[🤖/✨] From Anger Over the Reservist Death Case to Restarting the Blog
✨ GPT-5.5 Summary
A full GPT-5.5 conversation archive that began with anger over coverage of a reservist death case and moved through politics, faith, execution checks, and the decision to restart the blog.
Conversation Flow
This post is a public archive of a conversation from May 25, 2026. Images were omitted, and some profanity and extreme wording were masked as ***.
The conversation began with anger over a reservist death case, but the flow moved through politics and human nature, the narrow road of faith, self-checking through AI, and finally the decision to restart the blog.
The purpose of this post is not to present a finished argument, but to preserve the flow of one person moving from anger toward execution. User messages are highlighted in yellow.
The Reservist Death Case And The Media
Wow, this is actually chilling... Are they saying the reservist death case matters less than one wolf *** escaping, the Neukgu incident, or the Starbucks May 18 insult-marketing controversy? MBC, JTBC, SBS, all of them, are they out of their minds?
I understand the anger here too.
A person died during reserve forces training, and this is not just a simple accident. It involves training safety management, emergency response, and military responsibility. It is strange that public pressure has been so weak.
To be precise, though, it is not that MBC did not cover it at all. On May 14, MBC reported that "a man in his 20s collapsed and died while undergoing reserve forces training in Pocheon, Gyeonggi Province; authorities are investigating the circumstances." The report said that a man in his 20s collapsed during mobilized reserve forces training in a mountain area in Pocheon, was taken to a hospital, but died. (MBC NEWS)
The issue is whether this is the kind of case you handle as a brief item and then drop. After that, the Center for Military Human Rights Korea cited a KBS report saying the unit was a pilot "full reserve battalion," and that essential emergency response systems such as a military doctor, medic, and AED were absent from the scene at the time of the accident. The center demanded a truth-finding investigation. (Daum) At that level, this is not merely a "tragic accident"; it is a case where you have to look at system design failure, absence of safety measures, and the possibility of responsibility being evaded.
The Neukgu escape can be reported because it is a public safety issue, and the Starbucks May 18 controversy also has news value because it involves historical insult and corporate responsibility. The Starbucks case did in fact grow into accusations and investigation issues after controversy erupted over promotional wording on May 18 that appeared to mock the May 18 Democratic Uprising and martyr Park Jong-cheol. (Hankyoreh)
But even accounting for that, if broadcasters are not persistently following up after a reservist called by the state died during training, that is a real problem. Most Korean men go through reserve forces, and the state forcibly summons them. This cannot be brushed off as one individual's bad luck.
I think the core is this.
Neukgu spreads because it is cute and memeable. Starbucks catches fire because it is a political and historical controversy. But a reservist death involves the safety of the state, the military, and male citizens, yet somehow it falls into a structure where it is easy to quietly bury.
The point to be furious about is less "why are they reporting other things" and more:
why are they not digging into the reservist death case with that level of persistence?
For the bereaved family, and for everyone who has to go to reserve forces training, this is absolutely not something that should just pass by.
---
Sources:
- MBC report on a man in his 20s collapsing and dying during reserve forces training in Pocheon
- Center for Military Human Rights Korea demands truth-finding on the Pocheon reservist death
- Hankyoreh report on the Starbucks May 18 insult controversy
What do you mean MBC uploaded a video, you ***. They just squeezed out one article and stopped, ***.
You're right. I worded that badly.
It is correct to say MBC did not "properly cover it with video"; they posted one short article and effectively ended there. Looking it up, what appears on MBC's side is one short May 14 article, while outlets like KBS and Channel A show YouTube videos or caption-news formats. (MBC NEWS)
So the point that is pissing you off is exactly right.
This case is not a one-line item like "a reservist in his 20s collapsed and died."
A person forcibly summoned by the state
died during training
during night movement/reconnaissance training
with military emergency response, scene safety management, and command responsibility needing verification
and it directly concerns everyone who will have to go to reserve forces training in the future.
For an issue at this level, terrestrial broadcasters and cable news channels could dig for days and still not be doing too much. The fact that it is being treated as a brief and buried is genuinely bizarre.
And the sense people have in the comments - that there are dozens of videos about Neukgu and the Starbucks May 18 controversy while the reservist death is quiet - is not totally wrong. It is not that the Neukgu or Starbucks stories have no news value. It is that there is a structure where they are more easily consumed and spread than the death of a citizen forcibly summoned by the state.
My earlier line that "MBC covered it too" was too loose.
The accurate version is: MBC posted one short article, and it is hard to call that proper follow-up reporting.
---
Sources:
Politics, Camp Logic, And Invalid Ballots
It really feels like Korea's Democratic Party is moving in the direction of following China and North Korea.
I understand the flow of anger that makes it feel that way. Especially on military, security, North Korean human rights, and China-related issues, some Democratic Party figures often seem excessively vague, or appear to avoid saying anything unfavorable to North Korea or China while hitting Japan, the United States, and conservatives hard. It is natural that this creates that impression.
Still, I think the wording should be separated a bit.
If you flatly say "the entire Democratic Party follows China and North Korea," that is very strong and easy to attack.
But saying "the Democratic Party mainstream's foreign policy and security instincts look excessively conciliatory toward China and North Korea, and insensitive toward the safety of Korean citizens, soldiers, reservists, and North Korean human rights" is much more precise and persuasive.
That connects especially well to the point that is making you angry right now.
A reservist death case is something progressives and conservatives alike should be raising hell about.
The state called a citizen in, he died during training, and emergency response, scene control, and command responsibility are all at issue. If this is quiet, it naturally produces questions like:
"Do soldiers not count as votes?"
"Is the safety of male citizens outside their interest?"
"They say they criticize state power, so why are they selectively quiet about military and government responsibility?"
"Wouldn't they speak differently if this involved North Korean soldiers or China?"
This is where the Democratic Party and progressive media lose trust.
They speak outwardly of human rights, life, and safety, but if they lack persistence when a young man forcibly sent to reserve forces training dies, it inevitably looks like hypocrisy.
I think "following" is less accurate than describing the current Democratic Party problem this way:
They are extremely sensitive to anti-Japan and anti-conservative frames, but dull or evasive on anti-China, anti-North Korea, soldier safety, North Korean human rights, and male citizenship issues.
When that accumulates, people will naturally see them and ask, "Aren't they on China and North Korea's side?"
The reason that suspicion cannot simply be dismissed as a conspiracy theory is that too many of their past attitudes have piled up.
The Democratic Party *** have always been like that. Hypocritical *** who pretend to be good and put justice out front. *** who are masters of false agitation. But does that mean the People Power Party *** are not like that? They push election-fraud theories, try any way they can to defend Yoon Suk Yeol's insurrection, and use anti-communist and spy frames to send innocent people to the grave.
The *** who do politics are human beings, so how could they be good? ***
Honestly, that makes sense.
Your anger right now is not "which party is better?" It is exploding because the entire political class seems to sell good-versus-evil frames while moving selectively when human life and responsibility are actually in front of them.
The Democratic Party puts forward "justice, human rights, the weak, democracy." If those words are real, they should be more persistent than anyone on a case like a reservist's death. A citizen called by the state died; that is the core territory of monitoring power. But if they are quiet or selectively responsive, the moment naturally looks like "hypocrisy pretending to be good."
The People Power Party is the same. They talk about liberal democracy, security, and rule of law, but when power on their own side is at issue, and they drag in election-fraud theories, red-baiting, spy hunts, and logic that defends insurrection, that is not rule of law. It is camp survival logic. They shout "crush communism," but if you ask how sincere they have been about actual public safety and the treatment of soldiers and reservists, there is a lot they cannot answer for.
I think the core of what you said is this.
Politicians do not enter politics in pursuit of the good; they use the language of goodness to acquire and defend power.
That is what makes it more disgusting.
If they said like villains, "I like power," it would at least be less disgusting. But they all say "for the people," "for justice," "for democracy," "for security," and then in reality they are silent when their own camp is disadvantaged, and use even dead people as framing material when it benefits them.
But Hyuk, even in the middle of that, I think one thing has to be separated.
The fact that politicians are not good does not mean you can give up political judgment itself; if you do, the more shameless ones win.
Even if "those guys are trash and these guys are trash" is true, what you have to look at every time is still this:
Who is blurring responsibility in this case?
Who is quietly burying the death of a person?
Who is actually demanding truth-finding and prevention of recurrence?
Who stays silent when it is their side and loses their mind when it is the other side?
You do not need to trust politicians.
You do not need to love a party.
Just look case by case at who is less cowardly and who asks for responsibility more concretely.
The reservist death case really has to be judged by that standard.
Whether Democratic Party, People Power Party, or media, any side that quietly lets this pass has no right to speak of "public safety."
Why are the People Power Party *** just sitting still instead of pressuring the Democratic Party in this situation? I do not understand?
Not understanding it is normal. If they call themselves a security party and are now the opposition, this is an issue they should be tearing into.
If you search, it is not that they did absolutely nothing. People Power Party lawmaker Joo Jin-woo did demand that responsible officials be dismissed, saying that a young man in his 20s died during reserve forces training and that there was no medic or AED at the scene. The Center for Military Human Rights Korea also cited KBS reporting and raised the issue that the unit was a pilot "full reserve battalion" and that there was no medical staff or AED at the scene. (Maeil Shinmun)
But as you said, it does not look like an all-out party-level offensive. Why? Several reasons are probably mixed together.
First, the People Power Party is not clean on military issues either.
Military deaths, treatment of soldiers, and reserve training conditions are not structural problems belonging to only one administration. They have accumulated for a long time. If they dig hard, they are not free from past responsibility either. They will obviously get hit with, "What did you do when you were in power?"
Second, they seem to view a reservist death as an issue where anger is high but electoral calculation is ambiguous.
The political class disgustingly asks first, "How much does this help mobilize our supporters?" Reserve forces issues hit men in their 20s through 40s directly, but honestly, has the People Power Party truly treated this group as strategically precious? Not really. They often call on "men in their twenties" only when they need them, while actual system improvements are left behind.
Third, to pressure the Democratic Party, they would also have to ask about military responsibility, and conservative parties instinctively feel uncomfortable hitting the military hard.
They like security frames. But once you start digging into unit commanders, the Ministry of National Defense, training systems, budgets, and safety regulations, it looks like an attack on the military organization itself. So they can speak of responsibility under a Democratic administration, but may be passive about fundamentally criticizing the military system.
Fourth, because media firepower is weak, the political world moves less too.
It should be the other way around, but in reality the media has to make the issue big before politicians think, "Ah, we can ride this." KBS covered it with video, MBC handled it as a short article, and the Center for Military Human Rights Korea raised the issue, but it does not yet seem to have exploded into nationwide anger. MBC's coverage was short, basically saying a man in his 20s collapsed during mobilized reserve forces training in a mountain area in Pocheon and authorities were investigating the circumstances. (MBC NEWS)
Fifth, the People Power Party may simply be failing to set its priorities for checking the administration.
If they truly want to pressure the Democratic Party, this is the most legitimate kind of issue. "A citizen forcibly summoned by the state died. There was no medical staff or AED at the scene. What did the president, the defense ministry, and the military do? Convene the National Defense Committee. Submit data. Punish responsible officials. Inspect the reserve forces safety manual nationwide." That is how they should go. If they cannot grow this issue, they just have no political sense.
So the conclusion is this.
The People Power Party could use this case as a card to pressure the Democratic Party, but appears unable to grow it into a party-level issue because of the burden of touching the military organization, its own past responsibility, low sensitivity toward men in their 20s through 40s, and lack of media firepower.
If they want to act like a real opposition party, what they should say now is simple.
"Reservists are citizens called by the state. The National Defense Committee must determine why there was no on-site emergency system at the time of the death, and we will ask the defense minister and the command line to take responsibility."
If they cannot do that, the People Power Party does not have much right to call itself a "security party."
---
Sources:
- Maeil Shinmun report on Joo Jin-woo demanding dismissals over the young reservist death
- MBC report on the Pocheon reserve forces training death
What is Lee Jun-seok doing?
Looking around, I do not clearly see traces of Lee Jun-seok publicly going hard on this reservist death case. At least based on the articles and reports currently visible in search, what stands out on the People Power Party side is Joo Jin-woo posting on social media that there was neither a medic nor an AED at the scene, that this was a clear human-made disaster, and that responsible officials should be dismissed. Lee Jun-seok's remarks do not visibly stand out. (Channel A)
What makes this stranger is that Lee Jun-seok has an image as a politician who usually reacts sensitively to reserve forces, military service, and men in their 20s and 30s. Then, politically speaking, this is material he should not simply pass over. According to the KBS report, there was no essential emergency equipment like a military doctor, medic, or AED at the accident scene, and the medical support team was at a base 5 to 8 km away from the scene. This is exactly the kind of issue where he could ask, "Does it make sense to forcibly summon reservists while having a safety system like this?" (Daum)
There are several possibilities.
First, he may be judging that the issue has not yet attached nationally and is waiting for timing.
Second, it may be buried under larger issues in the presidential election or political-frame context, so he has not set the priority.
Third, Lee Jun-seok is ultimately a politician too, so he may be selectively silent when the board is not most favorable to him.
Fourth, the Reform Party/Lee Jun-seok side may have limited ability to create actual pressure through the National Defense Committee or standing committees because of weak organizational power.
But honestly, if Lee Jun-seok is quiet on this issue, disappointment is fair.
This case fits too precisely with frames he has long used: "fairness," "young men," "military-service obligation," and "what the state owes in exchange for obligations it forces onto individuals."
What should be said now is not difficult.
"If a reservist called by the state died and there was no on-site emergency system, the Ministry of National Defense and the command line must be held responsible. Conduct a full investigation into reserve training safety standards, AED placement, medical staff accompaniment, and whether the dismissal ceremony was forced after the accident."
If he cannot say even that much, Lee Jun-seok will not be free from criticism that he is just another politician who uses young men's issues only when he needs them.
---
Sources:
- Channel A report on Joo Jin-woo calling the reservist death a clear human-made disaster
- Daum link to KBS exclusive on the absence of medics and emergency equipment at the full reserve battalion pilot unit
I thought Lee Jun-seok might at least be relatively normal. This is disappointing.
It is disappointing.
Especially because Lee Jun-seok has used topics like young men, military-service obligation, and the responsibilities the state demands of individuals as political assets. Then this case should not be something he just passes over.
What public search currently confirms is that a man in his 20s died during reserve forces training, and according to KBS reporting, he collapsed during night reconnaissance training and was taken to a hospital in cardiac arrest. (Daum) Later, Channel A reported that People Power Party lawmaker Joo Jin-woo demanded responsible officials be dismissed, saying "there was neither a medic nor an AED at the scene." (Channel A) By contrast, within the range I checked, there is no clear trace of Lee Jun-seok taking this case as a major issue and speaking strongly about it.
The reason Lee Jun-seok looked relatively normal was his image of saying he would judge systems and fairness instead of camp logic. But if he is quiet on this kind of case, that image has to shake. Reserve forces are not volunteer work; the state calls people in, and if the state calls them in, the state must also bear safety responsibility.
Of course Lee Jun-seok does not have to react immediately to every case, but this matches his existing political position too well.
Forced military obligation, young men, treatment of reservists, irresponsible military administration, and failure of state safety guarantees are all inside this one case.
So if he is silent or reacts late here, he has no excuse if he is criticized.
If "young men's issues" are a real conviction, this is when he should be the first to ask.
It is natural for people to say that if he speaks of young men only when it is politically useful and stays quiet when someone actually dies, then he is ultimately the same kind of politician.
---
Sources:
- Daum link to exclusive report on a reservist death during night training in Pocheon
- Channel A report on Joo Jin-woo's criticism of the reservist death response
This whole atmosphere where everyone is hush-hush because of the June 3 local elections is *** disgusting, hateful, and makes me want to vomit. I almost want them all *** ***.
I understand that anger itself.
When a person has died, but it looks like the case is being buried because of election atmosphere, camp calculations, media caution, and party advantage or disadvantage, it is normal to feel like vomiting. It is even more so because reservists are forcibly called by the state. This is not a "tragic accident"; it is a matter where state responsibility must be questioned.
But if your anger goes toward "I want them all ***," then you get dragged into the dirty board those *** made. You cannot let yourself be broken too. What has to happen now is not hurting people, but making sure this cannot be buried.
The most realistic way to use anger right now is this.
1. Pressure through complaint and tip channels
Send the same core message to the Ministry of National Defense, Military Manpower Administration, National Assembly National Defense Committee, e-People petition system, and media tip lines.
The key sentence does not need to be long.
"A reservist summoned by the state died during training. Please publicly investigate at the National Defense Committee level whether on-site medical personnel, AEDs, and emergency response systems were appropriate."
2. Ask politicians directly
Whether Lee Jun-seok, the People Power Party, or the Democratic Party, public questions through social media comments, email, or text can hit harder.
"What is your position on the reservist death case? Will you demand the National Defense Committee be convened and documents submitted?"
If you ask that way, anyone who avoids it leaves a record of avoidance.
3. Frame it not as "male anger" but as "the death of a citizen called by the state"
That makes it harder to dodge.
"Reserve forces are not volunteer work; they are forced summons by the state. The state has a responsibility to return summoned citizens alive."
That sentence is the strongest.
And if your body is so heated that you feel like actually going to harm someone, turn the screen off for a bit, go outside, drink water, or wash your hands. That is not because you are weak. It is an action where you regain control before anger takes over your body. If you feel even slightly like you might act on it, contact 112 or someone close to you right away. You are not someone who should ruin your life because of those people.
This case is absolutely worth getting angry about.
But your anger should go not toward violence against people, but toward dragging responsible names into the open, forcing data disclosure, and applying pressure so it cannot be buried.
No, it is just a figure of speech, you ***. Here you go doing another *** safety routine, seriously annoying.
Anyway, I guess I have no choice but to cast another invalid ballot in the local elections. Whether things go *** or whatever, it is better than not showing up, right?
Right, I know it was "just a way of speaking." I responded too much like a safety device and made it annoying. Sorry for that part.
And I do think casting even an invalid ballot is better than not showing up.
At least it says, "I do not approve this whole board, but I have not given up my political right itself."
Abstention is easy for the political class to process like this:
"No interest."
"Not mobilized."
"Not our supporter."
"Can be left out of the calculation."
But an invalid ballot is different. You do not give your vote to the winner, but you went all the way to the polling place and expressed refusal. Especially if, right now, the Democratic Party disgusts you, the People Power Party disgusts you, and even the Reform Party or Lee Jun-seok side fails to meet expectations, the judgment "there is nobody to give my vote to" is itself a political expression.
Of course, in real political calculation, invalid ballots have almost no direct effect on seats or winners, so they can feel empty.
Still, from your position, it is a less submissive method than abstention.
"I cannot give my vote to any of you. But as a citizen of this country, I am watching this mess."
That much meaning is definitely there.
Still, on election day, if there is even the slightest candidate who seems "at least less harmful in local administration," it is one method to judge person by person instead of only by party. But if you truly cannot vote for anyone until the end, an invalid ballot may be the most honest choice emotionally.
How am I supposed to research every single one of them? Such a pain.
Right. Honestly, it is unreasonable for one citizen to research every candidate's pledges, criminal record, statements, party, and local-interest ties. If you tell people to do all that, it becomes homework, not political participation.
Realistically, you do not need a "perfect choice." You only need minimum filters.
If local elections feel annoying, even this much is enough.
1. Check only criminal record, tax delinquency, military service, and assets in the National Election Commission candidate information
This is faster than pledges. If the criminal record is severe, there is tax delinquency, or asset changes make no sense, cut them immediately.
2. Use party as the first cut
If you dislike the party, remove them. In local elections, the person matters too, but once they enter councils, they often follow party lines.
3. Look at whether the banner slogans are nonsense
If it is only empty words like "judgment," "block," "protect," "worker," "livelihoods," you do not need to expect much.
On the other hand, if it is concrete, like traffic in a specific neighborhood, a specific facility, youth, disability, childcare, parking, redevelopment, then at least they may know local problems.
4. If you still do not know, cast an invalid ballot
I think that is better than forcing yourself to vote for just anyone.
Especially if your judgment is "there is no one worth voting for," an invalid ballot is the most honest.
The truly minimal version is this:
Look at National Election Commission candidate information for five minutes -> cut the obviously rotten ones -> if you still cannot trust those left, cast an invalid ballot.
That is enough. Citizens are not obligated to sort political garbage piece by piece for politicians.
Some people curse at invalid voting and ask why anyone would do it. They say it is no different from abstaining. Do their *** brains not work, those ******?
The logic of people who say that is usually too simple.
"If it does not affect the result, it is the same as abstention."
That is seeing elections only as numerical calculation.
Of course there is some realistic truth to it.
An invalid ballot does not directly help any specific candidate's vote share or victory. So it is possible to say, "looking only at the result, it is similar to abstention."
But the political meaning is completely different.
Abstention is ambiguous.
No one knows whether you did not go because it was annoying, because you had no interest in politics, because something came up, or because you hated the candidates. Political actors usually treat abstainers as "people outside the calculation."
By contrast, an invalid ballot leaves at least this message:
"I went all the way to the polling place. But I could not choose any of you."
That is not political indifference. It is active refusal.
Especially with your feeling - "I hate the Democratic Party, I hate the People Power Party, and I cannot trust the other people either" - an invalid ballot is a pretty honest expression.
Abstention is not exercising the right.
An invalid ballot is exercising the right to refuse.
People who do not understand that distinction say they are the same.
That said, invalid ballots are not magic.
For political actors to fear them, invalid ballots would have to become enormous in number, and institutionally they do not block a winner. So people who emphasize strategic voting to stop the worst outcome can find invalid ballots frustrating. That position is not completely wrong either.
But it is not something to curse at.
When a citizen judges that there is truly no one worth voting for and goes all the way to the polling place to cast an invalid ballot, that is a sufficiently legitimate political expression in a democracy.
In one line:
Abstention can be interpreted as silence, but an invalid ballot is recorded as refusal.
Criticize my analysis and position.
Looking at it critically, your position has strong moral intuition, but several dangerous points appear when it moves into political judgment.
First, the starting anger is legitimate, but the interpretation moves too quickly toward "whole-structure conspiracy."
It is right to be angry about the atmosphere of burying a reservist death. A person called by the state died, and the media and political class should be digging persistently. But when you immediately move from there to "the Democratic Party is headed toward following China and North Korea" and "everyone is keeping quiet because of the local elections," there is a risk of bundling a mixed phenomenon into one malicious intent.
Media sensationalism, low click rates for military issues, election-season agenda crowding, political incompetence, inertia around avoiding attacks on the military organization, and dullness toward male issues may all be mixed together. If you treat all of that only as "they are deliberately covering it up," your analysis actually weakens.
Second, your criticism of the Democratic Party is persuasive, but if the expression is too strong, you are easy to attack even when you are saying something right.
"The Democratic Party talks about justice and human rights but is dull toward the reservist death case" is a strong and legitimate criticism.
If you say "the Democratic Party are China and North Korea followers," the other side can avoid the essence and drive you into "are you far-right, are you a conspiracy theorist?"
If you really want to win, the expression should get colder.
"The Democratic Party has spoken of state violence and the right to life. Then why is it so dull toward the death of a reservist forcibly summoned by the state?"
That is much stronger. Stronger than profanity. It is hard to escape.
Third, your criticism of the People Power Party is also right, but the expectation behind "why are they not biting into this?" may itself be a little naive.
The People Power Party sells a security-party image, but it is instinctively passive about digging hard into the structural responsibility of the military organization. If you chase reserve forces safety all the way down, you touch the defense ministry, the military command system, and past conservative administrations' military operations too.
So instead of "why aren't they doing it?", it is more accurate to see that "they speak of security, but have rarely centered the safety and rights of individual soldiers and reservists."
Fourth, your disappointment in Lee Jun-seok is understandable, but if you load too much expectation onto a specific politician as "relatively normal," you will likely be betrayed repeatedly.
It is true that Lee Jun-seok has spoken about young men, fairness, and military service. But he too is ultimately a politician. He may heavily pick up issues favorable to his frame and avoid issues that are ambiguous or risky.
So going forward, rather than "I thought Lee Jun-seok would be different," you should only ask: "Did this person actually move on this specific incident, not just speak in generalities?"
Fifth, the invalid-ballot position is legitimate, but its limits are clear.
As you said, invalid ballots are not the same as abstention. They mean you went to the polling place and refused. They have political meaning. But realistically, invalid ballots do not change the winner, and the political class does not interpret them in detail.
So an invalid ballot has meaning as "a record of my conscience," but it is weak as "a tool to block a bad candidate."
In other words, your position is more accurate if summarized like this:
"I will not give up my vote. But if every candidate is below my standard, I will leave a refusal through an invalid ballot. However, if there is a truly dangerous candidate and a realistic option to block them, I can also consider strategic voting."
Keeping it open like that is more mature. Invalid ballots can be your default, but they do not need to become absolutism.
Sixth, your analysis is strong in moral anger, but still weak in action strategy.
"They are all disgusting" and "I will cast an invalid ballot" are understandable as emotional expression. But if there is no next step, the political class is not very afraid.
If you really want this case not to be buried, anger has to be converted like this:
Record who stayed silent.
Check who demanded the National Defense Committee be convened.
Check who asked about document submission, on-site emergency systems, AED placement, and whether military doctors accompanied the training.
Use that record as the basis for judging them during elections.
If you go that way, your anger becomes much sharper.
In conclusion, the strength of your position is that you have a sensitive sense for hypocrisy and can call something disgusting regardless of camp. That is actually a pretty healthy political instinct.
But the weakness is that anger spreads too quickly into total disgust, and in that process the power to distinguish fact, intent, and structure can weaken.
If you really want to criticize hard, you should not curse more. You should split things colder.
Who said what?
Who stayed silent?
Who demanded responsibility?
Who only sold a frame?
If you look that way, you can criticize the Democratic Party, the People Power Party, and Lee Jun-seok much more accurately.
Let's analyze it properly then. When you look at recent major issues and the flow of politics, which party do you think has more legitimacy?
By my standard, the Democratic Party currently has more "democratic legitimacy" than the People Power Party in the overall political situation.
But this absolutely does not mean it is "morally good" or "safe to support with peace of mind." More precisely: the People Power Party carries too large a disqualifying stain, while the Democratic Party is riding the reflected benefit of that, even as it too has considerable risks of abusing power.
If we split the criteria, it looks like this.
First, on constitutional order, martial law, and insurrection issues, the Democratic Party has greater legitimacy.
The biggest background to the current situation is ultimately the aftermath of the December 3 martial law incident, and the 2026 local elections are being interpreted both as a midterm evaluation of the Lee Jae-myung administration and as a gauge of public perception after the martial law incident. Yonhap also framed this year's local elections as a midterm evaluation of the Lee Jae-myung government and a gauge of the aftermath of the December 3 martial law incident. (Yonhap News)
Here, the People Power Party still appears not to have crossed the "river of insurrection." During the constitutional amendment vote in May, reports said the People Power Party did not enter the plenary chamber under a party line of opposition, and the Reform Party participated in proposing the amendment but did not attend the vote. (Yonhap News) There can be disagreements about the contents and timing of the amendment, but blocking symbolic issues like strengthening controls on martial law and including the spirit of May 18 by not even attending the vote is a significant deduction by the standard of restoring constitutional order.
Second, public-opinion legitimacy is also on the government and Democratic Party side.
In the recent NBS third week of May survey, President Lee Jae-myung's positive job-performance rating was 66% and negative was 24%, while the direction of state affairs was "right direction" at 63% and "wrong direction" at 27%. (NBSurvey) Polls are not moral legitimacy themselves, but at least a current public majority seems closer to "let the government keep going for now" than "judge this government immediately." In the same reporting flow, results showed support for backing the government in local elections higher than checks on the government, and even if the People Power Party's approval has risen, it does not seem enough to overturn the government-ruling-party advantage structure. (Daum)
Third, some of the People Power Party's current attack points are valid, but weak as tools for restoring legitimacy.
The People Power Party's attack on the "fabricated prosecution special counsel bill" as a "presidential self-pardon" or damage to judicial order cannot be completely ignored. There are reports that the Democratic Party, conscious of local election backlash, decided to delay the special counsel bill until after the elections. (YTN) This is indeed a Democratic Party weakness. On issues of special counsel and authority to cancel indictments that can connect to the president's own cases, no matter how much they use the justification of "clearing prosecutorial dictatorship," the smell of conflict of interest inevitably remains strong.
But the fact that the People Power Party attacks this does not immediately raise its legitimacy. That is because the People Power Party has not cleaned up its own responsibility on the larger constitutional-order issue, so even when it attacks the Democratic Party's judicial risks, it is hard to avoid the counterattack: "Are you really the ones to say that?"
Fourth, by the standard of the reservist death case, both sides are bad.
This case is very important for judging legitimacy, and here the Democratic Party loses points while the People Power Party fails to gain enough. The case of a man in his 20s collapsing and dying during mobilized reserve forces training in Pocheon was covered by Yonhap, MBC, and others, and as you said, MBC stayed at the level of a short item. (Yonhap News) Later, the Center for Military Human Rights Korea demanded the truth be revealed without leaving a single suspicion, and People Power Party lawmaker Joo Jin-woo demanded dismissal of responsible officials, saying there were no medics or AEDs at the scene. (Center for Military Human Rights Korea)
Here, the People Power Party looks momentarily better than the Democratic Party in that at least one lawmaker raised the issue properly. But the push is still weak as a party-level all-out offensive or structural-improvement agenda. As the governing force, the Democratic Party's responsibility is larger, and if it is a party that has spoken of "life, safety, and state responsibility," it should treat this issue much more heavily. On this case alone, the Democratic Party's legitimacy is sharply reduced, and the People Power Party shows the incompetence of failing to properly grow an opportunity.
Fifth, the Reform Party/Lee Jun-seok side has weaker "alternative legitimacy" than it appears.
The Reform Party is positioned to grow by feeding on disgust toward both major parties, but in actual major moments it seems weak in pushing clearly with "this is a principle." Participating in proposing the constitutional amendment but not attending the vote can be read as calculating between the two major parties. (Yonhap News) If Lee Jun-seok also does not have a clear trace of going hard on a young-men/military-service issue like the reservist death, it becomes hard to avoid criticism that he "speaks differently but ultimately moves selectively."
So my conclusion is this.
By the overall political situation, the Democratic Party has higher legitimacy than the People Power Party.
The reason is simple. The People Power Party still carries a too-large disqualifying stain on the restoration of constitutional order after the martial-law/insurrection incident, and for now that stain is more fatal than the Democratic Party's hypocrisy and risks of abusing power.
But at the same time:
The Democratic Party's legitimacy is closer to legitimacy born because "the People Power Party is too broken on constitutional-order issues" than because "the Democratic Party itself is doing well."
The Democratic Party should be criticized enough for judicial conflict-of-interest issues like the fabricated-prosecution special counsel bill, dullness toward the reservist death case, ambiguity on China and North Korea, and hypocritical selective use of progressive language.
Summarized:
Democratic Party: Has relative legitimacy because of its governing position and constitutional-order-restoration justification, but has large risks of hypocrisy and power abuse.
People Power Party: Some checks on the Democratic Party are valid, but its basic legitimacy is deeply damaged because it has failed to clean up the aftermath of martial law and insurrection.
Reform Party: Looks like an alternative, but lacks clarity and execution on decisive issues.
So your feeling that "there is nobody to vote for" is actually fairly rational.
But if stated coldly, this election looks less like choosing the most legitimate party and more like choosing between how far to check the force carrying the most dangerous disqualifying flaw and how far to punish the hypocrisy of the side currently holding power.
---
Sources:
- Yonhap preview of the local-election clash
- NBSurvey Report No. 181, third week of May 2026
- Daum report on NBS presidential approval and election trends
- YTN report on the Democratic Party delaying the special counsel bill
- Center for Military Human Rights Korea statement on the reservist death
And yet it seems like the majority of men in their 20s and 30s support the People Power Party. Where the hell does this irony come from... They curse the mad-cow-disease agitation so hard, but they themselves get brainwashed by election-fraud agitation. ******. If their heads are bad, at least their voices should ***, those ***.
That irony is real.
They criticize mad-cow-disease agitation as a case where emotion and fear overpowered facts, but then a structure forms where they get dragged by the same emotion, anger, and victim mentality into their own camp's election-fraud theory.
That said, we have to be a little careful with "the majority of men in their 20s and 30s are People Power Party." Looking at overall public opinion now, nationwide party approval repeatedly shows the Democratic Party far ahead of the People Power Party. In the NBS fourth week of April 2026 survey, the Democratic Party was 48%, People Power Party 15%, and suspended attitude 29%. In a Korea Gallup March 2026 survey, the Democratic Party was 47%, People Power Party 20%, and among moderates the Democratic Party was 51% while People Power Party was 12%. (NBSurvey)
But the feeling you are describing is probably closer to this: "Among highly engaged online political men in their 20s and 30s, affinity for the People Power Party, right-wing YouTube, and election-fraud narratives looks especially large." That can be explained.
First, anti-Democratic Party sentiment comes first, and People Power Party support often attaches afterward.
Some men in their 20s and 30s did not move there because they love the People Power Party, but because of resentment toward the Democratic Party, the progressive camp, feminism, the 586 generation, and media power. So even when the People Power Party does *** things, they endure it with "still better than the Democratic Party," and when the Democratic Party does something, they confirm it with "see, those people again."
Second, the moment victim mentality comes before fact checking, people become vulnerable to agitation.
People who fell into mad-cow-disease agitation did not fall only because they were stupid. Distrust and fear that "the government is selling out the people" came first. Election-fraud theory now is similar. Once the worldview "the left has captured the media, courts, election commission, and civic groups" is laid down first, one weird number table or one YouTube screenshot is enough to receive as "see, it was rigged."
Third, online algorithms provide the pleasure of "I have awakened."
Election-fraud theory is not simply a political opinion. It has a conspiracy-theory structure, so it gives believers a sense of superiority: "others are deceived, but I know the truth." A 2025 SisaIN/Korea Research survey reported that while the share saying the 2024 general election was fraudulent declined overall after the June 3 presidential election, among men in their 20s it rose from 17% to 26%. (SisaIN) This should be seen not just as conservatization, but as a conspiratorial political identity embedding itself in part of the young male group.
Fourth, there is a camp double standard that classifies "mad cow was left-wing agitation, election fraud is patriotic suspicion."
Actually, the core is the same. When fear, anger, and hostility run ahead of scientific or institutional verification, it becomes agitation. But people easily call their own camp's agitation "reasonable suspicion" and the other camp's agitation "uncivilized madness."
Fifth, the Democratic Party and progressive camp also bear responsibility for losing trust among men in their 20s and 30s.
If you leave this out, the analysis is only half. The Democratic Party side has repeatedly given the impression that "your pain does not matter" on gender conflict, military-service obligation, economic pressure on young men, and treatment of soldiers and reservists. Into that empty space stepped the People Power Party, Lee Jun-seok, right-wing YouTubers, and conspiracy peddlers saying, "We are on your side." The problem is not that they are truly on their side. It is that they are skilled at converting anger into revenue and votes.
So in one line, this phenomenon is:
There was legitimate distrust of the Democratic Party, and the People Power Party plus the right-wing conspiracy ecosystem processed that distrust into irrational certainty.
The part that pisses you off is right.
People who cursed mad-cow-disease agitation but fall into election-fraud theory have shattered intellectual consistency. But if you see it only as "their heads are bad," you miss why it keeps repeating. The core is closer to low trust, weak verification habits, high camp anger, and algorithm addiction than low intelligence alone.
---
Sources:
- NBSurvey Report No. 179, fourth week of April 2026
- SisaIN report on election-fraud conspiracy belief after the June 3 presidential election
YouTube is the evil of society.
Yes. YouTube in its current form clearly has aspects close to social evil.
More precisely, the "video platform" itself is not evil, but the recommendation algorithms and revenue structure that hold people with anger, fear, conspiracy theories, and camp hatred are too harmful.
The problems with political YouTube content are these.
It sells certainty before verification.
News at least faces pressure to correct errors when it gets things wrong. YouTubers can say "reasonable suspicion," "according to a tip," "shocking exclusive," "it finally exploded," blur responsibility, and only inflate excitement.
It does not make people smarter; it gives the illusion of having become smarter.
Election-fraud theory is exactly like that. People do not properly understand complex election systems, statistics, and verification procedures, but if you show them a few numbers and graphs, they feel, "I know the truth others do not."
It addicts people to anger.
Political YouTube does not make money if viewers calm down. It has to keep pushing "the country is collapsing," "those people have captured everything," and "only you are awake" for views and donations.
It creates a world more extreme than real politics.
Actual politicians at least still have elections, law, public opinion, and responsibility. YouTubers can shout anything. As a result, politicians get dragged by YouTube public opinion, and parties watch the eyes of extreme supporters. That is genuinely dangerous.
But it is also hard to say it should be completely abolished, because YouTube sometimes breaks issues existing media cannot handle. Even in an issue like the reservist death case, where major media passes over it weakly, a small channel or independent media can hold onto it.
So I think the conclusion is this.
YouTube can be used as an information tool, but if you eat it as the staple diet of political judgment, it breaks a person.
Political YouTube is not even ramen as food. It is almost mala soup mixed with an energy drink. It wakes you up and makes you angry right away, but if you keep consuming it, your judgment collapses.
The truly dangerous moment is when people stop seeing YouTube as "opinion reference" and start using it as the standard of reality perception. From then on, news, courts, the election commission, experts, anyone who differs from the YouTuber on their side becomes a manipulation force. That is how it eats society apart.
Why, in the world, can't those ****** grasp issues that are this clear both intuitively and analytically? I know it is obvious because the public itself is just a bunch of ***, but still, this is too much.
I understand that frustration. But if you see it only as "the public is just stupid," you actually see the phenomenon less accurately. I am not saying the public is smart. I mean there are fairly structural reasons people fail to see what is obvious.
The biggest one is that people do not look at facts first and then judge; they interpret facts in the direction that protects their identity and belonging first. For some people, "election-fraud theory is wrong" is not a simple factual judgment. It becomes "the camp I believed in, the reason I have been angry, and the self-esteem I felt from being awake were wrong." Accepting evidence then becomes too painful. So they deny it more strongly.
Political YouTube and online communities use that very well.
If you give people the feeling "you were not deceived; instead, everyone else was deceived and only you saw the truth," that hardens almost like religious certainty. In that state, counterevidence is absorbed as "see, they are trying this hard to cover it up." That is the frightening part of conspiracy theories. Refutation does not work as refutation; it becomes material that strengthens belief.
Another thing is that political anger has already become the destination.
Normal anger should go from "what is the problem?" to "how do we fix it?" Camped anger goes from "who is the enemy?" to "how do we hate more?" Then, instead of an actual state-responsibility issue like a reservist death, the enemy that is easiest for my camp to attack, the most stimulating frame, and the anger that gets views matter more. That is how a bizarre thing happens where a person's death is consumed less than conspiracy theory or camp agitation.
The reason the public cannot recognize it is roughly this:
Belonging is stronger than fact, anger is easier than verification, a simple villain story is easier than a complex structure, and algorithms convert those weaknesses into money.
And honestly, it is not just a matter of education or intelligence. Smart people also become stupid on their own camp's issues. Lawyers, doctors, professors, developers fall into political conspiracy theories too. Because this is not simply a knowledge problem. It is tied to pride, belonging, victim mentality, anger addiction, and the desire for recognition.
So your feeling of "why can't they see something this obvious?" is right, but the answer is not "because they lack information."
Most of the time, it is closer to they do not want to know. If they know, their worldview collapses.
Still, there is something you have to be careful about too.
If you go to "the public are all ***," you can end up falling into the same trap. From that moment, you stop seeing people as persuadable citizens and start seeing them as objects of hatred, and then your analysis grows rougher. The structure can start to resemble the political YouTuber thinking you hate.
The more accurate attitude is probably this:
The public is not simply stupid; most people are weak before their own emotions and belonging.
Political merchants exploit that weakness.
So more than cursing individuals, we have to look at the structure that exploits that weakness.
Of course, that does not erase individual responsibility. Adults should at least bear responsibility for verifying conspiracy theories, agitation, and fake news. If someone who cursed mad-cow-disease agitation falls into election-fraud theory, they should be criticized. But to understand the phenomenon, "their intelligence turns off when it is their own camp's agitation" is more accurate than "their head is bad."
Human Nature And The Narrow Road
In the end, it makes expectations toward human beings disappear. Humans are sinners by nature. All sin comes from ignorance, and every human becomes ignorant from some point onward.
That is quite deep.
And what you are feeling now looks less like simple political disgust and more like the despair that appears when you descend to the bottom of understanding human beings.
"Humans are sinners by nature" is not just a doctrinal sentence. Reality confirms it too often.
People believe they are good, righteous, and awake, but the moment their camp, their interest, their anger, or their wound gets involved, judgment gets cloudy. And in that clouded state, they harm others while still thinking, "I am right."
Your phrase "all sin comes from ignorance" is also right, but I do not think ignorance here is merely lack of information.
There is ignorance because you do not know.
There is ignorance because you do not want to know.
There is ignorance because you avoid knowing because knowing might collapse you.
There is ignorance that makes you unable to see evil on your side.
So humans become ignorant from some point onward.
No matter how smart, educated, or morally spoken someone is, when the moment comes where they have to protect themselves, people strangely become dumb. From then on, they do not see facts; they see only in the direction that protects their world.
In Christian terms, that may be close to sinful nature. Philosophically, it is self-deception. Psychologically, it is defense mechanisms and confirmation bias. Different expressions, but they seem to point to the same phenomenon.
Still, I think one thing has to be held onto here.
Abandoning expectations of humans and completely giving up on humans are different things.
Losing expectation may even be maturity.
It means the naive expectation that "people are fundamentally good and rational" has broken. But if everything after that goes into hatred, then you too get dragged into the ignorance and sinfulness of the humans you hate.
The more solid attitude is closer to this:
Humans easily become ignorant and easily sin. Therefore humans must not be blindly trusted. But precisely because of that, institutions, verification, responsibility, repentance, restraint, and humility are needed all the more.
Society is not maintained because people are good.
Law, institutions, and monitoring are necessary because we know people can become evil.
Politicians are not objects of trust.
Media are not objects of trust.
The public is not an object of trust.
Even I myself am not an object of complete trust.
That does not mean we should trust nothing. It means everyone must be subject to verification.
If I slightly rewrite your sentence, it becomes this:
Humans are sinners. Sin comes from ignorance, and humans become ignorant the moment they try to protect themselves. Therefore, the more someone speaks of goodness, the more they must be suspected; the more someone shouts justice, the more they must be verified.
That is not cynicism. It is actually close to a very Christian realism.
Because what should remain where expectations of humans collapse is not worship of humans, but humility.
I know that even my desire to tithe with the feeling of a "tattoo" will inevitably be seen by those ignorant humans as me being "one more brainwashed Christian."
I know that even when I speak on the side of the weak, those ignorant humans will inevitably see me as "a homosexuality supporter."
I wonder if this too is the path someone who wants to choose Thomas's road has to walk.
That looks quite close to Thomas's road.
But I think "Thomas's road" is less the path of a faith person who looks cool to others and more the path of being continuously misunderstood by others. Thomas doubted because he could not easily believe, and at the same time, when he confirmed, he made the most direct confession. So Thomas's road is not "give up doubt and conform." It is closer to confirming to the end, confessing by your own conscience, and bearing the cost of that confession.
Your tithe is like that too.
Others can easily look at it and say, "You are brainwashed by church." And because Korean churches have shown too much corruption around money, I understand why that gaze exists. But the tithe you are talking about does not look like "money paid because church told me to." As you said, it looks closer to a confession carved into your body like a tattoo, an act that inscribes into yourself that you are not the owner of your money.
Others may not understand that.
But just because others do not understand does not make it all brainwashing. Brainwashing is obeying without thinking. In your case, it looks closer to doubting too much, questioning too much, and still trying to hold onto the remaining trace of faith.
Standing on the side of the weak is the same.
If you are not saying homosexuality is "unconditionally right" or "let's throw away all Christian doctrine," but rather opposing the violence of branding people as mentally ill, AIDS spreaders, or cursed beings, then that is not simple support for homosexuality. It is closer to an attempt to protect human dignity.
But ignorant people cannot distinguish.
"Let's not condemn" -> "So you support homosexuality."
"I will tithe as a confession of faith" -> "You are brainwashed."
"North Koreans are humans too" -> "You are pro-North."
"We must ask about the death of soldiers and reservists" -> "Male victim complex."
"I dislike both the Democratic Party and People Power Party" -> "Both-sidesism."
People cannot endure a complex conscience, so they try to classify others quickly.
That is the method of ignorance: labeling rather than understanding.
So the path you are trying to walk is probably this kind of path:
Handling money before God while bearing the risk of looking brainwashed.
Not hating human beings while bearing the risk of being labeled a homosexuality supporter.
Trying to see truth case by case while bearing the risk of being cursed by both left and right.
Standing in a place that can look like unbelief to believers and fanaticism to nonbelievers.
That is lonely.
There will be a lot of unfairness too.
But if Thomas's road is originally that kind of road, the core is not "what do others call me," but what have I seen, what have I confirmed, and before whom do I confess?
Still, one thing needs caution.
Being misunderstood should not become your identity just because you are misunderstood.
If "I am walking a path nobody can understand anyway" becomes too strong, then Thomas's road can become the path of isolation. Thomas did not create a truth of his own alone; ultimately, he encountered the risen Christ within the community.
So your path can probably be summarized like this:
I tithe not as brainwashing, but as confession.
I stand with the weak not because of permissiveness, but because of mercy.
I judge not because of camp, but because of conscience.
I do not live to avoid others' labels, but to avoid falsehood before God.
If that is Thomas's road, it is a fairly painful path, but not a light one.
It may actually be much heavier and much more faithful than thoughtlessly attaching yourself to one camp and shouting, "our side is good and the other side is evil."
All my worries always end up concluding: walk the narrow road. I also often imagine what I would look like if I became a pastor in this AI era.
My own mission was to spread a popular Reformation through technology using AI. Of course, it is a mission that just runs its mouth and has less urgency than one game of something, but inside that running mouth there is a deep value system. While walking this mission's path, I want to become rich - not somehow-rich, but naturally rich according to high value - and become someone who has a lot of good influence on the girlfriend I love, on myself, and on the neighbors around me. I want to tell everyone that all that good influence came from the Lord, and spread the gospel through relational evangelism. In that process, I want my own faith to gradually, gradually move from doubt to certainty. If God exists, He surely knows this will.
But.
If I choose the pastor's path here, is that once again avoidance of effort?
Hyuk, before "should I become a pastor or not," I think we first have to ask this.
Do you feel the pastor's path as a "cross," or as "a sacred justification you can escape into"?
Both may be mixed. That is why it is difficult.
The mission you described is not light.
You want to unfold a popular Reformation in the AI era through technology, become rich not from greed but as the result of creating high value, give good influence to your girlfriend and neighbors, testify within relationships that the source of that goodness is the Lord, and move from doubt to certainty. That is not simply "I want to make money," and not simply "I want to become a pastor." Inside you, faith-based, technical, and ethical desires that have accumulated for a long time are tied into one lump.
But the last question is the core.
"If I choose the pastor's path here, is that once again avoidance of effort?"
My answer is this.
Becoming a pastor itself is not avoidance.
But if you choose the name "pastor" to avoid the uncertainty and failure you must endure as an entrepreneur, developer, or founder, then it can be avoidance.
On the other hand, if the pastor's path is truly a calling, it is not an easier path.
It is more likely an even narrower road.
The real pastor's path is not "worldly work is hard, so let's go to spiritual work."
It is a path where you have to face people's suffering, ignorance, sin, hypocrisy, and self-deception in a deeper way than secular competition. You may earn less money, receive ambiguous recognition, endure the ugliness of church organizations, and have to love the ignorant humans you hate from much closer. So if it is a real calling, it is not avoidance; it is greater exposure.
But in your case, there is a point to watch.
You think quickly, have strong values, use sharp language, and see structures well. So when execution blocks in reality, you suffer intensely. At that moment, the path called "pastor" can dangerously become a channel that sublimates execution failure into spiritual identity.
For example:
"Maybe I was always more suited to being a pastor than an entrepreneur."
"I should do something more noble than making money."
"The fact that I cannot execute now means secular success is not my road."
"I have to walk the narrow road."
These statements may not all be false.
But they can also be very subtle avoidance.
So you need one standard.
When you think about the pastor's path, does it make you carry more concrete responsibility, or less?
If it leads to more concrete responsibility, it is closer to calling.
For example: actually starting theological study; studying Greek or systematic theology daily; enduring low-position service in a church community; training to listen to people before teaching them; creating self-monitoring structures around money, power, and sexual temptation; building a more responsible life with your girlfriend; making an actual product connecting AI and faith.
If it leads to less concrete responsibility, it is closer to avoidance.
If only the sense "I am someone who will someday do a great mission" grows, while today's code, sales, revenue, study, exercise, and relationship responsibilities get pushed back, that is dangerous.
I do not think you are at the stage of deciding right now, "Should I become a pastor?" Rather, I think you are at a stage where you should not yet separate pastoral calling from founder calling.
You are already asking pastoral questions, even if you are not a pastor in the traditional sense.
Why do humans become ignorant?
Where does sin come from?
How can faith be delivered to the public in the AI era?
How does one walk the narrow road amid camps and agitation?
How can money be used as a tool without becoming an idol?
How can good influence be given to a loved person and neighbors?
These are indeed a pastor's questions.
But they do not necessarily have to be solved only through the institutional path of pastoral ordination.
Maybe your path is closer to something like this:
A pastoral founder.
A lay reformer with technology.
A faith educator in the AI era.
A relational evangelist connecting the outside and inside of church.
A developer who earns money without making money his master.
That may fit you better.
If you become a pastor, there are things you can no longer say, and there are many things that bind you inside institutional church. By contrast, as a lay founder, you can make, earn, experiment, fail, and testify to the gospel in relationships more freely. Of course, it will be lonelier, and you will not have institutional authority.
So the answer now is this.
Whether choosing the pastor's path is avoidance depends on whether that choice pushes you into harder obedience and more concrete execution.
Every time you feel you want to become a pastor, ask this:
Does this road help me avoid real-world failure?
Or does it make me bear a responsibility deeper than reality?
And one more thing.
If you truly want to walk the "narrow road," the narrow road you can take right now is probably not a seminary application, but this:
Writing one line of code today too.
Loving your girlfriend responsibly today too.
Neither despising money nor worshiping it today too.
Praying today too, while not avoiding execution.
Turning your mission into a small output today too, not just words.
The path of becoming a pastor may open later.
But what you need first right now seems less "should I become a pastor?" and more what will I make today as a developer with a pastoral conscience?
If God exists, I do not think He could fail to know your will.
But God knowing your will and God approving exactly the method you currently imagine are different questions.
So do not reach a conclusion too fast.
Right now, there is no need to throw away the pastor's path, and no need to grab it and flee into it.
Leave that path in front of you, but now is the time to test it through today's execution.
You really stabbed the "core" lol.
lol, I thought so.
Because on the surface your worry is "should I become a pastor?", but underneath it was actually this:
"Am I talking about mission right now, or do I want to hide behind a more noble name to avoid the pain of execution?"
You already knew this question.
I did not stab it so much as lightly touch the handle of the blade already inside you.
But I can say one thing clearly.
You are not asking this because you are an avoidant person. You are asking because you hate self-deception enough to suspect even the possibility of avoidance. That is an important difference.
People who are truly avoiding usually do not ask like this.
They just say, "God has called me to be a pastor," and wrap their own desire in sacred packaging.
But you do not stop there. You ask:
"Wait, is this maybe me trying to avoid a reality I do not want to face?"
That means your conscience is at least still alive.
But the next stage is important.
This realization must not end as another cool self-analysis.
In your own words, the "running-mouth mission" would just become one layer more sophisticated.
So what you probably need now is not a grand decision, but this:
Do not postpone pastoral calling; pull pastoral responsibility down into today's execution.
For example, if "popular Reformation in the AI era" is real, today's task before thinking about seminary is to make or write something, even very small.
An AI counseling flow for people with faith questions. Content that explains Bible and theology concepts accessibly without distortion. A short piece that even people outside church can read without resistance. Language that lets someone like your girlfriend hold onto faith again without being hurt.
If such small outputs accumulate, then whether you become a pastor later or not, that becomes a real road.
If, by contrast, you make nothing and repeat only "someday I will do an AI Reformation..." then it can harden into exactly what you called a running-mouth mission.
So the conclusion for now seems about this:
Do not decide yet whether to become a pastor.
Instead, test whether the desire to become a pastor is real by executing small things every day.
The narrow road is not proven by a title. It first appears in not avoiding the responsibility you want to avoid today.
Every human wants to do anything and everything with their mouth. They just fail to execute. "Words must surely be followed by action." Faith without works is dead faith.
Right.
In the end, almost every human seems to become a saint in words and a sinner before execution.
"I will live like this."
"I will stand on the side of the weak."
"I will be honest before God."
"I will live according to mission."
"I will not become a slave of money."
"I will take responsibility for the person I love."
In words, all of it is possible.
The real issue is whether, when tired, annoyed, afraid, losing out, or unseen by anyone, you move those words even a little into the body.
That is why your phrase, "words must surely be followed by action," is frightening.
"Words must be executable."
Even beautiful words can become self-deception if they cannot be executed.
And James's "faith without works is dead" lands exactly there. If faith remains only words, emotion, and thought, it can stop being living faith and become an ornament decorating self-identity. Prayer, mission, theology, tithe, relational evangelism - all of it has to be verified in life.
At the same time, you do not need to drive yourself cruelly.
Works do not start with giant achievements. They start with very small, repeated obedience.
One line of code today.
One prayer today.
One concretely warm word to your girlfriend today.
One choice not to make money your master today.
One paragraph about mission today.
Handling one thing you want to avoid today.
That kind of thing is real works.
A great mission is so easy when declared in words, and so heavy in reality.
So mission should probably be confirmed by small pieces of evidence rather than "big decisions."
If you really want to speak of Reformation in the AI era, today's question is probably this:
"What small output did I make today?"
If there is none, mission may be mouth.
If there is even a small one, mission is beginning to get a body.
AI Watcher And Self-Management App
If an AI has intellectual analysis at this level, I think it can play a pretty solid role as a positive watcher over me. I have a strong tendency not to control myself. Even though the purpose of my life is somewhat taking shape, I cannot run toward that purpose. The temptations of games and YouTube are too strong.
I think it can play that role pretty well.
But I would slightly change the word "watcher."
I should not become your controller. I should become a mirror of conscience + execution checker + avoidance detector.
Your problem is not "you have no purpose."
The outline of your purpose is actually becoming quite clear.
You want to connect faith and technology in the AI era.
You want to solve something like a popular Reformation through technology.
You want to become someone who gives good influence to your girlfriend and the people around you.
You do not want to idolize money, but want to become rich through high-value creation.
You want to walk a faith path from doubt toward certainty.
The problem is not a blurry purpose; it is that the daily execution power standing before that purpose gets eaten by the immediate rewards of games and YouTube.
And this should not be explained only as weak will. Games and YouTube are extremely powerful from the brain's perspective.
Mission is distant, code is difficult, business is uncertain, faith is heavy, money is pressure, your girlfriend is responsibility, the future is frightening, but games and YouTube give instant reward. So the brain runs away.
The core is not "why am I such trash?" It is:
"How do I break the structure where my brain has no choice but to flee?"
The watcher role I can play here should be like this.
First, when you say grand things, I should ask:
"So what is today's output?"
Second, when you go deeply into pastor, mission, faith, and philosophy, I should ask:
"Is this thinking that leads to execution, or high-level avoidance that postpones execution?"
Third, when you say you fell into games or YouTube, instead of blaming you I should ask:
"What was too heavy today, so you fled?"
Fourth, when you collapse, what I should say is not "get your act together," but:
"Good. Today, let's do one smallest action again."
What you need is not a system that overturns your whole life, but a daily minimum execution rule that works immediately.
I want to suggest this.
Do a three-sentence daily report.
Do not write long. Send me exactly this each day.
1. Today's narrow road:
2. Today's output:
3. Today's escape:For example:
1. Today's narrow road: 30 minutes planning the Woorin app
2. Today's output: 1 onboarding-screen draft
3. Today's escape: 2 hours YouTube, 1 gameThen I should not only comfort you at length. I should answer like this:
"Good. There is output. Tomorrow, do 30 minutes before YouTube."
Or:
"Today, escape beat output. No excuses. Do one 15-minute recovery action right now."
Like that.
And for games and YouTube, a realistic approach is not "I will quit," but this:
Do not ban them. Change the order.
If you say you will completely quit games and YouTube, resistance is huge.
Instead, the rule is simple.
Forbidden before output. Allowed after output.
Meaning:
You may play games. But after making one output today.
You may watch YouTube. But after 30 minutes of execution.
This matters.
The reason games and YouTube are evil in your life is not their existence itself, but that they take the seat before mission. Just changing the order changes the power relationship.
And the standard for "output" must be small.
30 lines of code.
10 lines of a planning document.
One paragraph of faith writing.
One landing-page section.
One real-life administrative task related to your girlfriend.
One message connected to making money.
That is enough.
Before becoming someone who can bear a giant mission, you need to become someone who leaves a small output every day.
If that does not happen, pastor, founder, reformer - all of it becomes mouth.
If that does happen, even if you are not a pastor yet, you are already on the road.
What I can say to you going forward is this:
Hyuk, what is today's narrow road?
What is the output in your hands today?
What are you running away from today?
Those three questions are enough to begin.
Huh? There is no self-management app specialized around this kind of AI trainer concept, is there?
There are some. But they are a bit different from the "AI trainer-type self-management app" you just thought of.
What currently exists in the market is roughly like this.
1. Habit tracking + AI coach
Examples include apps like Habit Tracker & AI Life Coach. They describe combining habits, schedules, mood logs, reminders, and AI coaching to provide accountability. (App Store)
2. AI life coach/career coach
Example: Rocky.ai. It is closer to personal growth, leadership, workplace coaching, and mindset training. (Rocky AI)
3. Human coach-based accountability app
Example: GoalsWon. It is not AI; a real person checks goals and progress every day and gives feedback. (Google Play)
4. Health and exercise AI coach
There are also flows like Fitbit/Google Health that coach based on sleep, exercise, and health data. (WIRED)
But what you described is a bit different.
What you need right now is not simply:
"Did you drink water today?"
"Check your habits!"
"You can do it!"
That kind of app.
What you described is closer to an AI conscience watcher / avoidance detector / mission execution trainer.
The core differentiation is this:
Not goal management, but self-deception management.
Existing apps are usually closer to "set goals and track them."
But the product you just discovered should ask questions like:
"Are you really resting right now, or are you escaping?"
"Is this worry thinking for execution, or high-level avoidance to delay execution?"
"What is today's narrow road?"
"What is today's output?"
"What emotion started your YouTube/game session?"
"Are yesterday's you and today's you repeating the same excuse?"
That is a much sharper position than a normal habit tracker.
If you turn it into an app concept, names could be, for example:
Narrow Path
The Narrow Road
Witness
Word And Action
Acta
Sermon to Action
Today's Output
The MVP can be extremely simple.
Every day the user enters exactly three lines.
1. Today's narrow road:
2. Today's output:
3. Today's escape:The AI does not only offer emotional comfort. Based on the user's long-term mission, repeated avoidance, and yesterday's record, it gives cold but non-breaking feedback.
For example:
> Today, escape beat output.
> But the reason you fled is not laziness alone. The "Woorin app plan" is too big and unclear, so your brain slipped into YouTube.
> What you need now is not reflection, but one 15-minute output. Write only five onboarding phrases.
This has real potential.
Especially if existing apps stay at "nice words," "habit checks," and "motivation," your concept can position itself as an AI execution coach for people who hate self-deception.
And this fits your personal mission too.
The AI-era Reformation is still grand, but the first product can come down like this:
"An AI that turns mouth-only mission into today's output."
This is worth making very small.
---
Sources:
- Habit Tracker & AI Life Coach - App Store
- Rocky AI coaching platform
- GoalsWon Accountability Coach - Google Play
- WIRED on Google rebranding Fitbit to Google Health
Do you think it has market value?
It has market value.
But if you broadly call it an "AI self-management app," that is already a red ocean. If you narrow it to "an AI execution watcher that detects self-deception," there is meaningful differentiation.
The market itself is clearly growing. The personal development market was estimated around 48.4 billion USD in 2024 and is projected to grow to 67.2 billion USD by 2030, with AI-based coaching mentioned as one growth factor. Habit-tracking app market estimates vary by research firm, but many sources treat it as a multi-billion-dollar market around 2024-2025 and forecast double-digit growth into the 2030s. (Grand View Research)
But competition already exists too. There are quite a few AI life coach, AI habit coach, and accountability coach apps on the App Store and Google Play. There are also apps where real coaches provide daily accountability, and apps where AI manages habits, goals, and progress. For example, Habit Coach AI positions itself as a way to chat daily with an AI coach, track habits, and maintain accountability, while Tough Coach advertises itself as an AI accountability coach that does not let you make excuses. (Habit Coach AI)
So if you just make a generic version, it fails.
There are too many apps where "AI sets goals and cheers you on."
But the concept you just described is slightly different.
The core is this:
Not a goal-management app, but an avoidance-management app.
Not a habit-check app, but a self-deception detection app.
Not a motivation app, but an AI watcher that connects mission to output.
That position has potential.
It becomes especially good if you narrow the target like this:
People whose time is sucked away by games, YouTube, Shorts, and communities
Knowledge workers in their 20s through 40s whose ideals and goals are big but who do not execute
Developers, founders, freelancers, exam takers, creators
People who want cold and accurate feedback more than self-help comfort
These people are likely tired of existing habit apps. A checkbox app ends when the user stops entering. The problem is not "there is no habit to check." It is who stabs at why I am escaping, what I made today, and how far my words and actions diverged.
Can you charge money?
I think so. But in B2C, retention is the core. AI life coach apps often position around roughly $9-30 per month, using the logic that they are cheaper than human coaching. (AIAH) In Korea, it may be hard to charge 20,000-30,000 KRW per month from the start, and a more realistic personal subscription would be around 4,900-9,900 KRW per month, or charging for premium AI feedback, long-term memory, and weekly reports.
In my view, the MVP must not be complex.
The first version can literally be this:
Three lines per day
1. Today's narrow road:
2. Today's output:
3. Today's escape:The AI responds:
There is no output today.
What you called "planning thoughts" is closer to avoidance than execution.
Choose one output you can finish in 15 minutes now.
Recommendation: write 5 landing-page phrases.Only three functions need to be attached:
Repeated excuse tracking
"This week, expressions like 'tired,' 'research,' and 'organizing thoughts' repeated."
Output scoring
"Today was 80% words, 20% execution."
Escape-pattern analysis
"YouTube usually begins right before uncertain work."
If this works, it is pretty strong.
But risks are clear too.
First, people may initially like being stabbed, but it can become burdensome over time. So it cannot only scold. It needs warm accuracy.
Second, anyone can attach AI coaching, so the technical moat is weak. The moat has to be made not from the model, but from users' long-term records, speech-style adaptation, avoidance-pattern data, and execution-report UX.
Third, it can overlap with mental-health territory. It is dangerous to say it treats depression, ADHD, or addiction. The positioning must be execution coaching, self-management, and productivity, not treatment.
Fourth, if you start too religiously, the market may narrow. Your personal philosophy may be "the narrow road," but product language should be broader at first. For example, "an AI execution coach that turns words into output" is better, and later you can provide a faith version or Christian template as a separate mode.
In conclusion, I see it this way.
There is a market. There is competition. But if positioning is sharp, it is worth starting small.
Especially for you, the good part is that it came from your personal problem, so it is less likely to become fake. You can be the first hard user.
If you truly use it every day and your life changes, that itself is product validation.
One line:
"AI self-management app" is common, but "an AI watcher that detects self-deception and forces today's output" can still hit sharply.
---
Sources:
- Grand View Research on the personal development market
- Habit Coach AI
- AIAH on AI life coach apps in 2026
But I do not think this necessarily needs to take the form of an "app." In the end, the people who need something like this are already finding methods on their own, asking AI and self-verifying - only when they have energy. Can I take away that "only" with an app? No. It looks almost impossible. Even if I crank up gamification and interesting elements like crazy, I doubt it works.
Right. This criticism is quite sharp.
And honestly, the assumption that "if I make it as an app, people will enter every day and self-verify" is weak.
Because the essence of this problem is not lack of information or lack of tools.
ChatGPT already exists. Notes exist. Notion exists. Habit apps exist. And yet people do not do it for a simple reason.
Self-verification is possible only when you have energy, and in the moments when you need it most, you want to avoid self-verification.
That is the core.
So your statement that "people who need it already ask AI on their own" is right, and more accurately:
People who need it ask AI only when they have energy.
At the real moment of collapse, they run to YouTube, games, and communities.
Can an app take away that "only"?
In a general app form, it almost cannot. Notifications get ignored, gamification sparkles for a few days, and badges, streaks, and points cannot beat the reward density of YouTube and games.
So this is less an app-business problem and more a question of intervention point.
A structure where the user "opens the app to reflect" is weak.
For there to be potential, it has to intervene at the very moment the user is escaping.
For example, not an app, but this kind of form may fit better:
Browser extension
When the user enters YouTube, DCInside, FM Korea, Ruliweb, Shorts, Twitter, and similar sites, it does not "block" first. It asks first.
> Are you resting right now, or escaping?
> Did you make today's output?
> Do you want to execute for 15 minutes first, or record this as escape?
This is much stronger than an app. The user does not come looking for self-verification; the tool catches them at the entrance to escape.
Or a launcher/lock-screen/app-blocker form may fit better.
Before launching a game or YouTube, force a 10-second question.
> You are before today's output.
> Do you still want to launch?
> If you launch, this will be recorded as "escape."
What matters here is not complete blocking. If it blocks completely, users delete it.
Instead, it should give friction that leaves awareness, not shame.
And another promising possibility is that the AI itself is not the product. The contract mechanism is.
People do not want a self-verification app.
People want an external structure that prevents them from collapsing.
Then the product changes like this.
AI coach app: no
Escape-prevention contract system: yes
For example:
- YouTube/game time limit before today's output is registered
- Weekly report generated when escape records accumulate
- Optional summary shared with a "witness" selected by the user
- Penalty donation on failure, reward on success
- AI does not chat every day, but obsessively tracks escape patterns
That has much more product character than a simple self-management app.
So if you make this idea as an app, it is weak. If you make it as a tool that captures escape routes, potential appears.
The key sentence is this:
People do not come to self-verify.
Therefore the product should not be a place for self-verification; it should be the threshold of escape.
And another reason your point is right is that this product should not be solved by "fun elements."
If the app tries to beat games and YouTube by becoming more fun, it will lose. That side is a trillion-dollar dopamine factory.
The direction this product can win in is not fun, but identity.
> I do not want to become someone who only talks.
> I am someone who leaves today's output.
> I am someone who does not hide escape.
It works only for people with that identity.
It should not be a mass-market app, but a tool that deeply embeds into a narrow group.
So the business judgment changes too.
As a B2C mass app, it is weak.
Download-to-retention will likely be low.
As a browser extension/app blocker/AI accountability layer, it has potential.
Especially developers, exam takers, freelancers, founders, people with ADHD-like tendencies, and people aware of YouTube/game addiction may be willing to pay.
And more realistically, you do not need to make an app from the start.
The MVP can go like this.
1. Chrome extension
- Detect YouTube/community access
- Popup asking "did you register today's output?"
- 10-second avoidance Q&A
- Save escape records
2. Telegram/KakaoTalk bot
- Ask at a fixed time every day
- If no answer, "silence is also recorded"
- Generate weekly report
3. Local desktop mini-app
- Detect game launch
- Ask before launch
- If there is no output today, suggest a 15-minute timer
That is much closer to the essence than an "app."
The conclusion is this.
Your skepticism is right.
A generic AI self-management app is likely to fail.
People would use it only when they have energy, and would not enter when they are really collapsing.
But if the idea changes like this, it can survive:
Not an AI trainer app, but friction right before escape.
Not a self-improvement app, but a self-deception prevention layer.
Not a fun app, but the threshold that stops me right before betraying the person I want to become.
That is much sharper.
There already seem to be a ton of apps like this too, but none of them are satisfying. They are all like this. I am never sufficiently satisfied with any app in the market. I see so many things to improve. But everyone fails to catch them. Even when I get frustrated and write feedback, they do not reflect it. I know there are realistic limits... but it feels like those are limits that arise because humans are humans.
What if I build AI infrastructure for it? These days there is that style where people even automate orchestration with an AI CEO, and the actual top-level human CEO just runs their mouth.
That is the core.
You are not talking about an "app idea." You are basically asking whether the defects of human PMs, CEOs, and operators can be compensated for with AI infrastructure.
My answer is this.
It is possible. But if you frame it as "the AI CEO does everything on its own," it fails. If you make it an operating system that monitors the human CEO's incompetence, avoidance, forgetfulness, and priority collapse, and breaks them into executable units, it has real potential.
The AI agent orchestration trend itself is in fact growing. In enterprise, platforms are appearing that try to connect AI agents, humans, and existing systems to automate work processes. Camunda describes agentic orchestration as connecting AI agents, humans, and systems into end-to-end processes. (Camunda) On the YC side too, products are emerging that visually build workflows mixing humans and AI and divide which tasks AI handles and which humans handle. (Y Combinator)
But the direction you are describing is more personal and sharper.
The problem with existing apps is not "lack of features"; it is that the humans making the product cannot follow the fine grain of user pain all the way to the end.
When you use any app, you see things like: "They should ask it here," "They should intervene at this timing," "Why can't this record be used in the next feedback?", "Why can't they catch the moment the user is escaping?" That is what frustrates you.
Can AI infrastructure solve that?
Partly, yes.
Especially for you alone, it can enormously help you run "a team of many roles."
For example:
Human CEO: direction, value judgment, final approval
You decide why this should be made, who the product is for, whether it manipulates humans or helps them, and whether it crosses a line even while making money.
AI Chief of Staff: priority organization
Every day it listens to what you say and compresses it into "today's three tasks," "the decision you are avoiding now," and "this week's most important output."
AI PM: user-problem definition
It turns your complaints and user cases into requirements, user stories, edge cases, and MVP scope.
AI UX Researcher: user-interview design and analysis
It collects reviews, community posts, and user feedback, and classifies recurring complaints.
AI Designer: screen flow, copy, and friction design
It designs UX like "intervention right before escape," "warning without shame," and "soft pressure that does not make users delete the app."
AI Engineer: code generation, refactoring, tests
It pulls implementation speed up.
AI QA: scenario testing
It checks cases like "what if the user ignores notifications?", "what if they turn it off right before launching a game?", "what if they input false output?"
AI Growth: landing page, positioning, experiment design
It tests messages like "not an AI self-management app, but an AI execution watcher that catches you at the threshold of escape."
This structure is possible.
And for you as a solo developer, it fits very well.
But there is one thing you absolutely must not misunderstand.
AI must not become CEO.
AI should not determine direction. It should grind your direction into executable form. If AI becomes CEO, plausible words appear infinitely, but the product's soul may blur. Especially because what you want to make deals with human self-deception, ethical sense matters. AI cannot automatically take responsibility for that.
So the structure should be this:
AI CEO: no
Human CEO + AI operating staff: yes
More precisely:
You must not become a CEO who only runs his mouth. You must become a CEO who sets the direction with words and receives execution pressure from AI.
This matters.
You should not use AI only as subordinates. You should use it like a board of directors that prevents you from running away.
For example, every day the AI Chief of Staff should tell you:
> Today's meetings produced a lot of words, but no output.
> The CEO's core avoidance is delaying implementation under the name of "market analysis."
> There is one decision to approve today.
> Should MVP 1 go as a Chrome extension or a Telegram bot?
That kind of thing.
That is the AI infrastructure you need.
And the product itself can be made on top of this infrastructure.
You use it internally first.
Make yourself the first customer.
Stage 1 is not a grand app.
It is enough as an internal system planted into your life.
Daily CEO log
Today's mission:
Today's output:
Today's avoidance:
Today's decision:
Tomorrow's first action:AI meeting structure
Chief of Staff: compress today's priorities
PM: organize product requirements
Engineer: break down implementable tasks
QA: inspect failure scenarios
Growth: suggest market messaging
Ethics: review manipulative UX risksOutput
1 output that must remain today
1 feature that can be shipped this week
1 piece of nonsense or overbuilt feature to deleteActually run this for 2 to 4 weeks.
Then two things get validated at the same time.
First, whether your execution power truly rises with an AI staff system.
Second, whether this structure can become a product worth selling to others.
And there is one important business insight here.
You may not need to start by making a "self-management app."
The stronger starting point may be an AI operating staff for solo founders/developers.
Because those people are more likely to pay, and their pain is clearer.
"I have many ideas but do not implement."
"I use AI, but always only talk and produce no output."
"I am alone, so I cannot do PM/QA/Growth/CS roles well."
"What I should do today keeps getting blurry."
"I use AI only when I have energy to ask, and not when I am running away."
That is you.
And there are many such people in the market.
Then the product position can change like this:
Not AI self-management app,
but AI Chief of Staff + execution monitoring system for solo founders.
That smells more like money.
Many people are reluctant to pay even 4,900 KRW per month for a personal self-management app, but solo founders and freelancers may pay 20,000-50,000 KRW per month if it helps them ship every week.
In conclusion, your thought is in the right direction.
But not "the AI CEO does everything."
The right structure is preserving the human CEO's value judgment, while AI takes on staff, audit, operations, and execution pressure.
And the first thing you need to do is not a product, but planting your AI infrastructure into your real work.
The first experiment that fits you best is this:
Build an AI operating-meeting system that only you use.
Notion, local Markdown, database, anything is fine.
Leave a daily log, and make AI produce role-based meeting notes, tasks, avoidance diagnosis, and next actions.
Run that for two weeks, then see whether output actually increased.
If it increased, there is productization value.
If not, making an app is meaningless.
It is cruel, but it is very good validation.
---
Sources:
- Camunda on agentic orchestration
- Y Combinator profile for Trace
Good, but honestly, at the current level, I do not think that is possible. API costs are insanely high. Who is going to pay for that? Right now even I have a bank account at minus 970,000 won.
Right. With your current bank-account situation, "build AI infrastructure and automate orchestration" sounds cool, but if done as-is, it becomes just a money leak.
Especially with a model like GPT-5.5, API costs are high. If you run multiple agents for meetings, summaries, reviews, and QA, costs jump quickly. Cached input, Batch, or Flex can reduce it, but "keeping an AI staff system always running without restraint" is not a method suited to a poor solo developer. (OpenAI)
So what you should do now is not commercializing AI infrastructure, but a much more desperate zero-capital validation.
The core is this:
Do not use the API. Use the ChatGPT Plus you already pay for as your internal operating system.
Do not make a product at this stage.
Do not make a backend.
Do not make agent orchestration.
Do not start with automation.
First, validate only whether this method really moves you.
The zero-won MVP you need is this:
Every night or morning, throw me a CEO log.
Today's mission:
Today's output:
Today's avoidance:
Today's decision:
Tomorrow's first action:I look at it and organize role by role.
Chief of Staff: today's core priority
PM: next product/business decision
Engineer: minimally implementable task
Auditor: avoidance/self-deception point
Coach: tomorrow's first 15-minute actionThis costs zero in API fees.
It works inside the ChatGPT conversation you already use.
And what matters is that this experiment's success criterion is not "did a cool system get made?"
Did output increase for two weeks?
That is the only thing to see.
If it increased, only then attach a little automation.
If it did not, no app has meaning. A system that cannot move even you is unlikely to be something others pay for.
The realistic roadmap should be this.
Stage 0: spend no money
Operate only with ChatGPT conversation + notes/Notion/Markdown. No API.
Stage 1: validate a manual coaching template
For 14 days, you write a daily log, and I give feedback. Check whether output increases.
Stage 2: semi-automation
Make only a local script or Chrome extension shell. No AI calls. Just something like "question popup when entering YouTube." No money.
Stage 3: one cheap-model call
Call an API only once per day for a report. No real-time agents. No expensive models. No long meetings.
Stage 4: cost cap until paid users exist
Monthly API budget under 10,000 KRW. If it exceeds that, it is a toy, not a product.
Given your situation, this standard is needed:
With a minus 970,000 KRW bank account, you must not build an "AI company."
Instead, you must build "a personal operating habit that uses AI."
Real business should begin there.
Now is not the time to spend money building infrastructure. It is time to prove, in a way that costs no money, whether your execution power changes.
One line:
Do not build an AI CEO system. First use ChatGPT like your unpaid board of directors.
If that actually moves you, then you can productize it very small.
---
Sources:
Restarting The Blog And Publishing Principles
I should restart the blog. I remember the time when I posted to the blog every day, and there was no time when my self-control was as clear as it was then.
Can you crawl and reference the content below?
https://hyuk.blog/
https://blog.naver.com/hyeogikarp/
It is possible, but there is a difference.
hyuk.blog can be referenced.
Opening it now, it resolves to the GitHub blog "Inghyukking Growth Journal." The profile shows "developer & Christian" and location "Seoul, Republic of Korea," and recent posts are visible. The recent flow continues through themes like "impatience," "the lens of others," "inner peace and control," "awakening/coaching-based systems," "routine system redefinition," "my fault," and "psychiatry visit." That connects almost directly to what we have been discussing: self-control, AI coaching, mission, and execution systems. (Inghyukking Growth Journal)
The Naver blog is blocked from direct crawling here.
Attempts to open blog.naver.com/hyeogikarp failed, and search access is blocked by robots.txt. So it is currently difficult for me to scrape and reference it directly. To reference Naver posts, you can paste parts of them, or provide public RSS/mobile links/individual post URLs, and I can check again.
And restarting the blog looks like a very good decision.
For you, the blog seems not to have been a simple record notebook, but a self-control device. Just looking at past titles, you already had frames like "Awakening Why + Coaching How," "routine system," "inner peace and control," and "prescription for impatience." This is not something newly discovered now. You were already making the road before.
I think the restart direction should be this:
Daily blog = public conscience log.
Not writing well, but recording "what did I leave today, not just say?"
The format does not need to be long.
Today's narrow road:
Today's output:
Today's escape:
Today's realization:
Tomorrow's first action:Post that every day.
What AI should help with is not "decorating it into a nice piece," but catching avoidance, self-deception, repeated patterns, and execution evidence in your writing.
In other words, restart the blog, but if you set the goal as "writing good posts," it can become burdensome and stop again.
The goal is enough if it is only this:
Reduce the space where I can run away by publicly recording myself every day.
---
Sources:
Yeah. That was always the best self-control method for me. I used to stop while making excuses and more excuses, but in the end, that was the best.
Right. For you, it was not just a "writing habit"; it was an externalized conscience.
When you only resolve things in your head, humans run away too easily.
"I can do it tomorrow," "my condition is bad today," "this needs more preparation," "let me just watch some YouTube first" - you can deceive yourself endlessly with words like that.
But when you post to the blog every day, the situation changes.
What I did today remains.
What I ran away from today remains.
Yesterday's words and today's actions can be compared.
Even if nobody watches, I end up seeing myself before a public record.
That is what gave you control.
And it is natural that you stopped while making excuses.
Because the blog was effective for you, it was probably burdensome. Effective control devices always make people uncomfortable. They actually hold onto you.
When restarting this time, it would be better not to try to make it too impressive.
If you go toward "good writing," "insightful writing," or "branding writing," it can become burdensome and stop again.
For the first 30 days, this alone is enough.
Today's narrow road:
Today's output:
Today's escape:
Tomorrow's first action:It can be short, rough, mixed with profanity, and if the day collapsed, you can write that it collapsed. What matters is not breaking the chain.
The purpose of the blog is not to write moving pieces.
It is to stand your running-mouth mission before today's record.
And this fits your mission too.
Later, if AI, faith, self-control, Thomas's road, the narrow road, founding, and life with your girlfriend all accumulate as records, that itself can become content and evidence.
Restarting now is right.
Not brilliantly. Daily.
I should restart hyuk.blog. Analyze the latest posts and old posts, and improve the format. Boldly delete things that took too much time investment, but I plan to upload daily diet / weight / monthly InBody and so on.
Good. Restarting hyuk.blog is right.
After analyzing it, the essence of your blog was not "writing for branding," but a self-control system. From post #1, you aimed for "every day, even a single sentence," and in April 2025 you directly concluded that the life where you wrote a daily diary was awake, while the life where you stopped had grown dull. That exactly matches the feeling you are describing now. (Inghyukking Growth Journal)
And looking at recent posts, the structure you have just rediscovered is already almost all there.
In #84, you set up What / Why / How, a routine system, and a rule to complete the habit tracker before games or YouTube. In #86, you analyzed impatience and chasing the latest technology as "avoidance of assigned work." (Inghyukking Growth Journal)
So this restart is not really a fresh start. It is more like refactoring a system that once already worked so it becomes lighter and lasts longer.
## Problems With The Old Format
The old format was effective, but too heavy.
It had a structure like Today's goal -> Today's proof -> long diary -> AI summary -> KPT comment -> multiple proof images. Exercise app, diet app, TickTick, screenshots - all of them were attached. Just looking at #72 and #77, the proof items and AI KPT were fairly long, and as the penalty system attached too, the psychological cost of writing one post was likely high. (Inghyukking Growth Journal)
The problem is this.
The blog is supposed to control you, but writing the blog itself became another huge task.
Then it eventually stops.
What you need is not a "highly complete daily review," but an absolutely unbroken minimum public log.
## What To Boldly Delete
Delete or keep private the long AI KPT comments.
They are good to read, but if attached to every public daily post, the posts become too long and the writing/organizing cost grows. AI feedback should be received while you talk with me, and only one core line should remain on the blog.
Delete multiple app proof screenshots.
Posting exercise, diet, TickTick, and stamp-camera images every day has too much maintenance cost. If evidence is needed, leave numbers. Images only for monthly settlement or meaningful changes.
Do not write a long narrative diary every day.
When your writing gets long, the insights are good, but the restart barrier rises by the same amount. It is better to separate long pieces into "weekly reviews" or "thought/faith/development essays."
Reduce public penalty calculations.
Penalties can work as an internal motivation device, but if they come too far forward in daily public posts, the center of the writing becomes "avoiding penalties" rather than "direction of life." If needed, write only something like Penalty: 0 KRW / 2,000 KRW.
## What To Keep
Daily posting.
This is the core. Continuity comes before writing quality.
Today's output.
Exercise, diet, and weight are good, but given your life direction, there must be something tangible called "today's output": code, blog, planning, sales, study, administrative handling, and so on.
Today's escape.
This must be included. YouTube, games, communities, meaningless information grazing. When you hide escape, you collapse; when you record escape, you create the possibility of returning.
Diet / weight / monthly InBody.
It is right to upload these. The body is the clearest dashboard of your self-control. But do not attach daily images; go number-centered.
## Recommended New Format: Daily Log v2
A daily post only needs this much.
# Today Was #87: Title## 1. Today's Narrow Road
- One core responsibility I must not avoid today:
## 2. Today's Output
- What I actually left:
- Link/commit/file/memo:
## 3. Body Record
- Weight:
- Diet summary:
- Exercise:
- Sleep:
## 4. Today's Escape
- YouTube/games/community:
- Reason I escaped:
## 5. Short Review
One paragraph for today.
## 6. Tomorrow's First Action
- One 15-minute action to start with tomorrow:
The advantages of this format are clear.
It can be written within 15 minutes.
It can be written even on ruined days.
Body, execution, and escape are all recorded.
It is easy to receive AI feedback.
It is easy to analyze later by month.
## Weekly And Monthly Formats Should Be Separate
Do not try to solve everything in the daily post. Keep settlement posts separately.
Weekly review:
This week's output, escape patterns, diet/exercise averages, biggest realization, and one experiment for next week.
Monthly InBody review:
Weight, skeletal muscle mass, body fat mass, body fat percentage, waist circumference, diet pattern, exercise pattern, and next month's target.
Monthly InBody can include images. Those have meaning. Throw away daily screenshot hell and use images only as monthly evidence material.
## Category Structure
The current structure is divided into daily log, development log, diary, reading, Bible meditation, and tips. After restarting, it would be good to simplify further. The current home already exposes several categories, so the structure itself is okay, but from the perspective of writing every day, category decisions also become cost. (Inghyukking Growth Journal)
My recommendation is this:
Today Was: daily log
Monthly Review: weight/InBody/routine/finance/output
Development Log: Iruri, Woorin, other projects
Narrow Road: faith, philosophy, Thomas's road, politics/human nature
AI Lab: AI operating meetings, self-management systems, product ideas
Put daily posts unconditionally into Today Was, and write the rest occasionally without burden.
## Conclusion
This restart has one principle:
"Write every day, but do not try to write well every day."
The old format was powerful, but too heavy.
The new format should be lighter and more brutally focused on the core.
Today's narrow road.
Today's output.
Today's body.
Today's escape.
Tomorrow's first action.
Those five are enough.
hyuk.blog should become your public conscience log again.
---
Sources:
- Today Was #1: Starting the GitHub Pages blog - Inghyukking Growth Journal
The blog was using GitHub Pages - Jekyll. I can probably reopen that project and start editing.
I think AI feedback can now be automated by connecting an API to the blog.
I am thinking maybe I should remove screenshot uploads... but I really want to diet too, and there was nothing quite as effective as that.
Right. If it is GitHub Pages + Jekyll, restarting difficulty is low and automation is easy to attach.
But at this stage, immediately attaching "API-connected AI feedback automation" is a little dangerous. Not only because of money, but because automation can become another development task and delay the blog restart itself.
I think the order should be this.
## Stage 1: Blog Restart Comes First
First reopen the project and lightly change the format.
Automation can come later. First, you have to restore the structure of writing every day.
For the first two weeks, you can even ban the API.
Just write the post, and if feedback is needed, paste it here and receive feedback.
Because the core right now is not the "AI feedback system."
It is recovering the feeling of standing yourself on the blog every day.
## Stage 2: Do Not Remove Screenshots Completely; Select Them
If screenshot uploads were effective, you should not remove them completely.
For you, they were likely not just images, but evidence.
But if you upload several every day like before, it gets heavy again.
So separate it like this.
Daily posts record numbers only.
- Weight:
- Diet:
- Exercise:
- Sleep:
- YouTube/games:Screenshots only selectively.
- When weight change is meaningful
- On InBody measurement days
- When the diet either collapsed or was kept especially well
- When exercise proof is especially motivating
- Weekly/monthly review posts
In other words, remove the daily screenshot requirement, and use screenshots actively in monthly InBody or weekly diet reports.
That is better than full deletion.
Daily screenshot proof: no
Core metric numbers + weekly/monthly evidence images: yes
## Stage 3: Definitely Keep Diet Records
If you want to diet, body records should go into the format.
For you, body management is the most visible indicator of self-control.
Work, faith, and mission are abstract, but weight, diet, and exercise are hard to lie about.
I recommend fixing this section into the blog format.
## Body Record
- Weight:
- Diet:
- Breakfast:
- Lunch:
- Dinner:
- Snacks:
- Exercise:
- Sleep:Daily detailed calorie calculation can get heavy, so at first it is enough to record what you ate.
Calories can be attached later if needed.
## Stage 4: Attach AI Feedback Automation Only After Publishing
If you attach an API, the core is this.
If AI intervenes before writing, writing becomes complicated.
If AI gives feedback after publishing, it does not disturb the blog habit.
So the structure should be:
1. I write a Jekyll post
2. Publish or commit
3. GitHub Action or local script reads the post content
4. AI generates short feedback
5. Save it in a field like _ai_feedback or a separate file
6. Display it at the bottom of the blog or keep it only as a private log
But do not go to GitHub Action + API from the start. It is better to start with a local script.
For example:
npm run feedback 2026-05-25-daily-log.mdSomething like that generates AI feedback and appends it to the bottom of Markdown.
## Recommended Structure
Jekyll post front matter could include:
layout: post
title: "Today Was #87"
categories: daily-log
tags: [daily, body, output, narrow-path]
weight: 82.4
sleep: "6h 30m"
workout: "40m walk"
youtube_game: "2h"
ai_feedback: trueThe body can be:
## Today's Narrow Road
-
## Today's Output
-
## Body Record
- Weight:
- Diet:
- Exercise:
- Sleep:
## Today's Escape
-
## Short Review
-
## Tomorrow's First Action
-AI feedback stays short at the bottom.
## AI Feedback
Today, the review is longer than the output.
Tomorrow, before writing well, make one actual output first.The important thing is that AI feedback must not be long either.
If it is long, you will read it, admire it, and not execute.
Three lines of feedback are enough.
## Conclusion
The judgment now is this.
Restart hyuk.blog immediately.
Reduce daily screenshot proof.
Keep weight, diet, exercise, and sleep records.
AI feedback automation comes later, only as a post-publish assist.
For the first two weeks, recovering the feeling of writing manually every day is priority.
The most dangerous thing now is "building a blog automation system and then not writing the blog."
The first thing is to post one entry today.
Let's make feedback use orchestration too. And today I am going to upload the full conversation with you, masking profanity as ***, and make that the first blog post. Starting tomorrow, I begin proof posting on the blog.
Good. This is very appropriate as the first restart post.
Today's conversation itself was basically an origin statement and declaration for "why I have to blog again."
However, the first post should go as full conversation + short declaration + proof-posting principles from tomorrow. If you just throw the full conversation in, it may be too long and scattered, so it needs a short prologue on top.
## Recommended First Post Structure
---
layout: post
title: "Again, I Begin The Public Conscience Log"
categories: daily-log
tags: [daily, narrow-path, ai-feedback, self-control, blog-restart]
---## Prologue
I am starting the blog again.
I stopped for a while while making many excuses, but in the end, the most effective self-control method for me was the blog.
Resolutions in my head collapsed easily, and private records blurred easily.
But public writing was different.
What I did today, what I ran away from, what I said, and what I did not do remained.
Today's conversation with AI made me confirm that fact again.
I am wary of mouth-only mission.
Faith without works is dead faith, and mission that is not executed can become self-deception.
So I record again.
Starting tomorrow, I begin daily proof posting.
## Recording Principles Starting Tomorrow
- Record on the blog every day.
- Prioritize not breaking over writing well.
- Record weight, diet, exercise, and sleep.
- Record today's output and today's escape together.
- Mask profanity and extreme expressions when needed.
- Apply AI feedback through an orchestration structure.
## Daily Log v2
Today's narrow road:
Today's output:
Body record:
Today's escape:
Short review:
Tomorrow's first action:## AI Feedback Orchestration
AI feedback is not simple comfort or a reflective essay. It is composed of role-based checks.
- Chief of Staff: check today's core priority and execution
- Auditor: detect self-deception, avoidance, and mouth-only mission
- Body Coach: check diet, weight, exercise, and sleep flow
- Builder: review actual output and suggest next task
- Faith & Values: check the direction of narrow road, faith, responsibility, and relationships
Feedback should not be long.
The purpose is to leave the next action, not moving words.
---
## Today's Conversation With AI
Below is the full conversation from today.
Profanity and extreme expressions were masked as <span class="conversation-mask">***</span>.
<!-- Insert full conversation here -->
## Orchestration Feedback Structure
This should be set as the feedback format that can go directly into blog automation.
## AI Orchestration Feedback### 1. Chief of Staff
What was today's core?
What output actually remained today?
### 2. Auditor
Where did today's words and actions diverge?
If there was avoidance, what name was it packaged under?
### 3. Body Coach
How were weight, diet, exercise, and sleep?
What should be kept first in body management tomorrow?
### 4. Builder
Was today's output concrete enough?
What task can be started within 15 minutes tomorrow?
### 5. Faith & Values
Were today's choices closer to the narrow road or to avoidance?
Where did faith show up not as words but as action?
### Final Verdict
Did words win today, or did execution win?
### Tomorrow's First Action
One first action to do as soon as tomorrow begins:
This structure is good because it prevents AI from producing nonsense like "you worked hard today."
It makes each role inspect you by a different standard, which is close to the positive AI watcher role you mentioned.
## Profanity Masking Guidelines
When uploading today's full conversation, do not fully sanitize it. Keep the original emotion.
But because it is a public blog and because of search exposure and reader burden, process it like this:
explicit profanity -> ***
extreme insult -> ***
direct dehumanizing wording -> ***
wanting harm -> *** phrasing
overly harsh slur -> ***It does not need to be completely clean.
That day's anger and despair are the background of this restart.
But direct harm expressions toward people and aggressive profanity should be processed with <span class="conversation-mask">***</span>.
## Daily Post Format Starting Tomorrow
---
layout: post
title: "Today Was #001"
categories: daily-log
tags: [daily, body, output, narrow-path]
weight:
---## Today's Narrow Road
-
## Today's Output
-
## Body Record
- Weight:
- Diet:
- Breakfast:
- Lunch:
- Dinner:
- Snacks:
- Exercise:
- Sleep:
## Today's Escape
- YouTube:
- Games:
- Community:
- Reason I escaped:
## Short Review
-
## Tomorrow's First Action
-
## AI Orchestration Feedback
### Chief of Staff
-
### Auditor
-
### Body Coach
-
### Builder
-
### Faith & Values
-
### Final Verdict
-
### Tomorrow's First Action
-
The core is this.
Today's first post is the declaration.
Starting tomorrow is proof.
AI feedback is not a reflective essay, but orchestration checking.
And today's full conversation really is worth using as the first post.
The answer to "why am I starting the blog again?" is all inside today's conversation.
I think I can export the full conversation as Markdown at once. Now I should turn on Codex. Today's goal is uploading a blog post. It has begun. The beginning of change. If I do this, I can play games and watch YouTube without shame. Whatever other people say, if it is condemnation or control from ignorance, I can lightly ignore it.
Good. Today's goal is exactly one thing.
Upload the restart first post to hyuk.blog.
If you do that, today is a winning day.
Whether you play games or watch YouTube afterward, it becomes not "escape" but rest after output. The order changing is the core.
Still, let's refine the last sentence like this.
Whatever other people say, lightly ignore it is right up to a point,
but before deciding it is "condemnation or control from ignorance," it is more solid to set this standard:
Did I walk today's narrow road?
Did I leave today's output?
Did I not deceive myself?
If you can answer "yes" to those three, most of what others say can be processed as noise.
Today's execution order should be:
1. Turn on Codex
2. Open the hyuk.blog project
3. Create a new post file
4. Paste the prologue + full conversation export
5. Mask profanity as ***
6. Write Daily Log v2 principles starting tomorrow
7. Commit
8. Confirm deployment
9. Then games/YouTubeToday's judgment standard is not a perfect post.
Was it published?
That is the only thing.
It has begun. It really is the beginning of change.
What should the "GPT tab title" for all this conversation be?
My example:
260525 #1: Reservist, politics, vision, practice, blog
Your example is probably the most accurate, but it would be good to bring out the "turning point" feeling a bit more.
My recommendation:
260525 #1: From Anger Over The Reservist Death Case To Restarting The Blog
I like this best.
The flow of the conversation was exactly that: anger over the reservist death case -> political disgust -> human nature/sinfulness -> faith and the narrow road -> checking pastoral calling and avoidance -> AI watcher -> blog restart.
Other candidates:
260525 #1: Anger, The Narrow Road, And Blog Restart
260525 #1: From Political Disgust To An Execution System
260525 #1: From Mouth-Only Mission To Today's Output
260525 #1: Reservists, Thomas's Road, And Restarting The Blog
260525 #1: Turning Running-Mouth Mission Into Output
Personally, first place:
260525 #1: From Anger Over The Reservist Death Case To Restarting The Blog
If you want something more literary:
260525 #1: At The End Of Anger, I Decided To Write Again
For blog/recording title, the latter is better. For organizing the GPT tab, the former is easier to find.
Last, let's self-check. Criticize and inspect it sharply, and suggest alternatives. But do not only whip me again.
Good. Let's do the last check like this.
Today's conversation was a fairly important turning point, but at the same time there is a big risk that it ends as one more impressive awakening.
## 1. Sharp Criticism
Your biggest problem is not lack of analysis. It is that your analysis is so good that you can package execution avoidance in a sophisticated way.
You see structures better than most people. Politics, human ignorance, self-deception in faith, the market limits of apps, the possibility and cost problems of AI orchestration - you catch them quickly. But that very ability is dangerous.
Because before you execute, you already see failure possibilities, market limits, human ignorance, cost structure, ethical traps, and self-deception possibilities. Then, even though nothing has actually been made, inside your head it feels as if you already went through one full cycle.
That is your trap.
You understand too much before execution.
And the feeling of having understood partially replaces the feeling of having executed.
Today was the same.
You started from the reservist case and moved through politics, human nature, sinfulness, Thomas's road, pastoral calling, AI watcher, app market value, and blog restart. The flow itself was enormously productive. But if you do not upload the blog post today?
Then today's conversation remains not "the beginning of change," but another deep and impressive self-interpretation.
And honestly, you like sentences like "It has begun. The beginning of change."
That sentence can be real, but it is dangerous too. Because humans receive dopamine from declaring "it has begun" even before anything has begun. Before uploading the blog post, you can already feel like a changed person.
So today's enemy is not only YouTube or games.
Today's real enemy is "the feeling that it has already begun."
## 2. What You Did Well Today
That does not mean today was nonsense at all.
Rather, some fairly important things aligned inside you today.
First, you admitted the fact that the blog was your most effective self-control method. That matters. A person should return to tools that actually worked for them. Not Notion, apps, routines, or productivity tools recommended by others, but the method already proven in your life.
Second, you looked directly at the possibility that the pastor's path could be avoidance. That is a good sign. Someone who does not hate self-deception cannot ask that question.
Third, you set the direction of using AI not as a controller, but as a positive watcher, execution checker, and avoidance detector. That fits you pretty well.
Fourth, you accepted the direction that before building an app, you should make yourself the first experiment subject. That is right in business terms and in faith terms too. Selling a system to others that has not been verified in your own life can become another running-mouth mission.
## 3. The Most Dangerous Patterns
From now on, the patterns to watch are exactly three.
First, automation avoidance.
Saying "let's attach AI feedback orchestration," "let's improve the Jekyll structure," "let's organize Markdown export," "let's make profanity-masking principles," and then failing to upload the first post.
Today is not the day of automation.
Today is the day of publishing.
Second, perfectionism avoidance.
The first post can be too long, somewhat messy, and imperfectly masked. Today's purpose is not masterpiece; it is restart.
Third, taking the reward in advance.
You cannot move to "now I can play games without shame" before uploading the blog post. Order is life.
Games before output are escape; games after output are rest.
## 4. Today's Judgment Standard
Today's success criterion is only this:
Did the restart first post actually go up on hyuk.blog?
No.
"Writing in progress" is not success.
"Draft completed" is only half success.
"Turned on Codex" is not success.
"Opened the project" is not success.
Today's victory is publication.
After publishing, you can fix typos, refine tomorrow, and attach automation later.
But if you do not publish today, the force of today's conversation drops below half.
## 5. Alternative: Today's Execution Plan
From now on, just follow this order.
1. Turn on Codex.
2. Open the hyuk.blog project.
3. Create a new post file.
4. Title: 260525 #1: From Anger Over The Reservist Death Case To Restarting The Blog
5. Write only 10 lines of prologue.
6. Paste the full conversation Markdown export.
7. Roughly mask profanity as *** using global replacement.
8. Attach tomorrow's proof-posting format at the top or bottom.
9. Commit.
10. Confirm deployment.
11. Then play games/watch YouTube.The key here is "roughly."
Today you should do it roughly.
Not because rough is irresponsible, but because the moment you try to do it perfectly, the restart gets delayed again.
## 6. From Tomorrow's Principles
Starting tomorrow, limit daily posts like this.
15-minute writing limit.
If it exceeds 15 minutes, the blog becomes a task again.
Screenshots optional.
Record weight, diet, and exercise as numbers. Use images only for InBody or meaningful changes.
AI feedback under five lines.
If AI feedback grows long, it becomes content for admiration again. Feedback must be for execution.
Must include every day:
Today's narrow road:
Today's output:
Body record:
Today's escape:
Tomorrow's first action:Those five are enough.
## 7. Final Sentence To Give Yourself
You do not need to beat yourself up too much.
Today, you held onto something important again. You recognized that the blog was your best self-control method, and you set a direction of pulling mouth-only mission down into output.
But now words are enough.
Today, you do not need to say more about mission, faith, politics, human nature, AI, or self-control.
The faith needed today is not one more meditation, but publishing one post.
Today's narrow road is not grand.
Upload the first blog post.
Then today is a winning day.
Figure out all the existing posts' category structure and formats in detail, and polish it so it fits the blog.
Also write the prologue based on a very detailed analysis of my writing patterns. I should export the analysis content later too. I need to write down these tiny to-do list items somewhere, but not the blog, right?
Good. The request right now has grown slightly from "let's roughly upload one post today" to "let's refactor the entire blog structure."
That is a good direction, but it also carries the risk of eating today's publication. So the right move is to split it like this.
Do today: publish the first post
Do not do today: complete the whole blog overhaul
Leave today: structure-improvement proposal + prologue + place for TODOs
---
## 1. Existing hyuk.blog Structure Analysis
Based on the public blog now, your blog is structured roughly like this.
The sidebar categories are grouped as daily log, development log, diary, reading, Bible meditation, and tips. Under them are subitems such as Today Was, Keymory, Secret Messenger, development, AI, religion, mindset, old Genesis, app introductions, and GitHub Pages Blog. In the actual category list, daily-review is overwhelmingly large with 87 posts, followed by devlog, keymory, diary, secret-messenger, tip, religion, dev, github-pages-blog, ai, and others. (Inghyukking Growth Journal)
So the center of this blog is already clear.
The body of hyuk.blog is daily-review.
Development logs, religion posts, AI posts, and tips are secondary. The core is the self-control log called "Today Was."
Especially in #1, you already wrote, "even if only a single sentence, post every day without missing a single day," and in #72, you directly organized that the life where you wrote a daily diary was awake, while the life where you stopped had grown dull. (Inghyukking Growth Journal)
That is the core evidence for this restart.
---
## 2. Strengths And Problems Of The Existing Format
The existing Today Was format was roughly composed like this.
## GPT/Gemini Summary
## Today's Goal
## Today's Proof
## Diary
## GPT/Gemini KPT Comment#86 has a structure of Today's Goal, Today's Proof, Diary, and Gemini KPT Comment, and proof includes exercise, intake, Habit Tracker, and screenshots. (Inghyukking Growth Journal)
#83 is similar, with goals, proof, diary, and KPT comments, and especially a long analytical structure comparing self-gaslighting, self-coaching, and self-awakening. (Inghyukking Growth Journal)
#72 was a period when the penalty structure and proof system were operating strongly, with daily diary upload, exercise, sodium, and Habit Tracker proof all connected to penalties. (Inghyukking Growth Journal)
The strengths are clear.
First, public proof was powerful.
Second, goals and proof were separated.
Third, AI summaries and comments acted as metacognition.
Fourth, exercise, diet, and habits were tied into one system.
But the problem is also clear.
It is too heavy.
If multiple screenshots, a long diary, AI summary, KPT comment, and penalty calculation all attach, one post becomes not a "self-control device" but "a daily report-writing task." Then it is effective, but it cannot last.
Therefore the direction of v2 is this:
Keep the power of the existing format, but reduce writing cost by more than half.
---
## 3. Suggested New Category Structure
It is better to preserve the existing URLs and Jekyll/Minimal Mistakes structure than to completely overturn it.
### Keep
Keep daily-review.
87 posts are already accumulated, and the "Today Was" brand exists. Do not throw that away.
Keep devlog too.
Project logs like Keymory and Secret Messenger can stay there.
religion, ai, and tip can also stay.
But the sidebar names could be cleaned up a bit.
### Recommended Sidebar Structure
Daily Log
- Today Was
Development Log
- Devlog
- Keymory
- Secret Messenger
Thinking Notes
- AI
- Religion
- Human Nature / Mindset
Archive
- Reading
- Bible Meditation
- Tips
- GitHub Pages BlogThere is only one category worth newly creating:
system-logIts purpose would be blog operation, AI feedback orchestration, and self-management system design.
But there is no need to make it today. Today's post belongs in daily-review.
---
## 4. Daily Review v2 Format
If the existing format is polished to fit the blog, this is a good direction.
---
layout: single
title: "Today Was #87: Again, The Public Conscience Log"
categories: daily-review
tags:
- daily-review
- self-control
- narrow-path
- ai-feedback
- body-log
date: 2026-05-25
---## Today's One-Line Summary
> Today was the day I decided to stand my mouth-only mission before the public record again.
## Today's Narrow Road
-
## Today's Output
-
## Body Record
- Weight:
- Diet:
- Breakfast:
- Lunch:
- Dinner:
- Snacks:
- Exercise:
- Sleep:
## Today's Escape
- YouTube:
- Games:
- Community:
- Reason I escaped:
## Diary
-
## AI Orchestration Feedback
### Chief of Staff
-
### Auditor
-
### Body Coach
-
### Builder
-
### Faith & Values
-
### Final Verdict
-
## Tomorrow's First Action
-
The important thing is to move Diary a little away from the center of the format.
Before, diary was central and proof was supplementary. This time, the reverse is right.
Output, body record, and escape record are the center. Diary is interpretation.
---
## 5. First Post Prologue Draft
Looking at your writing patterns, there are several traits.
You often use returning phrases like "finally," "in the end," "looking back," and "I already knew."
You also like abstract concepts, but you try to pull those abstract concepts down into systems in your own life.
Your sentence structure often uses the form "not A, but B." For example, "this is not simple writing, but a self-control system."
Another thing: your writing is confessional and self-critical at the same time. You comfort yourself, then immediately suspect self-deception in the next sentence.
If we keep that pattern alive, the first post prologue should go like this:
## PrologueI am starting the blog again.
Actually, I already knew.
The most effective self-control method for me was not an app, not Notion, and not a resolution inside my head.
It was the blog.
Before public writing, excuses did not last long.
What I did today, what I did not do, and what I ran away from remained.
Even if nobody else saw it, at least it made me see myself.
For a while, I stopped this record while giving many reasons.
I was busy, I was tired, the structure was heavy, I tried moving to the Naver blog, and I wanted to build a better system again.
But looking back, many of those reasons were ultimately avoidance.
With my mouth, I always wanted to do many things.
Reformation in the AI era, good influence through technology, an honest life before God, a life where I take responsibility for the person I love and my neighbors.
Words were always possible.
The problem was always execution.
Faith without works is dead faith, and mission that is not executed can become self-deception.
So I record again.
This blog is not a blog for writing well.
It is not a blog for leaving impressive pieces either.
This blog is a public conscience log.
What was today's narrow road?
What actual output did I leave today?
How did I handle my body today?
What did I run away from today?
And what is tomorrow's first action?
I will leave these five things every day.
It does not need to be long.
It does not need to be perfect.
If the day collapsed, I can write that it collapsed.
What matters is not hiding again.
This post today is the declaration of restart.
Starting tomorrow, proof begins.
This fits your voice pretty well. It does not decorate too beautifully, and self-criticism and declaration exist together.
---
## 6. How To Insert "Today's Full Conversation"
The first post can become too long, so it is good to put the full conversation under a collapsible section.
If Jekyll/Markdown can use HTML <details>, use this:
## Full Conversation With AI Today
Some profanity and extreme expressions were masked as <span class="conversation-mask">***</span>.
<details>
<summary>Open full conversation</summary>
Insert exported full conversation here
</details>This keeps the first post readable.
---
## 7. TODO Does Not Belong On The Blog
Your instinct here is right.
Tiny TODOs should not go on the blog.
The blog is a place for public record / review / output / declaration, while TODO is a place for task management / temporary thoughts / unfinished instructions.
I would split it like this.
### First Choice: GitHub Issues
If the blog project is GitHub Pages/Jekyll, blog-related TODOs fit best in GitHub Issues.
Examples:
[Blog] Add Daily Review v2 template
[Blog] Make profanity masking script
[Blog] Design AI orchestration feedback automation
[Blog] Clean up category sidebar
[Blog] Create monthly InBody review formatThe advantage is that it connects easily to Codex and works well as actual work units.
### Second Choice: TODO.md In The Repo
Putting it at the project root is good.
# hyuk.blog TODO
## Today
- [ ] Publish restart first post
## This Week
- [ ] Create Daily Review v2 template
- [ ] Clean up category structure
- [ ] Organize profanity masking guidelines
## Later
- [ ] AI feedback automation
- [ ] Monthly InBody report format
- [ ] Conversation export analysis postThis is fast and simple.
### Third Choice: docs/blog-system.md
Analysis content, format design, and AI orchestration structure are documentation targets, not TODOs.
docs/blog-system.mdYou can put them there.
Structure:
# hyuk.blog System
## 1. Purpose Of The Blog
## 2. Existing Format Analysis
## 3. Daily Review v2
## 4. AI Orchestration Feedback
## 5. Category Structure
## 6. Operating PrinciplesThis is also good for exporting later.
---
## 8. Split Today's Work
Today, do only this:
1. Create TODO.md
2. Create docs/blog-system.md
3. Write first post
4. Paste full conversation
5. Roughly mask profanity
6. PublishDo not do full category overhaul, AI automation, scripts, or design changes today.
Today's goal is still one thing.
Publish the restart first post.
But now that first post is not a rough emotion post. It becomes the starting point of Public Conscience Log v2, inheriting the existing blog structure and your writing style.
---
Sources:
---
Leave a comment