But in digital systems, how can this procedure be implemented? It's impossible to just recite your Gmail password and gain direct passage, right? Would you feel safe if every tool could operate your email? Not to mention that passwords and ID numbers leaking on the internet is already a common occurrence.

Therefore, the authorization and authentication of different data access alone is a huge obstacle for these AI tool assistants.

Or stack the roles: if today an employee of your company goes to the Ministry of Economic Affairs to apply for a business alteration, how do they verify that this employee truly represents you?

Let's be a bit conspiratorial: it's not impossible for outsiders to hijack this employee, make them steal your company's certification documents, and forge your signature. So what about current Agent proxy tools? In the digital system, they don't even have emotions, expressions, or other distress signals for you to judge authenticity.

This is also why, in a situation where AI tools are expected to be smarter, information security policies are also taken more seriously and sensitively.

macOS has an inherent advantage in security: every type of operational permission, just like you going to a government department to apply for reading or writing certain personal data, has a complete process and restriction. Windows actually has this too, and it can be adjusted to be very strict, but by default, it is not as strict and organized as macOS. Thus, this is also a factor in why these AI tools are more suitable for operating on the macOS system.

Why AI Tools Have Not Truly Landed in the Hands of Ordinary People

This article has unknowingly become very long; let's wrap it up.

If you sincerely want to make good use of this AI employee, synthesizing the above factors, we find there are many friction factors. This is also why some people are already flying high with it, but it hasn't yet smoothly landed in the hands of ordinary people.

1. Lack of Direct Cognition of What AI Tools Can Do

With newly hired human employees, no matter what, there is a basic reality framework; you can anticipate what they will and will not do. However, for AI tools, most people's cognition remains at Q&A and information gathering. When it involves data alteration and modification, they cannot confidently hand it directly to them.

Additionally, AI tools do not have a tacit understanding of interacting with you like real people do. Part of this tacit understanding is that they will proactively and gradually adjust until you can directly assess the boundaries of their capabilities. AI tools will only wait for you to issue commands and prompt you on what else can be done.

At this point, engineers with programming and system operation knowledge are the best demographic to exert this capability, because the things engineers know that programs and text commands can achieve are the partial upper limit of AI tools. Meanwhile, the more powerful the model, the better it is at self-orchestrating the planning sequence and execution steps of these programs, which means it can better translate your ambiguous commands into concrete procedural steps.

2. Lack of Background Knowledge on How to Correctly Lower Friction

To bypass restrictions and correctly unleash AI tools, one must establish complete data boundaries or prepare relevant protection and backup measures. However, these abilities are primarily embodied by engineers familiar with digital systems.

You want ordinary people to prepare sandboxes and isolated environments? You want them to understand regular backups? You throw a bunch of jargon like git, docker, VM, WSL... I see most people getting dizzy right away. Even the author hesitated here, wondering if writing about Unix earlier was already too much.

Furthermore, the friction of digital authorization itself is an obstacle. Under increasingly strict information security precautions, if I want to allow AI tools to operate my digital assets scattered across different cloud services, such as Gmail, Google Calendar, or other services, just exploring whether the access interface has a complete authentication mechanism is a cost in time and energy, much like those appliances lined up in a department store.

3. Insufficient Control over Personal Privacy Data, or Trust Issues with AI Service Providers

Basically, when you start using AI tools to process these documents, you have to accept that the information contained within will absolutely be uploaded to cloud servers and might be processed and utilized by the companies. For instance, my ID photos or my bank account plain text, as long as I tell it to organize them, will definitely be uploaded.

When you have even the slightest bit of unease, yet cannot anonymize (de-identify) your own data, you won't be willing to use AI tools. This is also why there is a faction hoping to develop local AI.

Returning to Codex for Windows

Returning to the beginning of the article, although the Windows version of the Codex tool has been released, and I encourage everyone to explore it, rival Claude probably already has similar tools.

At the same time, my tone isn't very enthusiastic. Why?

Because of the problems discussed above, and the inherent aptitude of Windows. I even saw that this Codex desktop tool runs via the WSL tool, which means setting up an isolated Linux virtual machine within Windows. But based on my understanding of Windows running into problems every other day (the author hasn't used Windows for serious business in a long time; the last time was using a Citizen Digital Certificate), I feel its usage rate will definitely not reach the levels on macOS (don't come to me if things go wrong, just passing the buck X).

So this piece is probably more like a layman's popular science murmur article.

I saw a fellow patient crying.
Because they wanted to share, to acknowledge their vulnerability, and to disarm everyone's defenses,
They bared their soul nakedly,
And wept.

In that instant, a soft spot in my heart was touched.
But it remained a sigh of, "Ah, what can be done anyway?"

What truly ignited the fire happened later.
Someone used AI tools, far too habitually.
And so, an unproofread, bullet-point transcript
Was casually shared out just like that.

"So, it turns out our experiences can just be reduced to a few words of summary?"
"Is the suffering of patients nothing more than one little creative blurb after another that glides past people's eyeballs every day?"
"Does the outcry of what has been taken away, of what can never be possessed again, lose all meaning under arbitrary transcription?"

I profoundly realized why some people harbor such a hatred for AI summaries.
Because the creator's intentions, what they wish to convey,
Can all be lost in the nonchalance of a transcription,
Just as we are habitually ignored amidst the rolling tides of this world.

I am not as brave as this girl, for even now, I lack the courage to acknowledge my vulnerability.
I still arm myself with numbness and rationality.

But all of this must mean something.
It is quite ridiculous; I had almost adopted
A mindset of just getting by, of muddling through my own life, and letting things pass as they may.
Yet, for that crying voice and figure,
I cannot swallow this grievance.

But the focus of this article is not on detailing those events, but rather on examining what constitutes true change and growth.

The Interaction Between Health and Psychology

In the first half of 2023, my mind was entirely focused on the doctoral qualifying exam and finding a stable income to get by. Looking back now, with a weak body and mind, there wasn't really much growth.

Starting in the second half of the year, following my original plan, I started going to the gym and hired a coach to supervise my weight training. Fortunately, there are gyms within five hundred meters of both my workplace and my home.

During my fitness training, the thing the coach most frequently singled out to supervise was the impulsiveness in my movements.

For every movement, I just desperately wanted to complete the specified number of reps and quickly maneuver the equipment to the designated position, but the completeness and correctness of the movement were often neglected.

So the degree to which each movement actually trained the muscle was compromised; a rushed finishing motion also hurts the joints and muscles, which will accumulate over time and turn into long-term sports injuries.

But the "impulsiveness" in these movements also feels related to the interaction between my long-standing psychological state and my upbringing environment.

When interacting with the outside world, hearing loss—coupled with the delay in language processing speed training due to not wearing hearing aids early on—makes my reactions a beat slower.

This anxiety of failing to keep up with the outside world's reactions is already deeply engraved in my bones, and my gradually deteriorating visual field, lacking proper psychological cognitive construction,

only deepens this further.

Before starting the fitness plan, I was vaguely aware of this. And during the execution of fitness movements, through feedback from others and the most direct feelings of my body,

I became even more determined that I must improve looking "impulsive."

Focusing on the completeness of every movement during fitness training is also a mirror for myself.

How to breathe properly, how to stretch and tighten the muscles well, and how to properly control the position of the equipment—

besides the physical training itself, these are also rehabilitation for a soul that was once overwhelmed to the point of breathlessness.

If there has to be a conclusion, it is that

I still have to work hard, but the true goal of my efforts,

Is not pursuing perfect external reaction speeds,

But pursuing the calmness, composure, and leeway coordinated jointly by the body and mind.

Sleep

My sleep is an absolute disaster.

The uncontrollability of my sleep state is also likely the fundamental reason why I had to leave the hospital.

Because I cannot guarantee that my mind will remain completely clear during working hours.

No matter how I adjust, no matter what work situation I am in (unemployed, working full-time, or working part-time during the week),

my mental state in the morning is invariably worse; it feels as if something hasn't completely woken up.

I've gotten used to forcing myself into a working state with a cup of black coffee.

And if I am busy all week, there will definitely be one day where I need to sleep until noon just for the mental fatigue to ease.

It is indeed related to sleep quality; if the previous night's sleep quality is very good, the morning grogginess improves.

But it seems a night of perfect sleep quality is unattainable.

Currently known alleviating factors are:

1. Strong sunlight exposure in the morning.
2. Avoiding caffeine intake in the afternoon, or only drinking weak green tea.
3. A cooler ambient temperature when falling asleep.
4. Small doses of SSRIs and BZDs seem to offer some improvement.

The hospital environment is cold. Under the old air conditioner at home and a house rule that the AC is only turned on when it exceeds 27 degrees Celsius,

I spent at least two years working in cold temperatures during the day and returning to high temperatures at night. This may have worsened my sleep quality.

During the preparation for the doctoral qualifying exam, this severe groggy state did not completely disappear despite leaving hospital work.

After passing the qualifying exam and starting the fitness plan, there was some improvement, but not as much as I had expected.

I briefly went to a psychiatric clinic and took SSRI and BZD medications, feeling some improvement, but when I got too busy and forgot to return for follow-ups, I didn't get refills.

I suspect that the genetic defect in USH2A might have effects that are currently unknown on melatonin or the pineal gland.

Furthermore, an almost completely degraded retina is indeed discussed in literature as being associated with the regulation of circadian rhythms.

So... there really doesn't seem to be much I can do to improve my sleep quality... tragic.

Now, I can only take things one step at a time.

At least, through over a year of recuperation, I can naturally wake up around 8 or 9 AM.

Family

The meaning of family to me is actually very complex; to me, it is a mire I cannot break free from.

On one hand, I believe that for a person to be mentally independent, they should move out from their parents and live on their own.

Even renting in another city and working alone would be fine, so that my social life and privacy can have their own space.

On the other hand, my mother seems extremely unwilling to see all her children leave, even if she doesn't say it explicitly.

But from the emotions she arbitrarily spills and vents, I read a fear of being abandoned and having no one to accompany her.

And so I became the only child left at home.

And indeed, when crises occurred at home a few times, they were handled promptly by me staying at home and my sister who was nearby.

In the closed storm of the family, the only thing I can do is perhaps to earnestly cultivate myself,

in order to maintain the autonomy and stability of my emotions and judgment.

The Meaning of Life

Currently, I do not have a strong sense of identity.

There are no values I absolutely must protect, nor life goals I absolutely must achieve.

The obsession I fought tooth and nail for in the past—looking back, who was it for? Who instilled this obsession?

Was it the product of many values pursuing an external image, self-hypnotized heroism, and pandering to multiple expectations?

Was it a constant amalgamation of the images of all the successful people I had seen?

Also, are many of my implicit emotions

derived from the relative sense of deprivation regarding the life experiences of many normal people?

If I see clearly that no matter how hard I try, these life experience opportunities—lost as my eyes gradually go blind—cannot be recalled,

then what was the meaning of the effort driven by that burst of energy in the first place?

I still don't know the answer to this question.

Now, pushing forward the Taiwan Usher Syndrome and Audiovisual Impairment Association is what the professor who originally helped me find the pathogenic gene mutation

came over to pull me into.

She said she couldn't bear to watch my eyes continue to worsen,

and there are indeed many areas in biomedical science, drug R&D, and even IT accessibility

where efforts can be made.

But this thing called my future has never had a clear bottom.

Many things I previously tried so hard to grasp would slip away through my hands like sand.

Yet during this period of rest and recuperation, there are many connections that I inadvertently run into just around the corner.

My New Year's outlook for 2024, my current wish,

is to continue these good connections, keep looking forward, and wait to see what else lies ahead.

At 1:00 PM, the opening address at the first session of the first general assembly of the association in Nangang, Taipei.

Conceived on the high-speed rail that very morning,

It includes some responses to phenomena observed by the co-founding professor.

I am very glad to see so many people here today, participating and offering their support.

At its inception, this association was based solely on the wishes of a few individuals.

Some wanted to help those whose vision and hearing are both affected.

Some hoped that what they study and continuously research could benefit more people in need.

And some wanted their plight to be seen, hoping that this endlessly turning world could take care of them a little more.

Under various circumstances,

We began to try, taking it one step at a time, slowly gathering

People in similar situations or with similar ideals.

From the middle of last year to the beginning of this year,

From Facebook groups to Line groups,

We were pleasantly surprised to find,

That our philosophy was receiving more and more responses,

Our actions were gaining more and more attention,

And today, we are just a little bit away from accomplishing

That small step of our initial wish.

Based on my own past experiences, I personally believe

That in the real world, nothing is certain.

Effort does not necessarily yield results. Dreams cannot always be realized. Under the turning of society, not the whole world can hear the sound of our cries.

For those whose vision and hearing are affected, Illness and defects are not necessarily a blessing or a curse. Role models, paragons, striving upwards, expecting to be of use— These are also not necessarily our labels.

I think what we should value the most Is how we become "human." How do we fight for our rights as humans? Do we also have the capacity for free choice? With free will, living well according to our own desires, Instead of entirely relying on the pity and expectations of others As the meaning of our existence.

The association cannot promise that anything will definitely change, For the whole society, the whole world.

But I hope that in this process of striving, Everyone can discover something together and make themselves better.

Thank you all.

• Between July and August 2022, at my former resident training hospital, I was evaluated as unfit, making it very difficult to continue.
• I was given a chance to improve, but the truth is that both my body and mind had reached their limits at that time, even feeling as though my life was hanging by a thread.
• Looking back now, I believe that under the same environment, I absolutely had no remaining energy to think about how to improve.
• I decided that no matter what, whether pursuing further academic studies or continuing to work, I had to leave first, take a breath, and only then would it be possible to rebuild from scratch.

The Doctoral Qualifying Exam and the Challenge of Restoring Physical and Mental Condition

• After leaving the hospital, the immediate challenge I faced was the doctoral qualifying exam.
• Because I was so busy at the hospital that I forgot to apply for a leave of absence, I was delayed by two years, meaning I was already a third-year doctoral student when I went back to study.
• The university regulations stated that the doctoral qualifying exam had to be passed within three years, and three subjects had to be passed. This meant that in the upcoming third year, I only had one year left to prepare, with only two chances remaining: the fall semester and the spring semester.
• Another trouble was that my psychological state was stuck in negativity; at all times, I felt a sense of powerlessness that no matter what I did or what effort I put in, there would be no reward.
• Therefore, the immediate challenges I faced right after leaving the hospital were:
  • Physical and mental exhaustion, a sense of powerlessness toward everything.
  • A broken physical fitness and broken sleep.
  • The loneliness, anxiety, and fear of stepping into an unfamiliar field alone (my undergraduate major was medicine, and I entered the computer science doctoral program with almost zero foundation).
  • A sense of uncertainty about where to go and having no direction for my career.

Progress of Improvement

• Acknowledging the current situation, separating subjective emotions and objective problems as much as possible.
• Treating the psychological and physical states as part of a systemic problem, and listing countermeasures.
• Analyzing the problems, identifying the scope of what can be controlled, and slowly beginning to improve from the bare minimum.
• Searching for usable paradigms and methods from past psychological counseling experiences, spiritual self-help, and popular science books.

Results

• Teased apart the impact of the habits and mindsets of my family of origin on my body and mind, and found ways to address them one by one.
• Established a list of scientific attributions for chronic fatigue and poor sleep quality, alleviating the anxiety of self-doubt.
• Starting from the most basic act of sitting alone in a cafe, to traveling alone, and reading alone, gradually rebuilding and strengthening my self-awareness and value in every exploratory journey.

Phased Tasks Accomplished

• Spent nearly a year passing the doctoral qualifying exam; unexpectedly failed biostatistics the first time, and spent half a year passing it with fear and trepidation.
• Briefly revised a paper I had written before, establishing some basic workflows for writing papers.
• Reviewed basic calculus and took courses to acquire fundamental concepts in linear algebra and vector calculus.
• Used the Flutter framework to write a cross-platform App capable of audio recording, speech recognition, and processing text using a GPT model.
• I explored career directions outside the hospital, built on my medical license — no specialist certification, but two years at a large hospital — and sketched out possible development paths.

Phased Tasks Yet to be Completed

• Exercise Habits for Strengthening the Body:
  • Currently, I only do basic cardiovascular spinning and social badminton.
  • Targeted physical strengthening pathways are needed, including scientific training goals for cardiovascular endurance and muscle strength.
• Habits of Thinking, Summarizing, and Long-Term Repetitive Note-Taking:
  • My ability to gather themes and elaborate is too weak.
  • My old writing habits are too long-winded and not concise enough, failing to convey the main points.
  • Not focused enough, easily having my attention consumed by fragmented information.
  • The threshold is set too high, resulting in articles constantly failing to be finished.
• Time Tracking and Management Techniques:
  • The framework established a long time ago is too cumbersome and not concise enough.
  • It was temporarily abandoned after leaving the hospital to accommodate physical recovery.
  • I should slowly compress time utilization to the extreme to improve productivity.
• Training and Improving Social Skills:
  • To be frank, the environment of my family of origin is not a good environment for practicing expression.
  • It is not easy to build relationships with strangers.
  • If I need to rely on being a physician as a basic livelihood, a certain level of social communication skills is still required.
  • If I want to move towards a consultant, manager, or PM role, clear and powerful expressive abilities are still needed.
  • In fact, there is a record of perfect performance when physical fitness is intact, the brain is supportive, and sleep is sufficient.
  • Therefore, part of the improvement plan should be merged with the aforementioned physical training habits.

Upcoming Specific Goals

• Find a gym and set scientific goals to improve physical fitness by at least fifty percent to cope with increased challenges.
• Complete an online basic course in Python data science.
• After passing the qualifying exam, take two to three departmental courses, which must include NLP.
• While working part-time at a clinic, familiarize myself with and enrich my knowledge of lasers, medical injectable fillers, skin diseases, and chronic disease medication management, and absorb clinic management experience.
• Find research topics and generate at least two to three papers within a year.
• Find practical topics and write Flutter mobile applications to embody the value of my own medical information integration.
• Prepare the Usher Syndrome Association with partners.

Honestly, I am somewhat tied to my family of origin. I am the youngest of three children in my family; my brother and sister are over a decade older than me. Now I am the only one left in our hometown. Otherwise, after leaving the hospital and cooling down, discussing with others makes me feel that going to a medium-to-small hospital in another city that isn't as tiring, and getting a family medicine specialist certification, might still be an opportunity.

I am about to start working part-time at a clinic, which means I have to face the inner demon of being told I am not suited to be a doctor once again.

Perhaps I truly lack the extra capacity to become an extremely professional, top-tier doctor,

but I know that some strategies I have worked on are effective and can be carried forward, at least enough to slightly fade

the barrier between myself and this world.

💡

TIP

Attempting to analyze from systemic causes why I give the impression of lacking composure, and trying to break them down step-by-step for improvement.

Background

Extremely Narrow Field of Vision

• I can only focus on one person at a time; I cannot respond in real-time to the reactions of multiple subjects.
• Even with the same subject, I can only focus on their facial expressions; other body language might be missed.
• During dynamic activities, I cannot observe the interaction of things in three-dimensional space, making it difficult to step in to help or react immediately.

Hearing Loss

• More loss in mid-to-high frequencies.
• There is a chance I may miss what other people say.
• Once the initiation of a conversation fails to establish, the resulting communication hindrance has a certain probability of impeding further social interaction.

Background of Upbringing

• Authoritarian discipline that comprehensively overrides independent opinions.
  • Being a yes-man.
  • Lacking self-confidence and the ability to think critically.
• Poor feedback from the surrounding environment leads to uncertainty about interactions.

Psychological Factors

• Subconsciously believing that all unsuccessful negotiations are caused by my own flaws.
• Setting an image that can perfectly respond to all communication and social expectations, which instead makes it easier to get stuck.
• Overly demanding state; exerting full effort and extra attention to meet the standards of a normal person paradoxically increases anxiety.
• The interaction between environmental anxiety and pressure easily causes a negative feedback loop.

Outward Manifestations

Speaking

• Pitch easily rises and becomes high-pitched.
• My pace of speech easily quickens.
• Volume easily increases involuntarily.
• A habit of speaking while staring intensely (placing all attention on the other person's facial responses).
  • Similar to how a cat receiving excessive attention from a stranger triggers a natural avoidance response.
• All the above can give people a sense of extra oppression, or even create unnecessary hostility.

Spoken Content

• Imperative and ordering sentences learned from the family environment easily make people unhappy.
• Habitually negative opening particles also hinder the unfolding of normal communication.
• Too many "I feel..." and "I think..." easily increase the uncertainty of the speech, reducing its persuasiveness to others.
• Many redundant and superfluous words stem from an extreme fearful thought of being unsure if the other person understands.

Body Language

• In order to compensate for visual field loss, increasing the frequency of eye scanning and head-turning movements gives an impression of anxiety.
• Hunched, withdrawn body movements tend to suggest submission rather than dominance.

Countermeasures

• Practice slowing down speech speed, and modulating voice and tone.
• Self-imagine the sincere voice of a senior elder, and use his image to shape my statements.
• Mental preparation: practice perceiving communication barriers and anxiety reactions caused by heightened attention, to stabilize my overall demeanor.
• Remember that it is not true that there are no external means to counter not hearing clearly; adjusting the frequency of a hearing aid or politely asking the other person to speak louder are viable means, rather than just stubbornly struggling in silence.
• Moderately lock eye focus at a slight distance away from facial expressions, maintaining an impression that you are paying attention to them while possessing the poised elegance of a little distance.
• Reduce the obsessive thought of hoping the other person must absolutely understand and that communication must be flawless. Be more confident in your own words, and let the rhythm of the conversation unfold slowly.

Role Model: The King's Speech

• An unfavored heir to the throne underwent a period of arduous training to treat his stutter.
• At a critical moment, he used a sonorous and powerful tone to inspire the entire nation.
• The meaning was not as simple as stutter correction; it also represented the establishment and certainty of self, self-esteem, personality, and core values.
• Constantly thinking, refining, and polishing these, and finally radiating them outward with confidence.

Prologue

• Spoilers ahead.
• The worldview of Fate/Grand Order continues the settings from the Fate/stay night games, anime, and related works:
  • In this worldview, "Masters" can perform summoning contracts to summon Heroic Spirits to fight for them.
  • Heroic Spirits originate from the "Throne of Heroes" and could be warriors with glorious ancient battle records, emperors who conquered far and wide, or gods and magi existing in myths and legends. Later, this extended to well-known modern creators and scientists who deeply impacted history.
  • Different facets of a Heroic Spirit, different specializations of their abilities, or different anecdotal stories can manifest as different personas recorded on the Throne of Heroes, allowing them to be summoned in different classes. Even their betrayed, darkened, or vengeance-ridden sides can be specially summoned as Anti-Heroes.
  • Originally, Fate/stay night was set merely as a summoning system in a single urban area in the Far East (Japan), aimed at completing a magical ritual through the mutual struggle of seven Heroic Spirits. Once completed, the Master could use the Holy Grail, filled with the magical energy of defeated Heroic Spirits, to grant an "omnipotent wish."
  • Fate/Grand Order further extends this summoning system, where players act as Masters summoning Heroic Spirits to resolve world crises, traveling across different worlds to meet various Heroic Spirit characters.

The Fairy Round Table Domain

Brief Story Background

• The Fairy Round Table Domain is a newly opened hypothetical world in the Taiwanese version of Fate/Grand Order.
• This hypothetical world can be seen as a timeline branch distinct from the current world.
• In this worldview, the setting borrows from the famous British legend of King Arthur, but incorporates the story writer's own interpretation:
• The original mythical Morgan in the Arthurian legend has a three-fold nature:
  • An early fairy of the lake who was helpful and kind to humans.
  • The witch within the Arthurian legends after Christianization, depicted with various malicious descriptions.
  • Yet sometimes she possesses a protective nature toward King Arthur, guiding him to the legendary utopia, Avalon.
• In the story, the background established by the author is that Morgan was sent from the utopian paradise Avalon on the reverse side of the planet to the island of Britain to save it.
• After a major event that caused the timeline to branch, the land of Britain disappeared, and the six ancestors of the fairies committed the sin of deicide to create their own living space, resulting in a curse.
• Aside from the original curse, the true nature of fairies is to act entirely on momentary impulses, likes, and dislikes, possessing almost none of the moral ethics of modern humans.
• The sinful fairies continued to reproduce, but the catastrophes born from the curse and the warring nature of the fairies eventually led to the destruction of the British land that had lasted for a millennium, reducing it to a wasteland. This was Morgan's first failure.
• The intervention of the Heroic Spirit summoning system in the story caused the Morgan from Pan-Human History to accidentally intervene, transmitting information back to the Morgan who failed in the past, giving the Morgan of this branching hypothetical world a chance to save Britain thousands of years ago.

The Savior Era

• In this branched world, she served as a savior, saving Britain an unknown number of times and resolving countless calamities.
• However, a fairy from the paradise of Avalon was instinctively despised by the native fairies of Britain because her existence reminded them of their obligation to atone and repent.
• Every time she dedicated herself to saving Britain and the fairies, resolving disasters, she was met with imprisonment, water dungeons, burning at the stake, rumors, exile, betrayal, and assassination—all kinds of things, without the slightest gratitude or any sense of atonement or repentance. This continued for nearly a millennium.
• In the final instance of faithfully playing her role, she mediated a war alongside her companions, established an ideal political system between humans and fairy clans, and planned to enter the holy hall of marriage to be crowned queen alongside her lover, the human knight Uther. The era of peace she had anticipated for a millennium seemed finally to have arrived, and all her efforts appeared destined for a happy and perfect conclusion.
• Yet, on the very day of her coronation, she witnessed Uther being poisoned to death with her own eyes. The fairies rebelled once more, the savior was framed with unwarranted crimes, and everything once again came to nothing.

I will never save any fairy again. Every single fairy in this Britain is unworthy of being saved.

Morgan, the Winter Queen

• No longer trusting anyone, her myriad ideals turned into coldness.
• The dream of saving Britain transformed into an obsession to establish and maintain a beautiful British land (changing from Ruler class to Berserker class).
• Morgan directly ruled as a sovereign. After letting catastrophes wipe out the fairies, because the continuation of the kingdom still required the existence of fairy subjects, she directly used grand magic equivalent to a summoning spell—a ritual akin to the Holy Grail Heroic Spirit summoning contract—to forcibly summon all the fairies back.
• She implemented a tyrannical rule, which surprisingly and effectively maintained the beautiful Kingdom of Britain for over two thousand years.
• Although, as the queen of a hypothetical world, she was an enemy to the protagonist's party, easily crushing the players with her overwhelming magical prowess in the final battle.
• Yet, she still died by the malice of fairies. While Morgan was distracted, they threatened her with her most beloved adopted daughter, inciting her most loyal subordinate to backstab her, ultimately resulting in her dying completely disfigured.
• In the Queen's reminiscence, a single panel showed that Morgan loved the land of Britain so deeply that if visitors from the outside world ever arrived and sincerely praised the beautiful splendor of the kingdom she maintained, Morgan would be incredibly happy.
• Nasu, the game's scenario writer, truly has no human heart.

How is my country? Is it a beautiful kingdom? Is it a dreamlike kingdom? If you think so, I will feel the utmost joy. Welcome to the Fairy Kingdom of Britain, guests. I only hope that this scenery, Will remain forever in your memories.

Resonance

• When I played through this part of the story, I happened to be in a somewhat extreme state of mind.
• I had fought for a long time for the ideals and family members I once deeply loved, only to find that many things casually said were nothing more than momentary likes or dislikes.
• There are too many things that cannot be changed, especially the most difficult things to change: the human heart and ingrained behavioral patterns. Communication paths took countless twists and turns; I put in immense effort for a long time, but nothing ever took effect.
• When I saw Morgan's millennia of exhaustion, her ultimate breakdown, and the corresponding cold reversal, it felt as though it mirrored my own state of mind. The reason I no longer want to bother, the reason I no longer want to spend the effort to save things, is precisely because having the luxury to think about saving things only leads to falling down again and again.
• I longed to possess the kind of dominating, ruling power that Morgan, the Winter Queen, eventually obtained.
• The reflection of these emotions made Morgan the character in Fate/GO with whom I have resonated the deepest in its history. Previously, I might have rolled for characters just for their pretty artwork or strength, but this time I was determined to roll until I got her maxed out. Of course, it could also just be that the cold, aloof onee-san archetype happens to be exactly my type.

Origins

Nothing much, I just wanted to splurge (X The impulse to buy arose because I saw a discussion group mention that Apple had added a feature to adjust audio based on an audiogram in the "Accessibility" options for their own headphones.

Previously, when purchasing or looking for headphones, I always had to consider the limitations of wearing hearing aids (I have a mild-to-moderate hearing impairment), so I could only choose over-ear headphones. Without wearing hearing aids, using in-ear headphones has the following disadvantages:

1. Uneven frequency loss causes the heard sounds to distort.
2. To compensate for the sounds I can't hear clearly, I turn the overall volume up too loud, running the risk of damaging my hearing.

However, over-ear headphones are somewhat bulky and easily get hot and stuffy in the summer; secondly, they sometimes easily interfere with the hearing aids themselves.

This time, hearing that AirPods Pro could be adjusted based on an audiogram, and seeing foreign articles mentioning that in Transparency mode, after audiogram adjustment, the normal conversation ability of people with mild-to-moderate hearing impairment is on par with professional hearing aids, I wanted to buy them and try them out. I didn't expect that trying them would feel like discovering many new worlds.

Adjustment Mode Based on Audiogram

• Go to Headphone Accommodations > Audiogram, and follow the instructions to enter the decibel numbers from previous hearing tests.
• However, parts of my audiogram are missing, so I had to guess and enter more conservative data.
• Once filled in, it applies to Transparency mode, media, and phone calls.

Scenario One: Daily Focused Work and Noise Cancellation

Before

- Previously, when using over-ear headphones, the situation was awkward. When wearing hearing aids,
- even using the over-ear headphones' own noise cancellation mode, due to the characteristics of the hearing aids, the leaking sound would still be amplified, making the effect less comprehensive.
- If I wanted to block out sound completely, I had to take off the hearing aids. But if someone wanted to talk to me while I was focused on work, I had to go through a series of taking them off and putting them back on, which was somewhat troublesome.

After Using AirPods Pro 2

- Now, when focusing on work or studying, I use AirPods Pro directly instead of hearing aids.
- The noise cancellation effect is excellent, perfectly shielding the background chatting noise in a cafe at its loudest, achieving absolute silence.
- Transparency mode allows normal conversations; general daily conversations are no problem, making it suitable for dealing with occasional communication events.
- They are also very suitable for temporary online meetings. Because AirPods Pro 2 integrate seamlessly with Apple ecosystem products, the sound is very clear when using Google Meet and Webex paired with the audiogram enhancement.
  - Also, because Apple uses iCloud to unify Bluetooth pairing information across devices, I can switch between computer and phone at any time. I can keep a meeting going on my phone and quickly step away to the restroom without dropping the call.

Scenario Two: Exercise

• This was a scenario I didn't think of before buying, but it is surprisingly practical.
• Existing hearing aids are not waterproof or sweatproof. Once the weather is humid and stuffy or there is too much sweat, they easily get damp and emit abnormal sounds, which also affects the lifespan of the parts.
• There are special waterproof hearing aids, but they are expensive.
• But I still want to keep up my usual social and exercise routines.
• Because AirPods Pro 2 have a certain degree of water and sweat resistance, combined with the audiogram enhancement, I can maintain minimum daily communication during exercise.
• I now use them temporarily at regular badminton gatherings, making them effectively a second, cheap, temporary pair of hearing aids.
  • The danger is that if they drop during high-intensity movements and get stepped on, they're done for.

Scenario Three: Audio/Video Entertainment

• AirPods Pro 2 support Spatial Audio. Paired with online streaming like Netflix and Apple TV, the 3D immersion is very strong.
• The sound information provided by existing hearing aids fails to clearly convey the directionality of sounds coming from speakers, but wearing AirPods Pro makes it very clear.

Scenario Four: Sleep Use

• This is also a situation I hadn't considered before buying.
• While observing factors affecting sleep, I guessed that the low-frequency noise of the air conditioner might be affecting the peace needed to fall asleep, because I previously thought that having some hearing loss meant it shouldn't have an effect.
• I could use earplugs, but then I fear the alarm wouldn't wake me up. I probably could buy a vibrating electronic alarm clock to use.
• I wanted to try and see if the AirPods Pro's noise cancellation mode helped -> It was effective; after the low-frequency AC noise disappeared, it was easier to fall asleep.
• I planned to pair them with phone alarm reminders -> The reality is that the headphones' battery life cannot last through normal sleep time when noise cancellation mode is on.

Battery Consumption

• Under normal heavy usage, they might need to be put back in the charging case after four to five hours of continuous use to recharge.
• The charging case charges very quickly. I haven't timed how fast it takes to fully charge, but under the condition of an urgent meeting, charging for ten-plus minutes can last for roughly another two-hour meeting.
• With a fully charged case, they can be used about two to three times a day.

Contraindications

• Usage is prohibited in environments requiring high observation and attention, such as traffic.
• I once tried them in the dead of night when traffic was light. The conclusion is that although AirPods Pro 2 are excellent, the range of their microphone's sound pickup is not as good as professional hearing aids; environmental sounds are noticeably reduced, making it very dangerous to use them under these conditions.

• Purpose: To stop myself from impulse buying and establish a quantified basis for assessing purchases.
• Necessity Assessment: Prioritize, or see if it fits multiple criteria simultaneously:
  1. Is it immediately needed for work, research, or study?
  2. Health-related.
  3. Related to skill exploration.
  4. Entertainment or interpersonal relations.
  5. Lifestyle decoration or aesthetics.
  6. Just wanting to explore a new use or function?
• Do I already own a similar item?
  • Is the wear and tear of the old item to the point where it is about to break?
  • Are the existing functions of the old item no longer sufficient?
  • Will replacing it with a new item increase life efficiency?
• Prioritize items that have been on the list for a longer time and have undergone careful planning.
  • Items not on the list are just a spur-of-the-moment whim.
• Price Range:
  • Set a basic spending limit for a single month.
  • For items with a unit price exceeding a certain value, do not concentrate multiple purchases in one month.
  • If it is a planned purchase and the unit price exceeds 10,000 (NTD), make good use of installment payments to spread out the cost.

• Replaced the soft router hardware and updated my notes accordingly.
• This post is a continuation of Proxmox VE + PfSense 安裝.
  • The network installation architecture within Proxmox VE remains similar, utilizing the paravirtualized network interface controller (Virtio Net).
  • Testing showed that the Intel J4125 paired with the Intel i225v network card can achieve the full 300M / 100M speeds without configuring PCIe passthrough.
• The switch to OPNsense was made with the following considerations:
  • PfSense requires a subscription (PfSense Plus) for more aggressive updates. Although it's currently free for home users, future charges cannot be ruled out.
  • OPNsense has a more proactive update strategy and a shorter security update cycle.
  • Zenarmor and other third-party packages can be installed, though I ended up removing Zenarmor later on.
  • The OPNsense UI is more user-friendly, allowing quick navigation to desired settings via search.

Installation

• Assign the Port vmbr0 to WAN and vmbr1 to LAN (Refer to the images in the PfSense notes)
• Skip the lagg and vlan configuration
• Default account and password for enter installation
  • account: installer
  • pass: opnsense
• VM settings in Proxmox VE
  • machine type: Q35
  • processors type: host
  • OS type: other
  • Adjust other CPU and memory settings according to needs.

Configuration

PPPoE settings with IPv6 (Applicable for Hinet)

• Fill out PPPoE info to establish connection
• Tick "Use IPv4 connectivity option"
• at LAN IPv6 section, IPv6 configuration type -> Track Interface, and the options below interface select "WAN"
• In the Firewall options, Tick "allow IPv6"
• Add IPv6 ICMP allow rule in WAN firewall rule
• If you want to use a self-hosted DNS (Adguard Home or Pi-Hole) and apply it to IPv6:
  • Tick the "Manual Router announcement management"
  • Fill the DNS IPv4 settings in the "Router announcement"
  • Hinet IPv6 is Stateless (Stateless DHCPv6 + SLAAC)
  • Disable the DHCPv6 service in the LAN
  • The logic here is that IPv6 DNS requests can simply be sent to the IPv4 address DNS server, which will still return IPv6 resolution addresses.

Security

System

• Disable the listen service including WebUI, ssh, Unbound on WAN surface
• Install the CrowdSec, and enable Intrusion Detection
  • Disable hardware net acceleration related "Interfaces" > "Settings"
• Configure the SSH Key, disable the password login

Intrusion Detection (Suricata) (IPS/IDS)

• Download rule sets based on service used
• Rule set with using sites name (p2p, Facebook, Youtube) do not apply
• ET Pro rule set need suscription

Crowdesc

• Connected to the cloud database to detect the attackers IPs and block
• Collection for different scenarios (windows, nginx, ...)can only be added through shell command
• The hub for adding the scenario rule Hub |

Firehol IP list subscription

• FireHOL Block List ( Botnets, Attacks, Malware....)
• Follow the guide to add alias of Firehol level 2
• Add the Cron tab to update the Firewall alias

VLAN Configuration (For creating a guest network or dedicated IoT network)

• Add VLAN, assign Tag, and make Proxmox VE vtnet aware vlan
• Assign DHCP server
• Add Firewall rule to make the VLAN network unable to access the LAN

GeoIP and Ailases for Firewall Block (If you want to block specific countries/regions)

• Register the Maxmind GeoIP database
• Follow the guide to add Firewall aliases
• Configure to block the specific countries

Cron (Remember to set up updates for each list after security configuration)

• System and packages update
• Suricata blocklist update
• Firewall aliases updates (FireHol, GeoIP)

Others Packages

• Netdata: Another monitoring service
• Wake-On-LAN: For waking up machines remotely
• UPNP: Use with caution if you have security concerns; used for punching holes for services within the LAN
• Tailscale OPNsense安装配置Tailscale | 鐵血男兒的BLOG
• Wireguard [How to Set Up WireGuard in OPNsense in 2023 - WunderTech](https://www.wundertech.net/

• Due to eye conditions, I started seeking a special display mode with white text on a black background (Dark mode or night mode).
• I believe it puts less strain on the eyes, making long working hours less tiring.
• Some research: Effects of Dark Mode on Visual Fatigue and Acuity in Optical See-Through Head-Mounted Displays
  • Improves Visual Acuity
  • Reduces Visual Fatigue
  • Improves Usability and Preference in Dark Environments
• Some also believe that Dark mode makes content harder to read, thus increasing the strain on the eyes.
• It is recommended to consider one's own condition; there are also practices of automatically switching based on sunrise and sunset times.

Direction

Operable Targets

• Monitors:
  • Some monitors have built-in grayscale or e-paper modes; it's not white text on black, but it's an option to consider.
  • Purchasing an E-ink monitor directly, but screen refresh rates and high costs are issues.
• Operating Systems:
  • Native dark UI system themes are provided by various OS vendors including iOS and Android, though older versions might not support them.
  • Accessibility features: High contrast themes in Windows, invert colors in the Apple ecosystem, which forcefully inverts everything to black and white but might affect application display, so use with caution.
• Software:
  • Software on computers or apps on phones: Some have built-in Dark mode, which will either detect the system theme to switch automatically or require the user to enable it manually.
  • The impact is limited to the interface or document content.
  • Code editors: Generally support Dark mode and have different themes to switch between. If they don't provide it, please abandon them 😛.
  • Office Word as an example: Initially provided Dark mode for the interface only, while the document itself remained glaringly white. Later newer versions began to provide document viewing and editing with white text on a black background.
  • PDF viewing software: For example, PDF Expert provides Sepia and Night themes, which can be applied to the PDF document itself for a more comfortable reading experience.
  • Websites:
    • CSS Media query standard supports detecting whether Dark mode is enabled in the system theme.
    • Well-known websites also provide Dark mode toggles, such as Facebook, Twitter, Google Search...
    • Not all websites provide Dark mode. Extensions can directly override styles to force white text on a black background.
      • Apply to all websites: Such as Dark reader, night eye, noir...
      • Stylish for custom website CSS; advanced users can fine-tune individual website displays.

Conversion Process

• Turn on the operating system's own Dark theme; typically, supported software, apps, and web pages will adapt accordingly.
• For those that do not adapt, you can look through the software's settings or use Google keywords: software name + Dark or night mode to search further.
  • For instance, with the Zotero reference manager, I found an extension written by someone through a Google search to apply Dark mode.
  • Windows Task Manager began supporting it after Windows 11 version 22H2.
  • Some e-books on Kindle do not support Dark mode, so the only option is to enable system-wide invert colors.
  • Complete all tasks via the command-line interface; vim is the daily code editor, GUI get out of the way

My home soft router initially only had OpenWRT installed, and it did its job competently for a while. I casually noticed that even under heavy network load, RAM usage was only at 1% of the total 8GB. Hardware resources were not being properly utilized, so my engineering soul burned with the desire to squeeze every drop out of it.

One recommended configuration is to install a Proxmox VE environment, virtualize the Router software, and then use the remaining resources to install other guest operating systems like Windows, Linux, or deploy Docker services.

Proxmox VE Installing pfSense

The installation process for Proxmox itself is quite simple: download the ISO file, burn it to a USB flash drive, boot up, and install step by step. On heavy-duty servers with high-end hardware configurations, you could also consider the ZFS file system, but for a small soft router, I kept everything simple.

Download the pfSense iso file, upload it to Proxmox VE, and you can start the installation. There are plenty of tutorials for installing pfSense, and even official documentation: Virtualizing with Proxmox® VE. The following only discusses my personal requirements and the issues encountered during setup.

Planned Architecture

• The soft router machine has four physical network ports: enp1s0, enp2s0, enp3s0, enp4s0.
• The first one acts as the WAN port, and the remaining three serve as LAN ports.
• Proxmox VE can be accessed through the LAN ports.
• The virtualized pfSense under Proxmox VE is used as the Router, dialing PPPoE via WAN.
• Other network devices at home get network access through the LAN.

Issues

• Initially, Proxmox VE has already created a Linux Bridge, vmbr0, associated with enp1s0, and assigned the previously configured IP 192.168.100.2.
• The computer can only connect to enp1s0 via a cable to configure Proxmox VE.
• The other physical network ports enp2s0, enp3s0, and enp4s0 are associated with vmbr1.
• After pfSense is set up, it can function normally for network dialing and LAN distribution by assigning vmbr0 to WAN and vmbr1 to LAN. However, if this environment is put directly into use, the Proxmox VE management interface cannot be accessed from any machine already connected to the LAN.
• Following the above, if you want to manage Proxmox VE, you have to take the whole machine down and connect via enp1s0, which is extremely inconvenient.

Solution

After researching for a long time, and even once encountering a situation where a mid-way power outage caused all configuration files to be lost and the system wouldn't boot, requiring a Proxmox VE reinstall, I finally figured it out... 🤷‍♂️

• Remove the default IP and Gateway from vmbr0, and set them on vmbr1 instead.
• Assign the IP setting of vmbr1 to the same subnet as the pfSense LAN, with the Gateway pointing to pfSense 192.168.1.1.
  • The default DHCP range for pfSense is 192.168.1.100 ~

This allows access to Proxmox VE through the pfSense LAN ports.

Potential Issues

• The Proxmox VE host must go through pfSense to connect to the external network.
• If pfSense crashes, is stopped during migration, or is down for other maintenance reasons, all devices, including the Proxmox VE host, might lose connectivity.

pfSense vs OpenWRT

Understanding the development context based on product goals is better than arguing over which is superior or inferior for specific uses.

pfSense

• Based on FreeBSD, designed with the goal of being a professional firewall.
• Comprehensive firewall management features and complete management and monitoring functions.
• User interface is well-organized with clear explanatory text.
• Netgate provides commercial paid support, thorough documentation, and corresponding reference documents for various application scenes.

OpenWRT

• Based on Linux, oriented towards small embedded devices, suitable for home wireless routers with relatively low hardware resources.
• Small and lightweight; aside from basic functions, community-maintained packages can be installed via opkg.
• There is documentation, but it is not as user-oriented as the documentation maintained by Netgate for pfSense.
  • Basic networking knowledge is required to understand it, or where details are not provided, you need to search for more discussion articles for answers.
  • Discussion articles may not always provide the correct answers, requiring self-exploration.

I think both systems are excellent. If it were a wireless router or a machine with very low hardware specs, I would prioritize OpenWRT. But for a machine with a quad-core J1900 CPU and 8GB of RAM, I would consider installing pfSense to play around with.

Reference

• 在 UniFi Controller 5.9.29 啟用中華電信非固定制 IPv6 服務
• pfBlockerNG设置指南 | 鐵血男兒的BLOG
• pfSense安装AdGuardHome | 鐵血男兒的BLOG

This post contains summary notes after reading the book 19 Practices for Highly Sensitive People to Love Themselves.

• Books.com.tw link to the original book

The Self-Perception of HSPs (Highly Sensitive People)

• The "fragile heart" that remains unchanged even in adulthood
• Suffering immensely from caring too much about others' feelings
• Being extremely timid, constantly startled by things around you
• Perfectionism? Drive to succeed? Actually, I'm just cowardly
• Wearing a "clown mask" just to cover up neuroticism
• The fantasy world feels more real

Not blaming yourself is the first step for a highly sensitive you to shine.

The fragile heart and life struggles of an HSP are actually two sides of the same coin alongside beautiful strengths such as acute senses and receptivity, a rich mind and emotions, boundless imagination, and a high degree of empathy.

Most HSPs do indeed possess negative traits like being prone to anxiety, pessimism, negative thinking, being easily influenced by others, and getting fatigued easily. It is precisely these negative traits that make them easily hurt and full of suffering. You must tell yourself that sensitive and delicate traits are innate; it is not because you are too weak, too lazy, or have too low a tolerance for stress.

https://www.suncolor.com.tw/event/books/highlysensitive/quiz.html - Self-Assessment Scale for Highly Sensitive People

Characteristics of HSPs

• Sensory sensitivity symbolizes a high capacity for information processing
• Producing reactions from minute stimuli, thus able to perceive subtle changes
• Possessing excellent intuition and rich inspiration
• Empathetic ability, the ability of the "mind's eye"
• Empathizing deeply with the joys, angers, sorrows, and happiness of others
• Able to grasp the essence of things
• About 70% of HSPs are introverted
• Prone to anxiety and fatigue
• Also having sensitive reactions to chemical substances and electromagnetic waves
• The "Three Major Natural Enemies" of an HSP: overlapping tasks, sudden changes, and time limits
• Most afraid of strangers and unfamiliar places
• Blurred "boundaries," easily influenced by others
• About 30% of HSPs are "HSS (High Sensation Seeking) type HSPs"

Analyzing the Innate Sensitive Traits of HSPs from a Brain and Neurological Perspective

• Characteristics of brain circuits in HSPs
• Prone to feeling anxious because amygdala activity is too robust
• Empathy stems from the "empathy circuit" and "attunement circuit"
• The nervous systems for dopamine and serotonin are prone to disorder
• Neurological tension is a state of hyperarousal caused by the noradrenergic nervous system
• The relationship between the autonomic nervous system and HSPs

Why Are They Easily Hurt? Why Do They Feel Like They Are Living in Pain?

• Prone to developing attachment disorders
• Once an attachment disorder develops, they start playing the "good child"
• Prone to falling into a loss of self or codependent relationships
• Insufficient "denervation supersensitivity" leads to sensitivity
• HSPs are the "canaries in the coal mine" responsible for detecting danger
• Comfort that can only be created by an HSP
• Genes for anxiety will exert their effects under environmental influences

Daily Practices to Protect Yourself

• Maintain your own pace
• Establish your own private space
• Try to stay away from chemical substances and electromagnetic waves
• Pay attention to your own emotions and feelings
• Don't just zone out (try to move around more)
• Refuse decisively if you are unwilling
• Try to say positive and optimistic things as much as possible
• Find someone to listen to you
• You don't need many friends; one is enough
• Strengthen your boundaries
• Build your own castle
• Set priorities
• There is nothing wrong with occasionally showing weakness
• Establish an imaginary secure base
• Admit honestly that you are an HSP
• Use a firm attitude as a weapon
• Live out your truest self
• Harbor the determination to survive as an HSP
• Choose a lifestyle where an HSP can shine the brightest

Emergency Measures for When the Pain is Unbearable

• Focusing your attention on the "body" and the "present" can eliminate anxiety
• Brain gymnastics—helps you coordinate your autonomic nervous system, eliminate anxiety, and boost energy
• A simple three-minute meditation—helps soothe emotions
• The Lion's Purge—spit negative emotions from others into an imaginary trash can
• The tapping method (EFT)—calms your emotions

Linux in My Memory

I have been in contact with Linux since elementary school. After slowly understanding the Linux ecosystem and architecture, I developed a love-hate relationship with its high customizability. In the later stages, Linux was mainly used for server web development. Attempts to use a desktop environment for actual work were not nonexistent, but being busy with my studies, I didn't have the extra energy to squash bugs, so it always ended in mental and physical exhaustion.

However, being accustomed to the Unix architecture, I wanted to find a relatively stable desktop environment that could at least balance academics and work. So in 2013, I bought a Macbook Air and stepped into the Apple ecosystem, where I remain to this day.

A Few Thoughts on the Linux Desktop Environment

Since I am, after all, a layman in software development, some of my thoughts on the development of the Linux desktop may seem superficial. However, I feel that the problems with Linux desktop environment development actually stem from the open-source development community. Different components are maintained by different individual developers or groups, and their ideas are not always aligned. You may even encounter situations where differing opinions lead to arguments that remain unresolved.

Furthermore, the components of Linux are scattered across many upstream, midstream, and downstream parts. It is difficult to demand that the development team of each component perform comprehensive compatibility testing for the countless other components. Consequently, users are required to have a certain level of debugging capability. To some extent, this has affected the popularization of the Linux desktop.

The Linux community is essentially composed of people from all over the world; it is vibrant but loose, free and open without binding constraints. Anyone who wants to contribute can join or quit at any time, and they can always disagree with others' ideas and build their own from scratch. On distrowatch.org, you can see countless distributions, and each distribution may have a different mission and philosophy. You could say that compared to corporations, the resulting products lack the guarantee of long-term support, but there is no right or wrong in this. It is exactly this characteristic of the Linux community that has shaped the landscape of Linux today.

Disadvantages and Progress Compared to Windows and macOS

The author feels that compared to complete operating systems, Linux has several disadvantages when it comes to developing a desktop system:

1. Graphics card vendor support
2. UI design guidelines, like Apple's Human Interface Guidelines
3. A complete debugging and testing team to minimize issues upon delivery to the user end
4. A mature application ecosystem

Regarding point 1, support for the Linux desktop from Nvidia and AMD has been improving in recent years, and it is no longer as fragmented as it used to be. As for the third point, it becomes necessary to consider the characteristics of each distribution. Some Linux desktop distributions are supported by comprehensive community teams and have clearly announced testing cycle schedules from package testing to official release.

For the fourth point, it must be said that with the development of Chrome and Node.js, Web Apps and the electron framework have narrowed the gap between Linux applications and other operating systems (macOS has also benefited from this). Many lightweight applications can even be handled directly through a browser.

Revisiting Linux - Starting from Ubuntu 20.04

I built a new PC to occasionally play single-player PC games. My mac mini is taken to the office for work purposes, and I didn't want to use Windows for my primary work environment at home, so the idea of using Linux again arose.

To avoid trouble, I chose Ubuntu 20.04, thinking that a long-term stable distribution supported by a corporation should be very stable, right?

But after using it for a while, I still changed the operating system and reinstalled. Reflecting on it later, Ubuntu didn't actually have major problems; it was just minor system UI bugs that affected the overall perception of the details.

For example, being accustomed to macOS's reverse scrolling direction, I enabled it in the system settings, only to be frustrated to find that it would reset every time I rebooted.

The default Gnome desktop was too plain. Although I tried to convince myself to just get used to it, I still wanted to tweak it. As a result, I found that installing too many Gnome Extensions would cause various weird bugs to appear on the desktop.

Overall, I was not quite satisfied, so I decided to switch to Manjaro.

Manjaro KDE - Self-Willed Chaos

From this point on, the author violated the initial intention: to lower maintenance costs and focus on production use. Instead, I started spending a lot of time playing around with experimental features, but it didn't result in a much better unified experience.

Manjaro's default GUI installer is actually quite good, but because I wanted to use Btrfs and its snapshot feature, I used manjaro-architect to build it from the command line. The snapshot feature of Btrfs is very convenient, but it caused the memory function of grub-reboot to fail, meaning I couldn't reboot into Windows and have it automatically switch back to Windows next time.

Unfamiliar with the compositional structure of Btrfs partitions, I also had to spend some time researching where to start when something went wrong and needed rescuing.

Additionally, the desktop features of KDE are very comprehensive, showing an all-encompassing development route. However, there are too many settings and hidden bugs. Frequently switching and checking felt very troublesome. Also, KDE didn't seem to get along well with Nvidia, resulting in screen rendering bugs. At the same time, Wayland completely failed to launch.

Later, trying to hold my ground, I also installed i3-wm, a tiling window manager, and stuck with it for a while. Once you get used to i3-wm, managing windows is indeed very smooth, and thinking about the purpose of a window before laying it out also helps keep your mind clear. However, the problem with i3-wm is that it is, after all, an external component developed independently from various desktop systems. Therefore, when the internal components of KDE or Gnome desktop systems integrate, they do not consider compatibility with i3-wm. After prolonged use, you can still clearly see the strange behavior of some system windows under i3-wm.

Manjaro Gnome - The Beauty of Default Installation

As mentioned earlier, after much tossing and turning, I still felt the system had various imperfections large and small. Later, I made a resolute decision and simply wiped out Manjaro and reinstalled it, using all the default values of the GUI installer and ignoring the Btrfs format entirely.

I chose Gnome for the desktop, selected the Traditional Chinese locale, and installed it all the way through.

To my surprise, I found it quite smooth to use, noticeably better than when I previously installed it via manjaro-architect.

I think the main reasons are as follows:

1. Manjaro Gnome comes pre-installed with a whole set of useful Gnome Extensions, making up for the deficiencies of the native Gnome desktop.
2. The provided Layout Switcher allows switching the Dock position according to your needs and also has a minimalist tiling style.
3. The Gnome-style system settings are simpler and not as complex as KDE's.

Of course, it's not without flaws. The drawback of the Gnome desktop is that many features rely on Extensions to be supplemented, but the compatibility among Extensions isn't always great. If you install too many, conflicts when adjusting the settings of various Extensions can easily cause the desktop to crash. This happens even when tweaking the default settings provided by Manjaro.

As for the backup solution after abandoning Btrfs... I just directly run a scheduled rsync of the /home folder to the NAS folder.

What surprised me the most about Manjaro Gnome is that it has the Pop OS! tiling window extension built-in, directly providing functionality similar to i3-wm. The text configuration files I used to write painstakingly for i3-wm now seem like a huge waste of time.

Conclusion

In short, I am now staying with the Manjaro Gnome desktop environment. At least with the built-in desktop extensions, code editor, browser, and VMWare installed, it is usable for everyday tasks. Although Manjaro is derived from Arch Linux, its stability means it won't blow up every few days.

In the Nanhua community, aside from learning about its geographical location and industrial characteristics, through teaching discussions with resident physicians and conversations with health center staff, I gradually came to understand local issues such as population aging and labor outflow. The social issues constantly described in textbook form during high school are now vividly presented before my eyes, allowing me to personally experience the mission and helplessness of frontline grassroots personnel.

Traveling back and forth to Nanhua multiple times, we were fortunate to be driven by a senior resident physician, sparing us the exhaustion of travel. However, on one occasion due to scheduling conflicts, we had to take public transportation ourselves, spending nearly two hours in a packed, bumpy bus like sardines. It was then that I deeply felt the preciousness of medical accessibility in urban Tainan and the inconvenience of seeking medical treatment in rural areas. In the Nanhua area, the health center serves as the first line of medical care, handling important local public health duties such as vaccinations and chronic disease tracking and control. The adjacent fire department's ambulance can, if necessary, immediately transfer acute and critical cases to the Xinhua Branch Hospital, and only then to National Cheng Kung University Hospital.

Home Visits

One very memorable experience was the home visit. The subjects of the home visit were a family very common in the Nanhua District: elderly parents in their empty nest phase, living idly at home. As we followed the map and chatted in halting Taiwanese according to the home visit questionnaire, we were surprised to find that the interviewed grandfather and grandmother did not have any social functional disabilities. Aside from the grandfather needing his children to drive him to National Cheng Kung University Hospital for follow-ups due to a minor stroke and a cardiac stent, both elders could walk to the health center for medical treatment on their own. They also actively participated in community activities such as karaoke competitions and volunteered at community care centers. Their sons and daughters also returned frequently to visit. However, if someone were to become bedridden, related care needs and burden issues would emerge, making the subsequent situation quite worrying.

Administrative Situation

Aside from general duties, what more can the health center do in Nanhua? Because the Nanhua District is close to the mountains with many mountain roads, making transportation inconvenient, there is a desire to maximize medical accessibility. Therefore, the health center conducts mobile medical clinics in Guanshan Village, with the health center director personally taking a vehicle along with accompanying nurses and pharmacists to see patients there. Additionally, one could consider leveraging the power of important regional gatherings or religious activities. For instance, Nanhua Village has the Tianhou Temple as a gathering hub; could health education campaigns be conducted there? Because a higher proportion of the elderly in the Nanhua area missed out on early education, the design of health education flyers or posters must take this into account. Only by making good use of neighborhood activities and the power of word-of-mouth is there a chance to deeply root important medical concepts in people's minds. However, facing the macro-environmental trend of young and middle-aged labor outflow and a scarcity of care resources, this aspect may still require seeking assistance and planning from higher-level authorities.

Because there are subsidies from reservoir conservation area compensation funds, many people are registered in Nanhua but do not actually live there. This causes administrative troubles for the health center, such as tracking the status of newborn vaccinations. Under pressure from higher-ups, the health center bears the responsibility of monitoring the vaccination rate and will even go to homes to evaluate vaccinations when necessary. Once, the head nurse took us over the mountains to a low-income family that had been repeatedly notified for vaccinations but consistently failed to show up. Because all three children in the family had developmental delays, the head nurse's face was filled with worry. The children outside in the courtyard were bouncing around; inside the house, while I amused the children and conducted developmental assessments, I thought about the urban-rural gap and balanced regional development. Higher government authorities constantly talk about improving these predicaments, but how many people stand on the front lines bearing endless heartache like the grassroots personnel of the health center?

Regarding Long-Term Care

In various teaching discussions, Dr. Chih-Hsing Wu shared his thoughts with us regarding the authorities' long-term care policies. Locally in Nanhua, after the Long-Term Care 2.0 policy began to be promoted, the district office started coordinating through the Social Welfare Section, planning to expand and strengthen the original meal-provision care centers to meet the Level A, B, and C grading structure in the Long-Term Care 2.0 plan. However, official documents sent to higher authorities have continuously met with no response, and funding has not been allocated. Regarding the promotion of Long-Term Care 2.0, the health center, which should be the directly affiliated institution, has received absolutely no news in this regard, revealing a situation of conflicting policies and unequal information. And while writing this report, I also heard that the Ministry of Health and Welfare will separately establish a Department of Long-Term Care to coordinate long-term care policies. Yet, the emperor is far away; officials sitting in offices set ideal performance goals based on written theories and statistics, but they will not tell the grassroots level exactly how to implement them. When the grassroots level cries out to heaven and earth with no answer, the ideal vision of the Long-Term Care 2.0 policy can only struggle along, fighting while trapped in the mud.

This was a young patient in his thirties, suffering from vascular events like stroke and myocardial infarction at such a young age, which likely affected his ability to work; there was also a note of outstanding payments in the electronic medical record system. Because the condition was critical, family members needed to be contacted. But whether due to estranged relationships or the late hour, neither of the two relatives (ex-wife, older brother) could be reached. Furthermore, according to the nurses, there had been no caregiver at his bedside since his admission. The ECG showed ST-elevation and T wave inversion. The on-call cardiologist was consulted; a bedside echo was performed, revealing that the anterior wall of the heart was indeed contracting poorly, but comparing it with the imaging record of the previous myocardial infarction, it could not be confirmed whether this was a new or pre-existing problem. With numerous risks present, the senior resident later decided to transfer him to the ICU.

For me, perhaps because of a period of clinical experience, what left a deep impression was not the sudden onset of shortness of breath. What touched me was the powerlessness and loneliness, unsuited for his age and tormented by illness, on a face ready to give up struggling on the margins of the world.

Besides adhering to their professional duties, what can medical personnel do for him? Through countless long nights, amidst sudden shortness of breath, chest pain, chest tightness, or even the emergency of cardiac arrest, physicians exhaust their minds to figure out if there is anything more to be done, playing tug-of-war with death on the line between life and death. But beyond vital signs, an on-call physician can only let out a sigh over a chance encounter in a fleeting life, and then continue to be a passerby.

However, by past four in the afternoon, the patient still hadn't arrived. A colleague nearby, who had previously rotated through Infectious Diseases, was already muttering complaints about the patient's past history:

"He lives in the mountains of Chiayi. The cause of his HIV infection was likely sharing needles due to drug addiction. Both husband and wife were infected, then they blamed each other, and eventually divorced."

"Originally, they both wanted children, and even thought about doing IVF. But because of this, they couldn't have any."

"The last time he came for admission, he was also scheduled for the day, but only called at 5:00 PM to ask if he could come tomorrow instead."

"Don't know why he went to Dalin Tzu Chi Hospital, and later they refused to take him."

"Ah, but he does frequently brag about his adopted daughter, praising how well she does in school."

Someone asked: "Didn't they divorce? And can he raise a child like that?" "I don't know, it seems she was adopted before the divorce."

The colleague remarked with annoyance: "Can we ask the admission service to just cancel the admission for anyone who hasn't arrived by 4:30?"

Later, the patient finally arrived to check in for admission close to five o'clock. But by this time, the day shift staff had all gone off duty. Just then, a few other beds had unexpected issues, and the night shift on-call doctor and nursing staff were overwhelmingly busy. Thinking I would help take the admission to lighten my on-call colleagues' burden, I finally saw the patient's frail appearance.

The patient was indeed experiencing such severe neck pain that he could barely turn it, affecting his daily activities. Furthermore, being a single parent who had to care for a child, he likely had quite a few matters to attend to at home. Getting everything sorted out and traveling from Chiayi could genuinely cause a significant delay.

However, amidst his descriptions of discomfort—aside from the neck pain and various other minor aches—we also couldn't rule out opioid withdrawal. In other words, the recoil of his past drug addiction was now returning to strike him.

The medical staff's impatience and the wearing away of their tolerance stemmed from an inability to forgive the patient's lack of self-care due to drug addiction and his abuse of medical resources. But what I found peculiar was that someone who seemed to have completely banished himself from society still, according to the stories, retained a passion for raising his child, subtly projecting the beautiful image shaped by this society onto his daughter.

Recalling everything I have seen in the ward from last month until now, I truly cannot help but ask: are the values by which each person judges, or the fairness and justice conveyed by society, really that absolute? Once we render a judgment based on our own perspectives, is it not true that, to some extent, simply by being human, we are bound to have prejudices?

I have seen a famous NTU obstetrics and gynecology doctor, who was both brilliant and enormously wealthy in his youth, surrounded by his reluctant, grieving children at his deathbed; I have also seen the bleakness of a terminal cancer patient dying alone, a wife suspected of being abandoned by her husband.

I have seen family members in their fifties, riddled with ailments, whose sisters were all deaf and mute, bearing the entire burden of caring for an eighty-something-year-old mother—helpless, anxious, and frantically applying pressure to the nursing staff; I have also seen a carefree patient, a member of a southern merchant family living off inherited wealth, who despite his young age suffered from multiple chronic diseases.

All beings will never be equal, at least not in their right to choose, nor in the breathing room life's pressures leave for seeking dignity. But if this society cannot pass judgment on the true efforts each person makes, must we really leave it to God? Yet, I both believe and do not believe in God.

The God I do not believe in is a God mixed with human desires, conceived from an anthropocentric perspective. Humans are imperfect, confined by their own prejudices, and arrogant.

The compilation of the Bible can be traced back to history selectively concealed and altered by the Roman regime for public and private reasons; the old customs of Islam reflect the ancient desert tribes' pursuit of social stability, but carried over to modern times, they have become shackles for those in power to restrict women's rights; the practices of Buddhism, after spreading far and wide, incorporated local characteristics but transformed into oppression, harboring decayed Confucian hierarchies or even amassing wealth.

Whether it is genuine, sincere faith that leads to aberrant and misguided actions, or the pursuit of selfish gain under the guise of God, I do not believe in a God shaped by fundamental human desires. Regarding good deeds done in the name of God, I only revere the purity of sacrifice, dedication, and wholehearted giving.

The God I do believe in serves to remind us that humans are insignificant. In this era of population explosion, where technology brings the illusion that man can conquer nature, there are still forces operating beyond human comprehension; there are too many unimaginable unknowns beyond our perspective. Christian doctrine once mentioned: "Do not presume upon God." Extending this interpretation, using "God" as a collective pronoun for everything transcending human thought, the God I believe in serves to constantly caution us to remain humble, not to rashly judge right and wrong based on our own thoughts, and not to speak recklessly of fairness and justice based on our own views.

In the end, perhaps the pursuit of absolute fairness and justice beyond one's own subjective views is itself a form of pride or obsession? The Buddha said, "Sever attachment to self, attachment to views." When we rack our brains wondering why the world operates this way, and feel indignant over injustice, have we not jumped into the dead end of this obsession? Focusing on our own duties, staying true to the fairness and justice we recognize, and striving to have a clear conscience—perhaps this is the best that beings as insignificant as humans can do.

At the beginning of every month, when a resident physician rotates to a different department or ward, there is always a transition period. It takes time to familiarize oneself with the characteristics of different nursing stations, as well as the detailed disease progression of patients who have been staying longer, from the previous month to the current one.

Having just arrived at the 10C ward, before my seat was even warm, I was urgently paged by a nurse.

Doctor, the patient has been on Ceftazidime since last Friday, but is still having a fever right now! Do you want to take a look and prescribe any treatments?

The patient, Xiaohui (a pseudonym), had been hospitalized for over sixty days. She had terminal ovarian cancer with multiple metastases. Without time to carefully review all her weekly medical records since admission, I first learned from the diagnosis and handover information that she had a history of intra-abdominal infection (IAI) and urinary tract infection (UTI) during her stay. The Ceftazidime she was currently receiving was prescribed to treat the UTI.

At the bedside, I saw that Xiaohui looked very sick and a bit drowsy, and jaundice could be identified at the same time. However, a blood culture from three days ago happened to report the growth of E. coli at this time. So I wanted to wait for the antibiotic susceptibility report to determine which antibiotic to switch to.

Doctor, the susceptibility report is out. The E. coli from the peripheral blood culture looks very easy to kill; it is susceptible to Ceftazidime.

What on earth! If Ceftazidime could kill it, why was she still having a fever? Moreover, based on the patient's condition, it looked entirely like a fever driven by a bacterial infection. Looking at the fever curve record, there was indeed no fever on the first day after starting Ceftazidime, but it flared up again afterward. I could only guess that a new resistant bacterial infection had emerged, and the culture report from three days ago could not be fully trusted. I had to redo the entire fever workup and switch to a different empiric antibiotic. I chose Tazocin and started it directly.

Doctor, Xiaohui's consciousness is less drowsy, and she is no longer having a fever. Also, the urine culture report from last week is out, please take a look.

On the second day of Tazocin, Xiaohui's condition improved and the fever stopped. The urine culture report grew two strains of bacteria, both resistant to Ceftazidime. Well, no wonder it couldn't be suppressed. Cefepime could cover both, and with the attending physician's consent, we switched to it.

Up to this point, I was still glad that I had helped Xiaohui and done the right thing out of my duty as a physician. But...

Pleading Before the Sickbed

Doctor, I beg you, let me be free sooner, okay? Every day I wake up, I feel so, so cold. There is no one I know around me. It is truly lonely, you know...

Before the sickbed, tightly holding the hands offered by the attending physician, Xiaohui pleaded weakly. I stood dejectedly behind the attending physician, gazing at the sorrow on the patient's face.

Outside the nursing station, the senior resident from the hospice team, the attending physician, and Xiaohui's mother discussed in low voices, deciding to invite Xiaohui's husband over to participate in the decision-making together. While waiting, I walked into the doctors' office with the attending physician to adjust Xiaohui's current medical orders.

If the family decides to respect Xiaohui's wishes, discontinue all antibiotics and nutritional IV fluids, and start Dormicum so that Xiaohui won't be in so much pain at the end of life.

Numbly recording the medical orders on the rounding list, and upon finishing the handover with the senior resident of the hospice team, the words stuck in my throat finally burst out as I was about to leave:

Teacher, was my aggressive management of the fever and switching of antibiotics a few days ago a mistake for Xiaohui?

No. In terms of respecting the patient's wishes, Xiaohui still wanted to survive at that time. This patient has been here for quite a while. Even though she knew it was terminal cancer with slim hope, she still begged us to save her when she first arrived.

Xiaohui had attachments to her family; her pair of children are not yet adults. But perhaps during her hospitalization, she was truly tormented for too long. Her infection worsened and improved, going back and forth, but the pain of cancer never ceased. Every time her condition deteriorated, I would think, if the patient just passed away like this, maybe it would be a release.

The Farewell at the End of Life

Pre-dying. This is the terminology used by the hospice team to communicate with each other, hoping to prepare early for a good death for the patient.

Meeting the people they need to meet; saying words of apology, gratitude, and love; reviewing their life; recalling and narrating past stories.

In other parts of the hospital, patients come and go, and the lifeline is maintained amidst the doctors' hurried busyness, laboratory tests, and the pulsation of heart rates, blood pressures, and oxygen saturations. But in the hospice ward, human hearts gather into thoughts, leaving a pure land for the patient to seek eternal rest beyond the emptiness of the secular world.

What is a physician's role in the hospice ward? Because of rotating departments every month, the establishment of relationships and spiritual support are far inferior to the long-term cultivation and deepening by the hospice team. When Xiaohui entered the pre-dying state and fell into a drowsy coma, I was stunned and regretful. Watching the vital signs curve slowly plummet, I felt for the first time that a physician's capabilities are so limited; at the end of life, we still have to bow our heads.

September 8, 2019. Having survived a busy on-call day with two beds of shock and one bed of hyperkalemia needing emergency dialysis. Late at night after the handover, right before getting off work, perhaps having a premonition that the time of parting was imminent, I returned to the hospice ward and quietly walked to the sickbed to take one last look at Xiaohui. Her family was peacefully accompanying her by her side as she slept. On the morning of September 9, Xiaohui's name could no longer be seen on the patient list in the electronic medical record system; only the application document for the death certificate remained on the desk.

An 80 y/r female was found to have liver cirrhosis complicated by ascites earlier this year and was referred to our hospital's gastroenterology outpatient clinic. Blood tests showed HBV and HCV were both negative. During the week of 8/11, the patient's consciousness gradually deteriorated, accompanied by nausea, vomiting, poor appetite, and constipation. Following an outpatient visit on 8/14, she was admitted on 8/15 for ascites tapping and lab testing.

At the bedside, the patient was assessed to have a GCS of E2V2M4. The patient appeared extremely emaciated overall, with almost no muscle, while presenting with abdominal distension that yielded a dull sound upon percussion. From her family's account and follow-up medical records, it was noted that her weight had dropped 15 Kg from March to the present.

In gastroenterology, the most common cause of altered mental status is Hepatic encephalopathy, but of course, other causes must be ruled out. Using the mnemonic A-E-I-O-U-T-I-P-S, considering the history, besides basic biochemistry, CBC/DC, we checked blood glucose, electrolytes, Ammonia, and Vein Gas. Although there was no fever, two sets of blood cultures were drawn to rule out infection under the VS attending physician's recommendation. It was also pointed out that regarding electrolytes, the main ones affecting consciousness are Na and Ca.

(A more senior VS mentioned he felt that checking routine Ammonia is meaningless because its blood concentration doesn't completely correlate with hepatic encephalopathy; even if you draw a value within the normal range, you cannot rule out hepatic encephalopathy.)

Because this patient arrived around 4 PM, when the hospital's lab value alert text message came, I was still in the hospital. Fearing the on-call physician's workload would be too heavy, I handled it myself, and ended up dealing with it until past 8 PM, so the impression is still very deep now.

Blood calcium was as high as 14.5 mg

Blood calcium was as high as 14.5 mg

Blood calcium was as high as 14.5 mg

Well then, after a search, emergency symptom management mainly involves aggressive hydration for Dilution and giving calcium-lowering drugs. After inquiry, subcutaneous Calcitonin injection was chosen on the spot. An EKG also had to be done. Being in my first month as a PGY, I was on edge about everything. The sticking point was: ah, there's already ascites present, wouldn't giving hydration just put more fluid into the abdominal cavity? But finally, weighing the dangers of hypercalcemia, the ascites could only be considered a minor issue. Other significant Lab data included Anemia, Thrombocytopenia, Acute Kidney Injury, and Hypoalbuminemia.

台灣內科醫學會 -高血鈣的診斷與治療

Later during admission, after emergency Hydration and Calcitonin injection, the patient's blood calcium dropped, and her consciousness gradually cleared. The original symptoms of nausea, vomiting, and constipation also subsequently improved. But the underlying cause wasn't resolved; the hypercalcemia kept relapsing. In the lengthy hypercalcemia workup process, every week when the blood calcium spiked, Hydration had to be continued. In between, there were even episodes of pulmonary edema and hypokalemia. The patient's family and primary nurse expressed mental exhaustion, and I felt very guilty whenever explaining the disease to the family and primary nurse...

The Survey process was very tortuous. Synthesizing various guidelines, we first had these results: 1. PTHi was low. 2. Vitamin D 25(OH)D was in the normal range. Our hospital didn't have 1,25(OH)2D or PTHrP. After checking thyroid function, it was found to be Subclinical Hypothyroidism.

Tracing back the patient's history, the family clearly stated that the patient only started lying in bed after her consciousness altered; otherwise, she could usually walk with a walker and do chores herself. The family also mentioned there was no family history of hypercalcemia, nor was she taking calcium tablets. Additionally, she had visited a pain clinic for back and bone pain due to HIVD.

Synthesizing all this, the investigation launched mainly in three directions:

1. Multiple myeloma or other bone marrow diseases
2. Tuberculosis infection
3. Cancer

Multiple Myeloma

Multiple myeloma was suspected because the patient actually met the symptoms of CRAB:

1. C - Hypercalcemia
2. R - Renal complications
3. A - Anemia
4. B - Bone disease

A reverse A/G ratio was observed in this patient, but this could appear in patients with either liver cirrhosis or Multiple myeloma. The former is due to low Albumin, while the latter is due to an abnormal increase in globulins. For this patient, it was hard to determine.

We checked IFE (Immunofixation Electrophoresis) and found increases in both IgA and IgG, presenting a polyclonal pattern, which is less like the monoclonal pattern of Multiple myeloma. An examination of the Skull X-ray did not reveal the osteolytic lesions of Multiple myeloma. Checking the kappa/lambda ratio, it was also normal.

Ascites

The patient's Chest X-ray did not show characteristics of pulmonary tuberculosis. Two sets of sputum left for Acid-fast stain were also negative; a third set was delayed at the turn of the month because the patient had no sputum. What raised suspicion was the abdominal ultrasound of the ascites.

Under ultrasound, characteristics of loculated ascites could be seen. After consulting a senior resident, we had to suspect peritonitis or peritoneal carcinomatosis infection. It could also present this way if Abdominal tapping had been done multiple times to drain ascites, but this was the patient's first time having ascites drained.

Hmm, so the ascites was sent to test a bunch of things.

At least the ascites Data didn't look like an SBP infection, but the strange point was the very high Total protein. If calculating the SAAG, it would definitely be less than 1.1. The disease could be TB peritonitis, peritoneal carcinomatosis, pancreatic ascites, or nephrotic syndrome. Among these, TB peritonitis and peritoneal carcinomatosis were exactly the possible causes we suspected for the hypercalcemia. After ruling out others from the Data, we sent the ascites again for TB-PCR and Cell cytology (to look for malignancy), and the results were all negative...

Possibility of Cancer?

Losing a lot of weight in a short period, combined with hypercalcemia, definitely makes the possibility of Cancer very high. But an awkward point for the patient was that she had just had an abdominal CT scan at the end of July when coming to the ER for fever and abdominal pain. If we were to do it again at the end of August, there might be issues with National Health Insurance deductions and radiation exposure. At that time, the abdominal CT did not show lesions suspicious of Cancer. The earlier pathology report of the ascites was also negative. Head and neck as well as breast Physical Examinations did not show obvious mass lesions. The only thing not done was a chest CT...

Latest Progress Before the Month Turnover

Later, after the attending physician coordinated with the family, they decided to directly do a whole-body CT scan and a bone scan. There might also have been a factor of the family requesting less torment. However, a glance at the report seemed to show no obvious lesions in the chest; the main issue was still the abdominal ascites over there. I will continue to follow up when I have time later.