YT YouTube2Text

YouTube2Text → Transcripts → 👓매일 정보를 인사이트로 만드는 미키김의 4단계 루틴 (feat. 투자 맥락 읽기)

👓매일 정보를 인사이트로 만드는 미키김의 4단계 루틴 (feat. 투자 맥락 읽기) — Transcript

by 미키피디아 · 3,047 words · 476 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00We've been taking Q&A questions lately,
0:02and there are so many comments asking
0:04where we get our insights and how we
0:06find them.
0:07Yes. First off, thanks for all the
0:11questions; we live in an era of
0:12information overload, but sometimes
0:15that just makes it harder to see where
0:17the world is heading. That's why I
0:20think it's even more important to
0:22filter good information, grasp the
0:24context, and develop your own insights.
0:28Stock prices are ultimately a result of
0:29how the world is changing, so those who
0:31can find insights from information are
0:33better at investing, too. I believe you
0:37need a framework, and I actually have a
0:39four-step process for turning
0:41information into insights. Today, I'm
0:44going to walk you through these four
0:46steps in detail. And since we’re
0:48talking about how to view the world, we
0:50have a really cool eye-related comment
0:51event, so make sure to watch until the
0:53end. The first step is, of course,
0:55gathering information. What is
0:57happening right now? Consuming a lot of
0:59information doesn't necessarily mean
1:00you understand the world well. We’re
1:03drowning in information these days, but
1:05it often feels repetitive, which I
1:07think is why we feel the need for more.
1:11YouTube is a prime example of a space
1:12filled with that kind of repetitive
1:14information. So, quality is more
1:17important than quantity, and since you
1:20can't know every subject, you need to
1:24limit the areas you want to
1:26consistently observe. This is where
1:29choice and focus become essential. For
1:31instance, I’m extremely interested in
1:33the stock market, but I don't follow
1:35political news very closely. I love the
1:38tech industry, but I'm not that
1:40interested in energy; I watch movies,
1:42but I don't really follow celebrity
1:43news. While you should keep an eye on
1:47general world news, I think it’s
1:49important to clearly distinguish
1:51between fields you follow deeply and
1:53those where you only consume big
1:55headlines. Doing that helps you control
1:59your own fingers in an age where
2:01everything on your phone is designed to
2:03make you click. You need to actively
2:06gather info in the areas you follow,
2:09and there are three main sources for
2:12that. The first is primary sources. For
2:14example, if there's an earnings
2:16announcement, the official report is a
2:17primary source, not a news article
2:19about it. Things like original policy
2:21texts or official reports—what we
2:22often call the "source." The surest way
2:25to grasp the facts, and the area I
2:27really want to dig into, is to get as
2:29close to the source as possible rather
2:31than relying on news articles. Of
2:35course, primary sources aren't neutral
2:37because they reflect the interests of
2:39the organizations that created them.
2:42The second is reports or
2:43interpretations based on these primary
2:44sources. It’s someone else’s
2:46summary and commentary on a piece of
2:48news. In the past, the media held a
2:51monopoly on this, but in the era of
2:53online and social media, individuals
2:55are actively playing this role, and
2:56even on Wikipedia, tech or
2:58stock-related news often serves this
3:00purpose. The shift of control over
3:03reporting and interpretation to
3:05individuals is a key reason for the
3:07weakened power of the press, and it’s
3:09also why we are flooded with 50,000
3:11opinions and unverified news. I get
3:13most of these news summaries and
3:15interpretations from YouTube or
3:17podcasts. I personally don't like
3:20portal news sites like Naver because
3:22there’s so much clickbait that once I
3:24go in, I encounter news in such a
3:26scattered way that I forget what I came
3:28to look for in the first place. So, as
3:31I mentioned before, it ruins my
3:33selective focus. On the other hand,
3:36YouTube’s algorithm limits exposure
3:38based on my interests—though I have
3:40to be careful as that can foster
3:42repetitive information and confirmation
3:45bias—but it’s good that it provides
3:47some limits, and podcasts keep me
3:49within the areas I follow since they
3:52only show what I subscribe to. And for
3:54news commentary, I try to watch
3:56channels from both Korea and the U.S.
3:58to maintain a balance. For example, for
4:01stock market news and commentary, I
4:02watch Korean channels like 3Pro TV or
4:04Hankyung Global Market, but I also
4:06enjoy U.S. channels like CNBC, the Wall
4:08Street Journal, and More Fool, where
4:10we’ve even interviewed the founders.
4:12The third is data. Data is information
4:15that provides objective signals, such
4:17as sales figures, stock prices,
4:19searches, ratings, and approval ratings
4:21. While primary sources and news often
4:24organize this data well, it’s more
4:26valuable to look at the data myself if
4:28I can. Instead of just reading an
4:31article saying a company's stock price
4:33went up a lot, I look up how that stock
4:35moved over a day, a month, or a year
4:37myself. An article might say it went up
4:39a lot yesterday, but when I check the
4:40data, it’s actually halved over the
4:41past year. This provides a completely
4:43different insight. And if you aren't
4:46sure what data exists, you can ask an
4:48AI to find it, and it will search and
4:50organize it for you. Acquiring
4:53information like this is step one, and
4:55if you don't want to dig any deeper,
4:57you can just stop here. But that won't
5:00lead to insights, so you have to move
5:02on to step two. Step two is the stage
5:04of grasping the context. If acquiring
5:07information in step one is a single dot
5:09, then connecting those dots creates a
5:11flow. In other words, you're looking at
5:13change, not just events. This is
5:14ultimately what grasping the context is
5:16, and that is step two. There are
5:18criteria you can apply to grasp the
5:20context more concretely, and the method
5:22I use is called "Won-Ban-Hwak-Ji-Gwan."
5:24I made that term up myself.
5:25Won-Ban-Hwak-Ji-Gwan.
5:27Yes. Let's look at each part of
5:28Won-Ban-Hwak-Ji-Gwan one by one. Won (
5:30Origin). Cause and Intent. You look at
5:33recurring patterns, like why this is
5:34happening and why such decisions were
5:36made. Check if similar phenomena are
5:39recurring and if they are happening
5:41elsewhere; 'Hwak'(Diffusion: is it
5:43spreading to other industries or
5:46regions?); 'Ji'(Persistence: is it a
5:48temporary phenomenon or a structural,
5:50lasting change?); 'Gwan'(Perspective:
5:53whose opportunity is this change, and
5:55whose crisis is it?). When you feel
5:58like you need a framework to grasp the
6:00context, it's good to look at
6:01Won-Ban-Hwak-Ji-Gwan. Analyzing news,
6:04in the end, is also a process of
6:06grasping this context. So,
6:08Won-Ban-Hwak-Ji-Gwan is a process that
6:10focuses on the past and present,
6:12without thinking about the future for
6:14now. For example, let's say Apple
6:16announces a new product. Then, in step
6:18one, you go close to the source to
6:20grasp what the facts are and what
6:22product was announced. Then, let's
6:24apply Won-Ban-Hwak-Ji-Gwan. Won. Why
6:26did they release this product now? Ban.
6:28Were there similar cases in the past?
6:30Hwak. Are competitors also releasing
6:32similar products? Ji. Is it part of a
6:35larger product strategy? Gwan. Which
6:38users will like this? It helps to grasp
6:41the context using these few criteria,
6:43and you can only do this properly when
6:45the areas you follow are limited. Just
6:48as mentioned earlier. That's why "you
6:50see only as much as you know" applies
6:52to grasping context, and it's also why
6:54pretending to know about a field you
6:55don't really understand makes it even
6:57more obvious that you're clueless. And
7:00if you can just ask the right questions
7:02, AI is really good at organizing this
7:04context for you. Grasping context is
7:07the task of connecting countless pieces
7:09of information, and connecting
7:10information is exactly what AI does
7:12best. So, don't just ask AI to find
7:15information; if you also pose questions
7:16that allow it to grasp the context, you
7:18will become that much smarter. I tell
7:21ChatGPT what a "disciple of the disc"
7:23is and have it analyze things that way;
7:25it provides five points of analysis and
7:27does a really great job.
7:28It's a fantastic tip to plant the
7:30concepts you hold into an AI beforehand
7:32and have it act accordingly.
7:34Yes. Also, what’s interesting is that
7:36the more you use AI, the better it
7:37knows you; if I ask about a certain
7:39piece of news, it analyzes it using the
7:41"disciple of the disc" method even
7:42without me asking. Hmm. This.
7:45If the first two stages focus on
7:47grasping the context of the past and
7:48present, the third stage is about
7:50thinking ahead to what changes might
7:52come in the future. Stage 3 can be
7:54called a "future hypothesis." Because
7:57just connecting information well
7:59doesn't mean you can accurately predict
8:00the future—just like we can't predict
8:02future stock prices—it’s more
8:03accurate to call this stage a "future
8:05hypothesis" rather than a prediction.
8:07Taking the Apple Vision Pro example
8:09again, how many units will sell in the
8:11future? What kind of changes will this
8:12product bring to people's behavior?
8:14What kind of ecosystem will be built
8:16around this product, who will make
8:17money from it, and how will it affect
8:19stock prices, and so on. Since it's a
8:23hypothesis rather than a prediction, if
8:25you want to dig deeper, it’s good to
8:28think about an upside scenario if
8:30things go better than expected, and
8:32conversely, a downside scenario if they
8:35fall short of the base scenario. It’s
8:38a method often used when reporting
8:40business plans at work, isn't it?
8:42Presenting a best-case and a worst-case
8:44scenario. If you want to go even deeper
8:47here, it would be best to add
8:48probabilities to the base, upside, and
8:50downside scenarios. Though it’s
8:52easier said than done. And to verify if
8:55the hypothesis you've set is correct,
8:57it's a good idea to think about what
8:59kind of data or changes you should look
9:01out for moving forward. Ultimately, an
9:03insight is something completed by
9:05combining stages 1, 2, and 3. What
9:07happened, why it happened, what kind of
9:09phenomenon this is in the big picture,
9:11and what changes are ahead—that, in
9:12the end, is an insight. The final stage
9:14, stage 4, is to organize and record
9:16these first three stages. When you
9:19record them, they are much better
9:20organized than if you were just
9:22thinking about them. I believe there
9:24are two main types of recording. One is
9:27to sit down at your computer and
9:29organize the aforementioned stages 1, 2
9:31, and 3 whenever you encounter news
9:32you're interested in or want to dig
9:34into further. And it’s convenient to
9:37have an AI by your side to help you
9:38with that process. The big and small
9:41news I often post on my Instagram
9:44stories is a summary of these organized
9:46notes, and we also compile these
9:49records to create videos for Mickey
9:51Pedia. The video where we collect and
9:54summarize insights from recent
9:55quarterly earnings reports is exactly
9:57made that way. But it’s not always
10:00done so formally; sometimes insights or
10:02thoughts just pop into my head, or I
10:04jot down new information or insights I
10:06pick up while listening to podcasts or
10:08audiobooks in my notepad—that’s
10:10record-keeping too. You have to write
10:13these things down immediately,
10:14otherwise, they get forgotten later.
10:16You know how you go, "Oh, what was that
10:17thing?" and it just doesn't come to
10:19mind. That’s why I have a separate
10:20note in the Apple Notes app
10:21specifically for "Thoughts and
10:23Information." So, whenever something
10:25crosses my mind, or I hear or read
10:27something and think, "Hey, this is good
10:29," I jot it down right then and there,
10:31and it becomes a great source for
10:32creating YouTube content.
10:34Could you give me one recent example of
10:36how you organized something like that?
10:38Oh, yes. Recently, some top developers
10:42left Google. It was all over the news.
10:45In particular, a legendary genius
10:47developer named Jay, who had been at
10:50Google for nearly 30 years and
10:51basically coded everything from Google
10:54Search to other major services, has
10:56left the company. I met him a few times
11:00when I worked at Google, but anyway,
11:01the news that Jay was leaving Google
11:03was the first step. I also verified it
11:06through primary sources, such as
11:08Jay’s post on X and the official
11:10Google blog. It wasn't solely because
11:12of this, but Google’s stock price
11:14fell by 4%following this news. This is
11:16the data material I mentioned earlier.
11:18Now for step two, understanding the
11:20context—if we look at the root cause,
11:22it’s likely that no matter how
11:24successful he was, he felt frustrated
11:26by the pace and internal politics of
11:28Google, which has become a massive
11:29corporation, and naturally wanted to
11:31start his own thing. This is just the
11:34cause as I presume it to be. Of course,
11:36this is not a fact. Second, recurrence.
11:39Developers leaving to start their own
11:41companies is a recurring event that
11:42happens not just at Google, but across
11:43all big tech firms. This is a fact. As
11:47for step three, what impact this will
11:49have on Google, losing such super
11:51developers will surely affect
11:52development schedules in the short term
11:54, and as his new startup grows, other
11:55Google developers may move there, which
11:57will continue to affect the internal
11:59atmosphere. However, since a large
12:03company like Google basically runs on
12:04systems, I don’t believe there is a
12:06possibility that their technical
12:07capabilities will collapse just because
12:09a few developers have left. The impact
12:12on earnings will be limited; while the
12:14stock price reacted to the news that
12:16day, it’s unlikely to have a lasting
12:18effect. However, if this talent exodus
12:23lowers Google’s long-term AI
12:25competitiveness, that would be negative
12:27, so we’ll need to keep a close eye
12:29on the new versions of Gemini and
12:31future Google AI products. That’s a
12:35good way to summarize it. I believe
12:37breaking down information into facts,
12:39context, and predictions is a great way
12:42to build your own insights.
12:44I think this is something you can
12:45practice, so it’s important to give
12:46it a try often.
12:47Of course. You can practice this every
12:49day. When you look at the news you
12:51encounter daily, pick the ones you want
12:52to delve into deeply, as I mentioned
12:54earlier. If you apply these four steps
12:57to at least one news story a day, you
12:59will really feel your insights
13:01skyrocketing. And when this topic comes
13:05up elsewhere, there’s a real
13:06difference between someone who just
13:08repeats the article they read and
13:10someone who shares actual insights. I
13:12believe this is entirely possible
13:14through training. Ultimately, it’s
13:17about how you tailor information to
13:19create your own insights, and in a way,
13:21I think it’s similar to building your
13:23own personal standards. It’s not just
13:26about information; when we choose
13:28products, I believe there is no better
13:30product than one that is perfect for
13:32you. Like this custom-made clothing.
13:34But in an age where technology has
13:36advanced so much, I wondered why we
13:38still force our faces and eyes to fit
13:40mass-produced glasses. Anyone who wears
13:42glasses knows that discomfort of having
13:45frames that hurt your nose and ears. So
13:47, I wanted to get glasses that fit my
13:49face and eyes perfectly, which is why
13:51I’m here at the Seongsu store of
13:52Brizm, a brand that makes custom
13:54eyewear. The name of this place is
13:56Brizm Foundry Seongsu. You know how we
13:58like the "foundries" that manufacture
14:00semiconductors for us. I had my own
14:01custom glasses made here as well. What
14:04was fascinating about the process was
14:06that they do a 3D scan of my face. They
14:08make the glasses by measuring not just
14:10the height of my ears, but even the
14:11distance from my eyes to my ears. Not
14:13only that, but everyone has a different
14:15distance between their pupils. The
14:18scanner adjusts for that distance
14:20perfectly, which was great, and there
14:23are so many different frame designs too
14:25. Once they scan your face in 3D, the
14:28AI recommends the designs that would
14:30suit you best. It even shows a
14:32percentage next to them. "Something
14:34like this would look good on you." But
14:35the design the AI recommended as the
14:37best fit for my face was also the one I
14:39found the most attractive. So, I ended
14:42up ordering that design.
14:44But I've never really felt much
14:45discomfort wearing off-the-shelf
14:46glasses, so what's so different about
14:48custom ones? Ah, this is my first time
14:50getting them custom-made. And wow,
14:52it’s truly different. First off, my
14:53ears are at different heights. So when
14:56I buy regular glasses, I always have to
14:57adjust them to match my ear heights.
14:58But this time, they used 3D printing to
15:00match not just my exact ear height, but
15:02the distance from my eyes to my ears.
15:04So the comfort behind the ears is
15:06honestly on another level. And the key
15:08to comfortable glasses is the nose
15:10bridge. Since everyone's nose is shaped
15:12differently, some glasses are annoying
15:14because they press down too hard, while
15:16others keep sliding off. But since
15:19these were made for me, the nose part
15:21feels incredibly comfortable. Also, one
15:23thing I found interesting while getting
15:25these made is that even if you like a
15:27design, there’s an "R size" that fits
15:29each person. With Breezm, there are
15:32seven R sizes, and I think being able
15:34to choose the one that fits me best is
15:36another reason why these glasses feel
15:39so comfortable. Usually, custom
15:41products are bound to be expensive.
15:45Since they’re made to measure, higher
15:47prices are expected, but looking at the
15:48frame prices, they range from 140,000
15:50to 250,000 won, which is very
15:51reasonable. Of course, the final price
15:55varies depending on the lenses, but
15:57even with 100%domestic production and
15:59an on-demand custom manufacturing
16:01process that you'd think would drive
16:03costs up, they maintain these
16:05reasonable prices by streamlining
16:07production, distribution, and sales.
16:10Breezm glasses are being made right now
16:13through various processes, including
16:15this foundry 3D printing. I’ve been
16:18raving about these glasses, and I felt
16:20I couldn't move on without a subscriber
16:22event, so they’re giving a 300,000
16:25won Breezm gift card to three people
16:27who leave nice comments on today's
16:29video. Also, you need to book an
16:32appointment to get fitted for Breezm
16:34glasses, and if you put "Micky" in the
16:36requests section, you'll get a 10%
16:37discount, so keep that in mind if
16:39you're interested in custom glasses.
16:42I’ll be back next time with more fun
16:45and informative topics. Thank you.

About this transcript

This page contains the full transcript of 👓매일 정보를 인사이트로 만드는 미키김의 4단계 루틴 (feat. 투자 맥락 읽기) by 미키피디아, generated from the public captions YouTube serves with the video. The transcript has 3,047 words across 476 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.