AI가 다하는데 왜 컴퓨터공학을 알아야 할까? (밑바닥 시리즈 OT) — Transcript
Full transcript
- 0:00I am the developer you know from Vibe
- 0:02Coding PT. These days, people doing
- 0:05Vibe Coding are very sensitive to AI
- 0:07trends, but they overlook the
- 0:08importance of computer science. What do
- 0:12you think is the foundation, the very
- 0:14bottom of Vibe Coding basics? It is
- 0:17computer science. Nowadays, LLMs or
- 0:20coding agents like Claude Code all
- 0:22operate on computers. And you know the
- 0:26results you create through Vibe Coding?
- 0:29Web pages, mobile apps, and automation
- 0:32tools—they all run on computers.
- 0:35However, you don't really know much
- 0:37about computers. I think this is a very
- 0:40big problem. So in this video, I will
- 0:43summarize exactly three reasons why
- 0:45those doing Vibe Coding need to know
- 0:48computer science and what the benefits
- 0:50are. This is also an orientation video
- 0:54that distills the essential computer
- 0:56science knowledge that non-developers
- 0:59must know. So, if you watch until the
- 1:02end, you will accurately understand why
- 1:04you need to know computer science. And
- 1:07I will summarize the seven topics we
- 1:09will cover in this "Bottom Basics"
- 1:10series, so watching until the end will
- 1:12be helpful. Now, at this point, you
- 1:15might be thinking like this. "I’m
- 1:18just doing Vibe Coding, do I really
- 1:20need to know computer science to this
- 1:22extent?" I would like to ask you this
- 1:25question. You are all studying AI very
- 1:27hard right now, aren't you? Right? But
- 1:31among these AI-related technologies,
- 1:32there isn't a single one that isn't
- 1:34connected to computers. So, to
- 1:37understand AI well, we need to
- 1:39understand the fundamentals of
- 1:40computers. You absolutely do not need
- 1:43to know as much as a major student. In
- 1:45this "Bottom Basics" series, I will
- 1:47teach you only the computer
- 1:48fundamentals that non-developers
- 1:49strictly need to know. Now, I will
- 1:52begin by explaining the three reasons
- 1:55why we need to know computer science.
- 1:58First, while doing Vibe Coding, you
- 2:01must have had this experience. That
- 2:05something feels very uneasy. This
- 2:07uneasiness comes from what we don't
- 2:09know well. In difficult terms, we call
- 2:11this cognitive debt. It looks like it's
- 2:14working well on the surface, but you
- 2:15don't know how it works internally, and
- 2:17you wonder, "Will it keep working if I
- 2:19add more features?" Even when deploying
- 2:21, you might wonder if there are any
- 2:23security issues. You probably don't
- 2:25feel quite confident. So, every time
- 2:27that happens, you need to resolve what
- 2:29you don't know. When AI mentions
- 2:32technical jargon and says, "Oh, I coded
- 2:34it like this," you need to dig into
- 2:36that technical term. It is important
- 2:40not to accumulate cognitive debt in
- 2:42this way. If you keep following those
- 2:45technical terms, you will eventually
- 2:48arrive at computer science knowledge.
- 2:51Therefore, the fundamental cause of the
- 2:53cognitive debt that occurs in Vibe
- 2:55Coding is simply because we don't know
- 2:58enough about computers. That is why you
- 3:01need to build a solid core in computer
- 3:03science. Now, for the second point.
- 3:06These days, Astra is out and it's
- 3:09really hot. These AI models ultimately
- 3:12run on computer systems as well. And
- 3:15there is a component that you cannot
- 3:17leave out when talking about AI. What
- 3:19is it? It is the GPU. GPU stands for
- 3:22Graphics Processing Unit, a concept
- 3:24that comes from computer architecture.
- 3:27It is a component that emerged to
- 3:29overcome the limitations of the CPU. So
- 3:32, the GPU, which plays a key role in
- 3:34running AI models, is ultimately
- 3:37related to computer architecture within
- 3:40computer science. Now, let me give you
- 3:43another example; if you use Claude Code
- 3:46, you might have heard of the concept
- 3:48of a context window. If you have a
- 3:50solid understanding of computer science
- 3:52, you can see that it is very similar
- 3:54to a computer's memory structure. It
- 3:56may seem like a new technology has
- 3:58appeared, but AI technology ultimately
- 4:00cannot escape the computer. So, while
- 4:04keeping up with trends is important, if
- 4:06you secure the fundamentals, you can
- 4:08quickly learn any new AI-related
- 4:10technology that comes out. And the
- 4:14third, most important point is that you
- 4:17gain the judgment to not be swayed by
- 4:20AI. It is precisely judgment. Let me
- 4:25explain what this means. As you know
- 4:29from using AI, it is getting more
- 4:30sophisticated. This means that it might
- 4:34say it has finished a task, but when
- 4:36you check the code yourself, it is
- 4:38often not fully complete. And there are
- 4:41also cases where the work has been done
- 4:43in a completely different direction.
- 4:45Therefore, we absolutely need the
- 4:47ability to accurately verify the work
- 4:50done by AI. The importance of this is
- 4:53growing day by day. When AI says, "I
- 4:56have done this and that," it will often
- 4:58throw around difficult technical jargon
- 5:01. At that point, we must be able to
- 5:03accurately understand that technical
- 5:05jargon. That understanding all comes
- 5:07from computer science. Now, if we don't
- 5:09know this, it will all sound like a
- 5:11foreign language to us. Cognitive debt
- 5:13regarding the work and results we are
- 5:15creating with AI will continue to
- 5:17accumulate, and we will completely lose
- 5:20control over those results to the AI.
- 5:23Conversely, if you have the power to
- 5:25accurately understand what AI is saying
- 5:28, you can play with the AI. So, I have
- 5:31explained three reasons why you need to
- 5:34know computer science. Now, finally, I
- 5:36will explain how the "From the Bottom
- 5:38Up" series will proceed. First, I will
- 5:42explain the history of the computer.
- 5:46The history of the computer. And second
- 5:50, I will explain the history of AI.
- 5:56Third is computer architecture. So, I
- 5:58will be explaining hardware components
- 6:00like the CPU, memory, and storage
- 6:02devices of a computer. Next, the fourth
- 6:04part is about operating systems. The
- 6:06fifth part covers networks. When we
- 6:09talk about networks, things like the
- 6:11Wi-Fi we use in our daily lives all
- 6:13fall under this category. The sixth
- 6:16part is databases. And the grand finale
- 6:18, the seventh part, is software
- 6:19engineering. Software engineering might
- 6:21sound quite unfamiliar to you. But it
- 6:23is one of the most important elements
- 6:25in computer science. I didn't just make
- 6:27up this curriculum; it’s based on
- 6:29computer science programs at Seoul
- 6:30National University and other
- 6:32world-renowned institutions.
- 6:34Universities like MIT, Carnegie Mellon,
- 6:36and Harvard all teach these subjects as
- 6:38requirements. I have structured this to
- 6:41include essential subjects that are
- 6:43recognized by experts all around the
- 6:46world. I will see you again in the next
- 6:48video, where I’ll explain the history
- 6:51of computers in Part 1. If you don't
- 6:53want to miss this "Bottom-Up" series,
- 6:54make sure to subscribe! I am creating
- 6:57videos to help developers achieve "vibe
- 7:00coding." If you subscribe, you will
- 7:02gain a lot of great insights. I was
- 7:04your developer friend who knows his
- 7:06stuff. Thank you.
About this transcript
This page contains the full transcript of AI가 다하는데 왜 컴퓨터공학을 알아야 할까? (밑바닥 시리즈 OT) by 아는개발자, generated from the public captions YouTube serves with the video. The transcript has 1,119 words across 174 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 choose a caption track available on the video.
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, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.