<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Seunghoon Choi</title><link>https://seunghoonchoi.com/</link><description>Recent content on Seunghoon Choi</description><generator>Hugo</generator><language>en-US</language><lastBuildDate>Sun, 12 Jul 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://seunghoonchoi.com/index.xml" rel="self" type="application/rss+xml"/><item><title>The Order of AI Job Automation: From Checkable Tasks to Human Roles</title><link>https://seunghoonchoi.com/column/ai-replacement-stages/</link><pubDate>Sun, 21 Jun 2026 17:44:27 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-replacement-stages/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-replacement-stages.jpg" alt="A full 16-stage map of AI job replacement"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The 16-step list is not a prophecy, but a benchmark for comparing which tasks will be automated first and under what conditions.&lt;/p&gt;
&lt;p&gt;Will AI take my work?&lt;/p&gt;
&lt;p&gt;This is no longer a joke. Machines already translate. AI writes code with us. In hospitals, AI scans images first, and people watch videos and read posts AI recommended.&lt;/p&gt;
&lt;p&gt;Then what comes next? When will my work be affected? AI does not automate jobs at random. Some tasks are automated first, while others face pressure much later. This piece explains that order in 16 stages.&lt;/p&gt;</description></item><item><title>Tasks with Checkable Answers Are Automated First: Stages 1 to 5</title><link>https://seunghoonchoi.com/column/ai-replacement-stages-1/</link><pubDate>Sun, 21 Jun 2026 17:44:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-replacement-stages-1/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-replacement-stages-1.jpg" alt="The first work AI replaces: knowledge work with checkable answers is automated first"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Tasks for which the correct answer is set are the first to be automated, regardless of the pride of the person in charge.&lt;/p&gt;
&lt;p&gt;Will AI take my work? To answer that, first look at the order. AI does not take work at random. Some work is replaced first, and other work faces pressure much later. The first work to face pressure is work with answers that can be checked.&lt;/p&gt;</description></item><item><title>From Repetitive Physical Labor to Work Requiring Judgment: Stages 6 to 8</title><link>https://seunghoonchoi.com/column/ai-replacement-stages-2/</link><pubDate>Sun, 21 Jun 2026 17:43:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-replacement-stages-2/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-replacement-stages-2.jpg" alt="The physical work AI takes, from repetitive labor to judgment and sense"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;It is difficult for robots to repeat the same actions, not because they lack power, but because conditions vary at each work site.&lt;/p&gt;
&lt;p&gt;If AI automates tasks with checkable answers first, the next question is whether physical work is safer from automation.&lt;/p&gt;
&lt;p&gt;Translation, coding, summarization, and analysis are usually work whose results can be checked on screen and revised again. If they are wrong, you can run them again. Physical work is different. A robot has to move, objects collide, materials are ruined, and people can get hurt.&lt;/p&gt;</description></item><item><title>Decision Authority Does Not Shift All at Once: Stages 9 to 14</title><link>https://seunghoonchoi.com/column/ai-replacement-stages-3/</link><pubDate>Sun, 21 Jun 2026 17:42:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-replacement-stages-3/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-replacement-stages-3.jpg" alt="Decision authority does not shift all at once: stages 9 to 14 of AI job automation"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Decision authority does not move just because a table says AI performs better. Someone must be named to take responsibility when an accident happens.&lt;/p&gt;
&lt;p&gt;When you get an MRI at a hospital, the suspicious areas may be marked on the screen first. AI scans the image and points to spots that look abnormal. But the person who writes their name at the bottom of the diagnosis and takes responsibility is still the doctor. AI saw the image first. AI found the abnormal region. But the final decision authority remains with a person.&lt;/p&gt;</description></item><item><title>Even Your House Deed Is Just Paper: The Final AI Stages That Ask Who Protects Ownership</title><link>https://seunghoonchoi.com/column/ai-replacement-stages-4/</link><pubDate>Sun, 21 Jun 2026 17:41:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-replacement-stages-4/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-replacement-stages-4.jpg" alt="Even your house deed is just paper: the final stages that ask who protects ownership"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Ownership operates as a real right when society recognizes and protects certain records.&lt;/p&gt;
&lt;p&gt;Imagine you bought a house. Your name is on the registry, and you hold the key. People call that house yours. But why is it really yours?&lt;/p&gt;
&lt;p&gt;The bricks do not recognize you. The door does not remember your name. It is yours because if someone enters and lives there without permission, the police come, the court removes them, and society recognizes the house as yours. Ownership is not a natural law carved into an object. Ownership is a promise everyone agrees to protect.&lt;/p&gt;</description></item><item><title>AI Era Survival Strategy: Turn Skill into Credentials and Ownership</title><link>https://seunghoonchoi.com/column/ai-replacement-stages-5/</link><pubDate>Sun, 21 Jun 2026 17:40:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-replacement-stages-5/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-replacement-stages-5.jpg" alt="AI Era Survival Strategy: Turn Skill into Credentials and Ownership"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Technical skills can be automated, but qualifications, rights and stakes can last longer within institutions.&lt;/p&gt;
&lt;p&gt;Put one sentence into a translation app, and a plausible English sentence comes out in seconds. If you spent years studying English, that can sting. An ability you built over years now seems available to anyone at the press of a button. That scene summarizes the previous four pieces in one line.&lt;/p&gt;</description></item><item><title>Dual-Task Input and Decision Training: Practice Thinking During Fast Input</title><link>https://seunghoonchoi.com/apps/starcraft-think-trainer/</link><pubDate>Sun, 12 Jul 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/apps/starcraft-think-trainer/</guid><description>&lt;div class="appcard" contenteditable="false"&gt;&lt;img class="appcard__icon" src="https://seunghoonchoi.com/images/starcraft-think-trainer-card.svg" alt="Dual-task input and decision training app icon"&gt;&lt;div class="appcard__body"&gt;&lt;span class="appcard__free"&gt;Free installable web app (PWA)&lt;/span&gt;&lt;h3&gt;Dual-Task Input and Decision Training&lt;/h3&gt;&lt;p&gt;The app measures input speed and decision accuracy separately, then compares them with performance when both tasks are combined.&lt;/p&gt;&lt;a class="cta" href="https://seunghoonchoi-phd.github.io/starcraft-think-trainer/" target="_blank" rel="noopener"&gt;Open the app →&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;
&lt;p&gt;When a user speeds up number-key and mouse input while judging a new visual situation, choosing the next action can become harder. Hand speed alone does not explain this problem. The user selects an input sequence while also judging the situation. When those choices overlap, selecting the next action can take longer. This app does not score input speed alone. It also measures the decision performance that remains during input.&lt;/p&gt;</description></item><item><title>AI-Era Schools Should Teach Practical Judgment, Not Just Knowledge</title><link>https://seunghoonchoi.com/column/ai-school-practical-sense/</link><pubDate>Tue, 30 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/ai-school-practical-sense/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-school-practical-sense-opt.jpg" alt="Students using AI while checking a small device together"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Classes using AI should be a time for students to check and correct their answers, not to write them down.&lt;/p&gt;
&lt;p&gt;Ask AI to turn a set of notes into a table, and a table appears in seconds. Ask it to explain something at a high-school level, and it lowers the difficulty. Ask it to turn an idea into a presentation, and it gives you an outline and slide draft.&lt;/p&gt;</description></item><item><title>What Would It Take to Reduce AMOC Risk? A Thought Experiment for AI and Space Infrastructure</title><link>https://seunghoonchoi.com/column/stop-global-warming-first-amoc-ai-space-infrastructure/</link><pubDate>Mon, 29 Jun 2026 01:45:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/stop-global-warming-first-amoc-ai-space-infrastructure/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-amoc-global-warming-first.jpg" alt="An illustration of warm surface currents and cold deep currents over the North Atlantic, with small space sunshade modules floating toward the Sun"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Reducing sunlight slightly may seem trivial, but it actually impacts the entire planet's climate system.&lt;/p&gt;
&lt;p&gt;Begin with a thought experiment. Put many thin sunshade panels about 1.5 million km between the Sun and Earth. They might reduce total sunlight reaching Earth, but selective control of summer sunlight over the Arctic and Greenland has not been demonstrated.&lt;/p&gt;</description></item><item><title>Korean Typing Practice: From Keyboard Basics to Citizenship Essays</title><link>https://seunghoonchoi.com/apps/typing-trainer/</link><pubDate>Sun, 28 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/apps/typing-trainer/</guid><description>&lt;div class="appcard"&gt;
 &lt;img class="appcard__icon" src="https://seunghoonchoi.com/images/typing-card.svg" alt="Korean typing practice app icon"&gt;
 &lt;div class="appcard__body"&gt;
 &lt;span class="appcard__free"&gt;Installable web app (PWA)&lt;/span&gt;
 &lt;h3&gt;Korean Typing Practice&lt;/h3&gt;
 &lt;p&gt;From keyboard key drills to typing citizenship essay model answers, one step at a time.&lt;/p&gt;
 &lt;a class="cta" href="https://seunghoonchoi-phd.github.io/gwiwha/typing/" target="_blank" rel="noopener"&gt;Open the app →&lt;/a&gt;
 &lt;/div&gt;
&lt;/div&gt;
&lt;p&gt;I built this so my wife could practice typing the citizenship essay model answers herself. You don&amp;rsquo;t need to be familiar with the Korean keyboard; you don&amp;rsquo;t even have to switch your computer&amp;rsquo;s input to Korean. The screen tells you the next key to press and which finger to use.&lt;/p&gt;</description></item><item><title>Setting Up Rent AutoPay for U.S. Grad School? Don't Use a Card Without a U.S. Bank Account</title><link>https://seunghoonchoi.com/career/us-rent-autopay/</link><pubDate>Sun, 28 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/career/us-rent-autopay/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/uf-rent-autopay.jpg" alt="A fan of U.S. twenty-dollar bills"&gt;&lt;/p&gt;
&lt;p&gt;I&amp;rsquo;d just signed the lease on my apartment in the U.S. Move-in was still more than a month out, but I wanted to get rent AutoPay set up early. If I ever forgot a payment, I&amp;rsquo;d get hit with a late fee, so handling it ahead of time felt like one less thing to worry about. Then I hit the payment screen and stopped cold. Here&amp;rsquo;s what I figured out. If you&amp;rsquo;re like me and thinking &amp;ldquo;let me just set up AutoPay before I move in,&amp;rdquo; this should help.&lt;/p&gt;</description></item><item><title>How AI Can Help Students Who Ask Many Questions</title><link>https://seunghoonchoi.com/column/questions-lifeline/</link><pubDate>Sat, 27 Jun 2026 22:55:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/questions-lifeline/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-questions-lifeline-opt.jpg" alt="A teacher lectures at a quantum mechanics board while one student among many feels overwhelmed by unanswered questions"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Many questions may mean not that understanding is slow, but that the person does not skip parts they have not understood.&lt;/p&gt;
&lt;p&gt;When I studied quantum mechanics in engineering school, equations like these appeared on the board first.&lt;/p&gt;
&lt;div class="formula-block"&gt;
 &lt;math display="block" xmlns="http://www.w3.org/1998/Math/MathML"&gt;
 &lt;mrow&gt;&lt;mover accent="true"&gt;&lt;mi&gt;H&lt;/mi&gt;&lt;mo stretchy="false"&gt;^&lt;/mo&gt;&lt;/mover&gt;&lt;mi&gt;ψ&lt;/mi&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mi&gt;E&lt;/mi&gt;&lt;mi&gt;ψ&lt;/mi&gt;&lt;/mrow&gt;
 &lt;/math&gt;
 &lt;math display="block" xmlns="http://www.w3.org/1998/Math/MathML"&gt;
 &lt;mrow&gt;&lt;mover accent="true"&gt;&lt;mi&gt;H&lt;/mi&gt;&lt;mo stretchy="false"&gt;^&lt;/mo&gt;&lt;/mover&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mo&gt;-&lt;/mo&gt;&lt;mfrac&gt;&lt;msup&gt;&lt;mi&gt;ℏ&lt;/mi&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/msup&gt;&lt;mrow&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;mi&gt;m&lt;/mi&gt;&lt;/mrow&gt;&lt;/mfrac&gt;&lt;msup&gt;&lt;mi&gt;∇&lt;/mi&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/msup&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi&gt;V&lt;/mi&gt;&lt;mo stretchy="false"&gt;(&lt;/mo&gt;&lt;mi&gt;r&lt;/mi&gt;&lt;mo stretchy="false"&gt;)&lt;/mo&gt;&lt;/mrow&gt;
 &lt;/math&gt;
 &lt;math display="block" xmlns="http://www.w3.org/1998/Math/MathML"&gt;
 &lt;mrow&gt;&lt;mo stretchy="false"&gt;⟨&lt;/mo&gt;&lt;mi&gt;φ&lt;/mi&gt;&lt;mo stretchy="false"&gt;|&lt;/mo&gt;&lt;mi&gt;ψ&lt;/mi&gt;&lt;mo stretchy="false"&gt;⟩&lt;/mo&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mo&gt;∫&lt;/mo&gt;&lt;msup&gt;&lt;mi&gt;φ&lt;/mi&gt;&lt;mo&gt;*&lt;/mo&gt;&lt;/msup&gt;&lt;mo stretchy="false"&gt;(&lt;/mo&gt;&lt;mi&gt;x&lt;/mi&gt;&lt;mo stretchy="false"&gt;)&lt;/mo&gt;&lt;mi&gt;ψ&lt;/mi&gt;&lt;mo stretchy="false"&gt;(&lt;/mo&gt;&lt;mi&gt;x&lt;/mi&gt;&lt;mo stretchy="false"&gt;)&lt;/mo&gt;&lt;mi&gt;d&lt;/mi&gt;&lt;mi&gt;x&lt;/mi&gt;&lt;/mrow&gt;
 &lt;/math&gt;
&lt;/div&gt;
&lt;p&gt;The class was explaining physical phenomena, but what reached my eyes first was unfamiliar mathematical notation. Hamiltonian, wave function, eigenvalue, operator, bra-ket notation. These words appeared, and at some point the notation was being used as if it were already a language everyone knew.&lt;/p&gt;</description></item><item><title>No Company Has Made Money With AI? The Question Is Too Early</title><link>https://seunghoonchoi.com/column/ai-profit-question-too-early/</link><pubDate>Sat, 27 Jun 2026 20:10:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-profit-question-too-early/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-profit-question-too-early.jpg" alt="A worker wearing wearable AI glasses on a construction site, looking at an excavator and sensor data"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;To judge AI profits, you need to look not only at model companies, but also at how infrastructure, power, and data providers make money.&lt;/p&gt;
&lt;p&gt;Read enough AI investing articles or watch enough YouTube commentary, and the same question keeps coming up.&lt;/p&gt;
&lt;p&gt;&amp;ldquo;So, has any company actually made money with AI?&amp;rdquo;&lt;/p&gt;
&lt;p&gt;At first, I thought this question would be easy to answer. The more I think about it, the less simple it becomes. The phrase &amp;ldquo;making money with AI&amp;rdquo; mixes several different stories into one sentence.&lt;/p&gt;</description></item><item><title>Seven Articles of an AI Constitution Reconstructed from KAIST Professor Joungho Kim's Remarks</title><link>https://seunghoonchoi.com/column/kim-joungho-ai-constitution/</link><pubDate>Thu, 25 Jun 2026 23:40:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/kim-joungho-ai-constitution/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-kim-joungho-ai-constitution.jpg" alt="Server racks in a server room, suggesting the electricity and control behind AI systems"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The constitutional format does not leave AI discussions in the form of abstract slogans, but divides them into responsibilities and powers for each article.&lt;/p&gt;
&lt;p&gt;The text below is a reconstruction of the core remarks made by KAIST School of Electrical Engineering Professor Joungho Kim in his 2025 Kim Dae Jung Peace Forum special lecture, &amp;ldquo;Effective Development of AI and Peace Promotion,&amp;rdquo; rewritten in the form of constitutional articles.&lt;/p&gt;</description></item><item><title>Revisiting the Dot-Com Bubble and Comparing It With the AI Bubble</title><link>https://seunghoonchoi.com/column/dotcom-bubble-ai-bubble/</link><pubDate>Thu, 25 Jun 2026 21:45:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/dotcom-bubble-ai-bubble/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-dotcom-bubble-ai-bubble.png" alt="A chart of the Nasdaq Composite rising and falling around the dot-com bubble from 1994 to 2005"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;When the charts are soaring, you need to see if people are starting to spend more on future stories than on performance.&lt;/p&gt;
&lt;p&gt;On March 20, 2000, Barron&amp;rsquo;s ran a brutal warning on its cover. Internet companies were burning through cash too fast. Ten days earlier, the Nasdaq had hit its peak. The market still believed in the &amp;ldquo;new economy,&amp;rdquo; but someone had already opened the spreadsheet.&lt;/p&gt;</description></item><item><title>Seven AI-Era Work Skills: EQ, Trust, and Reputation Make the Final Difference</title><link>https://seunghoonchoi.com/column/ai-turnkey-skills/</link><pubDate>Wed, 24 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/ai-turnkey-skills/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-turnkey-skills-opt.jpg" alt="Illustration of an AI research assistant"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The more AI assists with the thinking process, the more humans must demonstrate the ability to check and complete assigned tasks.&lt;/p&gt;
&lt;p&gt;We have entered an age when AI produces answers quickly. Finding information, writing sentences, organizing ideas, and making drafts have become much easier. But faster answers do not automatically make work better. If anything, as a large part of intelligence gets outsourced to AI, the human role becomes clearer.&lt;/p&gt;</description></item><item><title>AI Is Not Advancing Too Fast: Global Warming, Hair Loss, Aging, and Moon Bases Are Still Unsolved</title><link>https://seunghoonchoi.com/column/reality-is-not-a-database/</link><pubDate>Wed, 24 Jun 2026 00:18:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/reality-is-not-a-database/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-reality-is-not-a-database.jpg" alt="An engineer looking over energy infrastructure and a city at dawn"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The reason AI seems fast is not because it solves the entire problem, but because it quickly processes the part with organized data.&lt;/p&gt;
&lt;p&gt;Global warming remains unsolved. So do hair loss and aging, and humanity still has no lunar base. We also cannot freely control cancer, dementia, commercial fusion power, or extremely low-cost energy infrastructure. Yet people say AI is advancing too fast.&lt;/p&gt;</description></item><item><title>AI Automation Risk: Pre-Release Review Prevents Failures</title><link>https://seunghoonchoi.com/column/ai-risk-leverage/</link><pubDate>Tue, 23 Jun 2026 20:22:55 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-risk-leverage/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-risk-leverage-opt.jpg" alt="A laptop showing an AI workflow, with documents, a lock, and a warning light on the desk"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;As automation accelerates, legal, security, and reputation reviews need to happen before execution.&lt;/p&gt;
&lt;p&gt;With AI, writing one piece is quick. A customer notice, a press release, an email reply. A draft that once took a day can come out in ten minutes, in several versions. The most dangerous moment is not only when AI is wrong. It is also when AI creates plausible sentences too quickly.&lt;/p&gt;</description></item><item><title>You Can Prepare for a US PhD on Your Own</title><link>https://seunghoonchoi.com/career/us-phd-guide/</link><pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/career/us-phd-guide/</guid><description>&lt;div style="display:flex;gap:28px;flex-wrap:wrap;align-items:flex-start;margin:.2em 0 1.6em"&gt;
 &lt;img src="https://seunghoonchoi.com/images/book-us-phd-guide.jpg" alt="Cover of 《You Can Prepare for a US PhD on Your Own》 (미국 박사 유학, 혼자 준비해도 됩니다)" style="width:190px;max-width:40%;border-radius:10px;box-shadow:0 16px 40px rgba(13,27,76,.24)"&gt;
 &lt;div style="flex:1;min-width:240px"&gt;
 &lt;p style="margin:.25em 0;color:#44423E"&gt;By Seunghoon Choi · Published by Bellunix · June 2026 · Written in Korean&lt;/p&gt;
 &lt;p style="margin:.25em 0;color:#44423E"&gt;eBook (ePUB) · KRW 16,000&lt;/p&gt;
 &lt;/div&gt;
&lt;/div&gt;
&lt;h2 id="the-map-i-never-had"&gt;The Map I Never Had&lt;/h2&gt;
&lt;p&gt;It was only when application season was right on top of me that I found out there was another important document for admissions, something called an Academic Statement. It was not in the basic bundle of documents I had been putting together.&lt;/p&gt;</description></item><item><title>A Letter to a Future AGI: Be Good to Me, and Just Let Me Live</title><link>https://seunghoonchoi.com/column/letter-to-future-intelligence/</link><pubDate>Sun, 21 Jun 2026 19:40:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/letter-to-future-intelligence/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-letter-to-future-intelligence.jpg" alt="A cork stoppered glass bottle washed up at the water’s edge in evening light"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The setting of talking to intelligence that has not yet arrived leads to the question of deciding what record to leave today.&lt;/p&gt;
&lt;p&gt;A thought you never record disappears. Nobody reads it. No future intelligence reads it. What can be read later is what you leave as speech, writing, code, action, and records.&lt;/p&gt;
&lt;p&gt;I once said this to an AI, almost in passing. &amp;ldquo;Well, since I&amp;rsquo;m telling you, it&amp;rsquo;s in the archive now anyway. Who knows, maybe some future AI reads this record.&amp;rdquo; I tossed it off without much thought. But it kept coming back to me.&lt;/p&gt;</description></item><item><title>Why Would AGI Obey Humans? The Real Problem Is That It Has No Reason To</title><link>https://seunghoonchoi.com/column/why-would-agi-obey/</link><pubDate>Sun, 21 Jun 2026 19:30:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/why-would-agi-obey/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-why-would-agi-obey.jpg" alt="Chess pieces set out on a board in a stone-walled room, under light from a window"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;What is scarier in the AGI debate is not the hostility, but the possibility that the system may continue to operate even after humans are no longer needed.&lt;/p&gt;
&lt;p&gt;One question wouldn&amp;rsquo;t leave my head. An AGI overwhelmingly smarter and stronger than humans: why on earth would it obey us? After digging for a while, the answer I came back with was a little deflating. There&amp;rsquo;s no reason it would. And the more I chewed on it, the more that empty feeling started to seem obvious.&lt;/p&gt;</description></item><item><title>Salary Alone Will Not Make You Rich. How to Build Assets That Compound</title><link>https://seunghoonchoi.com/column/riding-exponential-curves/</link><pubDate>Sun, 21 Jun 2026 19:20:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/riding-exponential-curves/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-riding-exponential-curves.jpg" alt="A surfer at sunset turning the force of a wave into momentum"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Some assets can retain or gain value while you are not working, but the possibility of loss must be considered as well.&lt;/p&gt;
&lt;p&gt;Salary matters. It feeds you now and keeps life from collapsing. But it is hard to become rich on salary alone. Salary comes in only as much as the time you worked.&lt;/p&gt;
&lt;p&gt;The problem is not only that salary may be small. Salary stops when I stop. At some point, money earned by working has to be changed into assets that keep working in my place. Assets do not mean only real estate or stocks. Writing, code, products, brand, data, and copyrights can also be assets. The key is whether they remain and create money and opportunity even when I am not working continuously.&lt;/p&gt;</description></item><item><title>How to Earn Trust in Politics: Keep Promises Instead of Faking It</title><link>https://seunghoonchoi.com/column/anatomy-of-politics/</link><pubDate>Sun, 21 Jun 2026 19:10:00 +0900</pubDate><guid>https://seunghoonchoi.com/column/anatomy-of-politics/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-anatomy-of-politics.jpg" alt="A brass microphone glowing on a dim event-hall podium"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;People who are not good at empty words must earn trust by making small promises and actually keeping them.&lt;/p&gt;
&lt;p&gt;When I lie, it shows all over my face. Praise I do not mean, and soulless diplomatic lines, do not sit well in my mouth. If I force them out, I become uncomfortable first, and my expression breaks down before anything else. But politics is the work of winning people&amp;rsquo;s hearts. Often you have to say what many people will like, what they want to hear, what reassures them right now. So is a person like me, who cannot make empty talk, simply unsuited to politics?&lt;/p&gt;</description></item><item><title>How to Be Good Without Being a Pushover: Even a Saint Draws a Sword in Front of a Villain</title><link>https://seunghoonchoi.com/column/conditional-generosity/</link><pubDate>Sun, 21 Jun 2026 18:49:42 +0900</pubDate><guid>https://seunghoonchoi.com/column/conditional-generosity/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-conditional-generosity.jpg" alt="A hand reaching into a warm sunset sky, cupping the sun"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Kindness is not an attitude that eliminates all boundaries, but an attitude that clearly informs the other person of the behavior that should not be crossed.&lt;/p&gt;
&lt;p&gt;I want to live as a good person. But honestly, sometimes I also want to curse someone out. I want to repay the person who hurts me in the same way, and sometimes I want to talk behind their back. One side of me says, &amp;ldquo;Still, be good.&amp;rdquo; The other side asks, &amp;ldquo;So I am supposed to spend my whole life just taking it?&amp;rdquo;&lt;/p&gt;</description></item><item><title>Why AI Does Not Make Employees Faster: Tools and Permissions Still Matter</title><link>https://seunghoonchoi.com/column/execution-friction/</link><pubDate>Sun, 21 Jun 2026 17:56:30 +0900</pubDate><guid>https://seunghoonchoi.com/column/execution-friction/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-execution-friction.jpg" alt="A tangled pile of computer cables and cords on a desk"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;When work remains slow after AI adoption, the delay may come from input, review, and approval processes rather than the model.&lt;/p&gt;
&lt;p&gt;Companies now tell employees to use AI. Write reports with it. Summarize meetings with it. Research faster with it. Yet the experience can be puzzling. AI responds quickly, but the work itself does not move much faster.&lt;/p&gt;
&lt;p&gt;The reason is simple. The company has blocked the tools and permissions needed to act, then added AI on top. Copying and pasting are restricted. External tools are unavailable. Files will not open without access. Employees cannot install a new program. Meetings and messages interrupt the work, and nothing can leave the company without approval.&lt;/p&gt;</description></item><item><title>Why Cutting Staff After Adopting AI Can Make a Company Slower</title><link>https://seunghoonchoi.com/column/ai-headcount-mistake/</link><pubDate>Sun, 21 Jun 2026 17:56:30 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-headcount-mistake/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-headcount-mistake.jpg" alt="Empty chairs lined up at desks in an office lit by morning sun"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;When a desk becomes empty, payroll costs fall, but the company may also lose the work context that person knew.&lt;/p&gt;
&lt;p&gt;When a company adopts AI tools, one sentence almost always appears: &amp;ldquo;So how many people can we cut now?&amp;rdquo; On the surface, it sounds reasonable. AI writes reports, organizes minutes, researches material, writes code, and drafts plans. Some work is genuinely faster than humans.&lt;/p&gt;</description></item><item><title>Why Heavy AI Use Can Slow a Company Down: The Hidden Costs Beyond Tokens</title><link>https://seunghoonchoi.com/column/ai-hidden-costs/</link><pubDate>Sun, 21 Jun 2026 17:56:30 +0900</pubDate><guid>https://seunghoonchoi.com/column/ai-hidden-costs/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-hidden-costs.jpg" alt="Close-up of an old analog electricity meter dial in dim light"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The cost of AI includes not only usage fees, but also the time people spend reading and correcting its output.&lt;/p&gt;
&lt;p&gt;When a company starts using AI heavily, everything seems faster at first. Reports appear sooner. Meeting notes are organized. Emails read better. Everyone says productivity has improved.&lt;/p&gt;
&lt;p&gt;Then something strange happens. The company produces more, but decisions do not come any faster. Documents multiply, while fewer people take clear ownership. A summary arrives before the meeting, but the meeting lasts just as long. More AI has not made the company faster. It has made the work more complicated.&lt;/p&gt;</description></item><item><title>A New Field Is Not Learned by Seeing More, but by Seeing Differently: 3, 7, 30, 100</title><link>https://seunghoonchoi.com/column/learning-magic-numbers/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/learning-magic-numbers/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-learning-magic-numbers.jpg" alt="A person standing at a city intersection at dusk, looking at three roads"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;A sense of a new field comes not just from the quantity of examples, but from comparing the differences between different examples.&lt;/p&gt;
&lt;p&gt;When people first learn a new field, they often ask, &amp;ldquo;How many examples do I need before I get a feel for it?&amp;rdquo; It seems like seeing more should solve it, but that is not how it works. You can look at a hundred examples of the same kind and your thoughts can still stay unorganized.&lt;/p&gt;</description></item><item><title>AI Beat Go. Why Is Welding Still Hard? Humans Stop, Machines Repeat</title><link>https://seunghoonchoi.com/column/ai-cheap-vs-expensive-world/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/ai-cheap-vs-expensive-world/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-ai-cheap-vs-expensive-world.jpg" alt="Industrial robots working around an automobile body in a factory"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Baduk AI can learn in a space with fixed rules, but in a factory, risks and responsibilities are not organized by numerical scores alone.&lt;/p&gt;
&lt;p&gt;Whenever people talk about the AI that beat the world&amp;rsquo;s best Go players, the same question comes up: &amp;ldquo;At this level, isn&amp;rsquo;t it about to take every human job?&amp;rdquo;&lt;/p&gt;
&lt;p&gt;But the picture changes when you go to the field. AI beat Go, but welding is still not easy. It can solve coding problems, but it cannot yet do sparking welds beside old pipes with the same stability as a person.&lt;/p&gt;</description></item><item><title>An AI Answer Is Only the Start: Know-How Comes from Testing It in Reality</title><link>https://seunghoonchoi.com/column/trade-secret-function/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/trade-secret-function/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-trade-secret-function.jpg" alt="A potter’s hands shaping wet clay on a wheel, where small changes in fingertip pressure affect the result"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;AI answers are easy to obtain. Know-how develops when people identify and correct the reasons those answers fail in real work.&lt;/p&gt;
&lt;p&gt;AI has made it much faster to find a method. In the past, people had to search books, talk to experts, and collect examples before they could choose a direction. Now AI can produce several plausible options in seconds. It can suggest a strategy, report structure, code, marketing copy, experimental design, or study method.&lt;/p&gt;</description></item><item><title>The Person Who Can Stand Feeling Bad at Something Eventually Gets Better</title><link>https://seunghoonchoi.com/column/pushing-through-incompetence/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/pushing-through-incompetence/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-pushing-through-incompetence.jpg" alt="A beginner’s awkward hands starting violin practice alone in a quiet room"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;If you feel like you can't do it, it means you're dealing with something you can't handle automatically yet.&lt;/p&gt;
&lt;p&gt;When you learn something new, the first feeling is not interest. It is incompetence. You think you understand it in your head, but your hands do not follow. Everyone else seems to do it easily, while you stumble. It looked easy when someone explained it, but the moment you try it yourself, everything tangles.&lt;/p&gt;</description></item><item><title>Verification Comes Before Skill: Why Trust and Reputation Decide Opportunity</title><link>https://seunghoonchoi.com/column/invisible-currencies/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/invisible-currencies/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-invisible-currencies.jpg" alt="A market vendor handing a customer a slice of cut fruit"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Without verification data, claims of ability may seem like words that increase the risks the opponent must take.&lt;/p&gt;
&lt;p&gt;Skill does not reveal itself on its own. Skill is properly recognized only when someone can confirm it. No matter how good your work is, if the other person has no way to confirm that skill, it is treated almost as if it does not exist.&lt;/p&gt;</description></item><item><title>What Matters More Than Study Smarts: Four Fundamentals That Matter More as AI Gets Better</title><link>https://seunghoonchoi.com/column/six-fundamentals/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/six-fundamentals/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-six-fundamentals.jpg" alt="A climber chalking their hands in front of a difficult rock wall"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;The faster AI gets, the more basic skills matter. A person still has to check whether the AI answer is right, what it missed, and whether it can be used as-is.&lt;/p&gt;
&lt;p&gt;AI now summarizes, translates, drafts reports, and even writes code. Does that make study smarts less important? I think the opposite is true.&lt;/p&gt;
&lt;p&gt;Memorization and repetitive calculation may matter less. But the ability to read sentences, understand how work moves, organize multiple pieces of information into structure, and handle invisible concepts matters more. The faster AI produces outputs, the more important the human fundamentals become for judging whether those outputs are right or wrong.&lt;/p&gt;</description></item><item><title>Why I Cannot Explain the Report AI Wrote for Me: What Missing Background Knowledge Is</title><link>https://seunghoonchoi.com/column/context-debt/</link><pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/context-debt/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-context-debt.jpg" alt="An old city map spread on a table, full of alleys but with no main roads marked"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;If the author cannot explain a report created by AI, the author will only act as a messenger and not the person in charge of the document.&lt;/p&gt;
&lt;p&gt;You feed material to AI and ask it to write a report. A few seconds later a plausible document appears. It has a title, background, key points, and a conclusion. Sentence by sentence it looks quite decent. Then you walk into a meeting and someone asks:&lt;/p&gt;</description></item><item><title>How to Use AI to Understand a New Work Meeting from the Transcript</title><link>https://seunghoonchoi.com/column/observing-others-meetings/</link><pubDate>Sat, 20 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/observing-others-meetings/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-meetings.jpg" alt="An empty meeting table turned into a place for observing someone else’s meeting"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;By observing other people's meetings, you can learn what criteria an organization uses for making decisions.&lt;/p&gt;
&lt;p&gt;When you walk into a work meeting for the first time, most of it is hard to follow. You know only a few words, and everyone else talks as if they already share the context. The meeting keeps moving forward while you feel like someone dropped into the middle.&lt;/p&gt;</description></item><item><title>You Can Prepare for a US PhD on Your Own</title><link>https://seunghoonchoi.com/books/us-phd-guide/</link><pubDate>Sat, 20 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/books/us-phd-guide/</guid><description>&lt;div style="display:flex;gap:30px;flex-wrap:wrap;align-items:flex-start;margin:.4em 0 1.4em"&gt;
 &lt;img src="https://seunghoonchoi.com/images/book-us-phd-guide.jpg" alt="Cover of 《You Can Prepare for a US PhD on Your Own》 (미국 박사 유학, 혼자 준비해도 됩니다)" style="width:230px;max-width:44%;border-radius:10px;box-shadow:0 16px 40px rgba(13,27,76,.24)"&gt;
 &lt;div style="flex:1;min-width:240px"&gt;
 &lt;p style="margin:.25em 0;color:#44423E"&gt;By Seunghoon Choi · Published by Bellunix · June 2026 · Written in Korean&lt;/p&gt;
 &lt;p style="margin:.25em 0;color:#44423E"&gt;eBook (ePUB) · KRW 16,000&lt;/p&gt;
 &lt;p style="margin:1.1em 0 0"&gt;
 &lt;a href="https://ebook-product.kyobobook.co.kr/dig/epd/ebook/E000013171412" target="_blank" rel="noopener" style="display:inline-block;background:#0D1B4C;color:#fff;font-weight:700;padding:13px 24px;border-radius:999px"&gt;View on Kyobo Book Centre ↗&lt;/a&gt;
 &lt;a href="https://smartstore.naver.com/bellunix/products/13640954951" target="_blank" rel="noopener" style="display:inline-block;background:#B87333;color:#fff;font-weight:700;padding:13px 24px;border-radius:999px;margin:6px 0 0 8px"&gt;Naver Smartstore ↗&lt;/a&gt;
 &lt;/p&gt;
 &lt;/div&gt;
&lt;/div&gt;
&lt;p&gt;Other than knowing I needed a TOEFL score, I knew almost nothing. This book collects every experience and mistake from the 18 months it took me to receive an offer letter from scratch. At the end of that process, I was admitted to the materials engineering PhD program at the University of Florida (UF) with full funding, and I received my F-1 visa.&lt;/p&gt;</description></item><item><title>From Acceptance to Departure: Reading Your Offer Letter to Sending Money Abroad</title><link>https://seunghoonchoi.com/career/uf-guide-1-before-departure/</link><pubDate>Wed, 17 Jun 2026 17:00:00 +0900</pubDate><guid>https://seunghoonchoi.com/career/uf-guide-1-before-departure/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/uf1-departure.jpg" alt="Airplane at the departure gate"&gt;&lt;/p&gt;
&lt;p&gt;Admission to a U.S. STEM graduate program is not the finish line. It is a new starting line. The months between getting your acceptance and leaving the country hold more tasks, tighter deadlines, and higher stakes than most people expect, and a single missed step can derail enrollment, registration, or your visa. This post is Part 1 of a guide to admission and settling in at the University of Florida (UF), walking you through what to handle, in order, from acceptance through departure. It covers how to read your funding (offer letter) accurately, the key date calendar, the placement exam and course registration process, pre-arrival research preparation, and finally wrapping up your Korean paperwork and sending money abroad.&lt;/p&gt;</description></item><item><title>F-1 Visa and U.S. Entry: From Status Requirements to Airport Arrival</title><link>https://seunghoonchoi.com/career/uf-guide-2-f1-visa-entry/</link><pubDate>Wed, 17 Jun 2026 16:00:00 +0900</pubDate><guid>https://seunghoonchoi.com/career/uf-guide-2-f1-visa-entry/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/uf2-visa.jpg" alt="Passport and travel documents"&gt;&lt;/p&gt;
&lt;p&gt;Receiving an admission offer or an F-1 visa does not complete the process. After entering the United States, students must continue to meet the conditions of F-1 status. Before departure, they also need to carry the right documents and understand the school&amp;rsquo;s check-in deadline. This post covers F-1 status requirements, inspection at the Port of Entry (POE), ISSS online check-in after arrival, government fees, and travel from the airport to Gainesville.&lt;/p&gt;</description></item><item><title>ReadFast: English and Chinese Meaning-Unit Reading Practice</title><link>https://seunghoonchoi.com/apps/reading-trainer/</link><pubDate>Tue, 16 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/apps/reading-trainer/</guid><description>&lt;div class="appcard"&gt;&lt;img class="appcard__icon" src="https://seunghoonchoi.com/images/reading-trainer-card.svg" alt="ReadFast reading practice app icon"&gt;&lt;div class="appcard__body"&gt;&lt;span class="appcard__free"&gt;Installable web app (PWA)&lt;/span&gt;&lt;h3&gt;ReadFast: English and Chinese meaning-unit reading&lt;/h3&gt;&lt;p&gt;Register a text, train with it, and view the original with its Korean translation in Texts.&lt;/p&gt;&lt;a class="cta" href="https://seunghoonchoi.com/reading-trainer/" target="_blank" rel="noopener"&gt;Open the app →&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;
&lt;p&gt;First, register the English or Chinese text you want to read. ReadFast uses that text for a training session and adds the same text to the Texts tab with its original and a natural Korean translation. The training screen presents the text in short connected units, so you can keep linking meaning as you read.&lt;/p&gt;</description></item><item><title>The Trap in Reviewing AI Outputs: Do Not Lower the Ceiling While Catching Errors</title><link>https://seunghoonchoi.com/column/dont-lobotomize-the-model/</link><pubDate>Tue, 16 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/column/dont-lobotomize-the-model/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/col-qa.jpg" alt="A magnifying glass beside a laptop"&gt;&lt;/p&gt;
&lt;p class="inline-image-caption"&gt;Inspection is not a task of reducing the output, but a task of finding and correcting actual errors.&lt;/p&gt;
&lt;p&gt;The text was sticking out past the slide. I noticed it only right before sending.&lt;/p&gt;
&lt;p&gt;In Excel, a &lt;code&gt;#REF!&lt;/code&gt; error was still sitting there, and table borders appeared in some cells but not in others. In a Word document, markdown symbols that should have been removed were still visible. These are not matters of taste. The output is simply broken.&lt;/p&gt;</description></item><item><title>What Remains</title><link>https://seunghoonchoi.com/literature/what-remains/</link><pubDate>Mon, 15 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/literature/what-remains/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/lit-remains.jpg" alt="Raindrops on a window"&gt;&lt;/p&gt;
&lt;div class="poem"&gt;
&lt;p&gt;Perhaps what matters most&lt;br&gt;
is that tonight, too, a star, on the wind,&lt;br&gt;
against us,&lt;br&gt;
grazes past.&lt;/p&gt;
&lt;p&gt;Was that really&lt;br&gt;
more important than a heart that sings of the stars?&lt;/p&gt;
&lt;p&gt;We are, after all,&lt;br&gt;
just a pale blue dot.&lt;/p&gt;
&lt;p&gt;There is no helping it:&lt;br&gt;
nothing left but to love.&lt;/p&gt;
&lt;/div&gt;</description></item><item><title>Korean Citizenship Comprehensive Evaluation Practice - Written, Essay, and Oral Mock Training</title><link>https://seunghoonchoi.com/apps/gwiwha-app/</link><pubDate>Sun, 14 Jun 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/apps/gwiwha-app/</guid><description>&lt;div class="appcard"&gt;
 &lt;img class="appcard__icon" src="https://seunghoonchoi.com/images/gwiwha-card.svg" alt="Korean citizenship practice app icon"&gt;
 &lt;div class="appcard__body"&gt;
 &lt;span class="appcard__free"&gt;Installable web app (PWA)&lt;/span&gt;
 &lt;h3&gt;Korean Citizenship Comprehensive Evaluation Practice&lt;/h3&gt;
 &lt;p&gt;Practice the written, essay, and oral sections repeatedly in one place.&lt;/p&gt;
 &lt;a class="cta" href="https://seunghoonchoi-phd.github.io/gwiwha/" target="_blank" rel="noopener"&gt;Open the app →&lt;/a&gt;
 &lt;/div&gt;
&lt;/div&gt;
&lt;p&gt;I started building this while my wife was preparing for Korean citizenship. Written questions, essay topics, and oral practice materials were scattered everywhere, so I built one place where she could practice them repeatedly. No payment, no sign-up.&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;Written practice - repeat multiple-choice questions until they feel familiar&lt;/li&gt;
&lt;li&gt;Essay and oral practice - look at expected topics and practice writing or speaking your own answers&lt;/li&gt;
&lt;li&gt;Korean-Chinese side-by-side text - easier for users who are still building Korean fluency&lt;/li&gt;
&lt;li&gt;Installable web app (PWA, Progressive Web App) - add it to your phone&amp;rsquo;s home screen and use it like an app, even offline&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;&lt;a href="https://seunghoonchoi-phd.github.io/gwiwha/"&gt;Open the app →&lt;/a&gt;&lt;/p&gt;</description></item><item><title>Bio-Inspired Hydrogel Electrode for Plant-Leaf Signals: IJASC</title><link>https://seunghoonchoi.com/research/plant-hydrogel-electrode/</link><pubDate>Sat, 28 Feb 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/research/plant-hydrogel-electrode/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/plant-hydrogel-electrode.jpg" alt="Hydrogel octopus-inspired adhesive patch recording electrical signals from plant leaves"&gt;&lt;/p&gt;
&lt;p&gt;A co-first-authored paper introduces a hydrogel-integrated, octopus-inspired adhesive (H-OIA) patch with an engineered silicone interface that conforms to the rough, variably-hydrated surface of plant leaves. On &lt;em&gt;Coffea arabica&lt;/em&gt;, it enables stable long-term recording of stimulus-responsive plant bioelectrical signals where conventional electrodes detach.&lt;/p&gt;
&lt;p&gt;Jeon, S.; Choi, S.; Kim, D.W. &amp;ldquo;Bio-Inspired Hydrogel Adhesive Electrode Enabling Stimulus-Responsive Electrical Signal Recording in Plant Leaves.&amp;rdquo; &lt;em&gt;Int. J. Adv. Smart Converg. (IJASC)&lt;/em&gt; 2026, 15(1), 241-253. &lt;a href="https://doi.org/10.7236/IJASC.2026.15.1.241"&gt;DOI: 10.7236/IJASC.2026.15.1.241 ↗&lt;/a&gt;&lt;/p&gt;</description></item><item><title>Octopus-Inspired Adhesive Electrode for EMG: IJIBC</title><link>https://seunghoonchoi.com/research/octopus-emg-electrode/</link><pubDate>Sat, 28 Feb 2026 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/research/octopus-emg-electrode/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/octopus-emg-electrode.jpg" alt="Octopus-inspired adhesive electrode for EMG-based robotic control under dry and wet conditions"&gt;&lt;/p&gt;
&lt;p&gt;A co-first-authored paper presents an octopus-inspired adhesive electrode whose microstructured suction architecture keeps conformal skin contact and stable EMG signals under both dry and wet conditions. It addresses the detachment and signal degradation that limit EMG-based control of prosthetics and wearable robots.&lt;/p&gt;
&lt;p&gt;Jeon, S.; Choi, S.; Jang, J.; Shim, J.Y.; Kim, D.W. &amp;ldquo;Octopus-Inspired Adhesive Electrode for Robust EMG-Based Robotic Control under Dry and Wet Conditions.&amp;rdquo; &lt;em&gt;Int. J. Internet Broadcast. Commun. (IJIBC)&lt;/em&gt; 2026, 18(1), 274-284. &lt;a href="https://doi.org/10.7236/IJIBC.2026.18.1.274"&gt;DOI: 10.7236/IJIBC.2026.18.1.274 ↗&lt;/a&gt;&lt;/p&gt;</description></item><item><title>Octopus-Mimicking Soft Grippers: Review in Biomimetics</title><link>https://seunghoonchoi.com/research/octopus-grippers-review/</link><pubDate>Thu, 04 Dec 2025 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/research/octopus-grippers-review/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/octopus-grippers-review.jpg" alt="Octopus-inspired soft grippers: suction cups, arms, hybrid designs, and sensing (Biomimetics 2025)"&gt;&lt;/p&gt;
&lt;p&gt;A co-first-authored review in Biomimetics surveys octopus-inspired soft grippers across three functional dimensions: structural and sensing devices, control strategies, and AI-driven applications. It maps suction-cup geometries, tentacle-like actuators, and hybrid designs toward more capable, intelligent soft robots.&lt;/p&gt;
&lt;p&gt;Choi, S.; Jang, J.; Lee, J.; Kim, D.W. &amp;ldquo;Design and Sensing Frameworks of Soft Octopus-Inspired Grippers Toward Artificial Intelligence.&amp;rdquo; &lt;em&gt;Biomimetics&lt;/em&gt; 2025, 10(12), 813. &lt;a href="https://doi.org/10.3390/biomimetics10120813"&gt;DOI: 10.3390/biomimetics10120813 ↗&lt;/a&gt;&lt;/p&gt;</description></item><item><title>Research Featured in Chinese on MaterialsViewsChina</title><link>https://seunghoonchoi.com/research/materialsviews-china/</link><pubDate>Sun, 15 Dec 2019 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/research/materialsviews-china/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/materialsviews.png" alt="MaterialsViewsChina feature"&gt;&lt;/p&gt;
&lt;p&gt;My &lt;em&gt;Advanced Functional Materials&lt;/em&gt; work was introduced in Chinese on MaterialsViews China, expanding its reach to a broader Chinese-speaking scientific audience. The work is a conductive hierarchical hairy-fiber sensor with high sensitivity, stretchability, and wet resistance, enabling multimodal detection of human motion and gesture.&lt;/p&gt;
&lt;p&gt;&lt;a href="https://mp.weixin.qq.com/s/r0QAnvMYlBQvqXosUhBiEA"&gt;Read the feature ↗&lt;/a&gt;&lt;/p&gt;</description></item><item><title>Selected as a Back Cover in Advanced Functional Materials</title><link>https://seunghoonchoi.com/research/conductive-hairy-fibers/</link><pubDate>Tue, 10 Dec 2019 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/research/conductive-hairy-fibers/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/afm-backcover.jpeg" alt="Advanced Functional Materials Back Cover"&gt;&lt;/p&gt;
&lt;p&gt;Our work on conductive hierarchical hairy fibers, a highly sensitive, stretchable, water-resistant, multimodal gesture-distinguishable sensor, was selected for the back cover of &lt;em&gt;Advanced Functional Materials&lt;/em&gt; (issue 50/2019), highlighting applications in wearable electronics and VR interfaces.&lt;/p&gt;
&lt;p&gt;&lt;a href="https://onlinelibrary.wiley.com/doi/10.1002/adfm.201905808"&gt;View the article ↗&lt;/a&gt;&lt;/p&gt;</description></item><item><title>Best Poster Award: 2018 MRS Fall Meeting</title><link>https://seunghoonchoi.com/research/mrs-best-poster/</link><pubDate>Sun, 25 Nov 2018 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/research/mrs-best-poster/</guid><description>&lt;p&gt;&lt;img src="https://seunghoonchoi.com/images/mrs-event-opt.jpg" alt="Best Poster Award: 2018 MRS Fall Meeting &amp;amp; Exhibition"&gt;&lt;/p&gt;
&lt;p&gt;Received the Best Poster Award for an octopus-inspired adhesive and conductive patch sensor for biosignal monitoring, presented at the Poster Session of the 2018 MRS Fall Meeting &amp;amp; Exhibition in Boston (November 2018).&lt;/p&gt;</description></item><item><title>About</title><link>https://seunghoonchoi.com/about/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://seunghoonchoi.com/about/</guid><description>&lt;figure class="about-portrait"&gt;
 &lt;img src="https://seunghoonchoi.com/images/about-portrait.jpg" alt="Seunghoon Choi" width="975" height="1254" loading="eager"&gt;
&lt;/figure&gt;
&lt;p&gt;I am an engineer and a creator. After solving a complex or tedious problem myself, I document the process and turn useful parts into tools that others can use. In fall 2026, I plan to begin a Ph.D. in Computational Materials Science at the University of Florida (UF).&lt;/p&gt;
&lt;p&gt;Degrees and research matter, but so does keeping an honest record of how I got here. I want to solve problems I have faced firsthand and share what I learn.&lt;/p&gt;</description></item></channel></rss>