YouTube2Text

MIT 물리학&컴공생이 체감중인 AI가 바꾸고 있는 부의 미래 | 우연히 만난 사람들 ep.09 — Transcript

by 희야기 | HeeChan · 3,143 words · 500 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Excuse me, sir. May I ask you something?
  2. 0:03>> Yes, sure.
  3. 0:03>> I'm just a content creator from South
  4. 0:05Korea.
  5. 0:05>> Oh, okay.
  6. 0:06>> And I'm just taking some interview with
  7. 0:07MIT students. Is it okay to take your
  8. 0:10interview?
  9. 0:11>> Yeah, sure.
  10. 0:11>> Oh, really? Is it okay to sit on there?
  11. 0:14>> Yeah, yeah. Whatever is fine.
  12. 0:15>> Yeah. I heard that this is a exam
  13. 0:17duration.
  14. 0:18>> Uh we have
  15. 0:20some final projects now.
  16. 0:21>> Yeah.
  17. 0:22>> So, I have a final quiz next Monday. And
  18. 0:25then the week after we have our finals.
  19. 0:27>> Oh.
  20. 0:27>> So, I think starting the
  21. 0:3017th through the 20th
  22. 0:32>> Oh.
  23. 0:32>> are finals.
  24. 0:33>> So, you have a little time for preparing
  25. 0:35some exam.
  26. 0:36>> Yeah, we have we have like a
  27. 0:38last day of classes is next Wednesday.
  28. 0:40>> Mhm.
  29. 0:40>> And final exams start on Monday.
  30. 0:43>> Mhm.
  31. 0:43>> So, we have the couple days in the
  32. 0:45weekend
  33. 0:46>> Oh.
  34. 0:46>> to study.
  35. 0:47>> Sounds good.
  36. 0:47>> Yeah.
  37. 0:48>> Then, uh may I ask you about your major?
  38. 0:50>> Yeah, I'm a major in physics and uh
  39. 0:53computer science.
  40. 0:54>> Oh. Are you born in Boston in Boston?
  41. 0:57>> No, no. I'm born in California.
  42. 0:59>> California. Then, why did you came to
  43. 1:01Boston MIT? Because I heard that there's
  44. 1:03also Stanford in
  45. 1:05California.
  46. 1:06>> Stanford is a very good uh university. I
  47. 1:07I didn't like uh the culture at
  48. 1:09Stanford. I feel it's too focused on
  49. 1:11startups, too focused on I don't know,
  50. 1:14things in the Bay Area. MIT people are
  51. 1:16more interested in like learning. All
  52. 1:18more people here like have the Olympiads
  53. 1:20or they want to pursue a PhD and really
  54. 1:22like learning something very focused.
  55. 1:25Stanford, I feel people want to build
  56. 1:26something and build companies,
  57. 1:28organizations. And that's just not what
  58. 1:30my interests are.
  59. 1:32>> So, you're just interested in studying
  60. 1:34kind of thing.
  61. 1:35>> Yes, yes. I love learning. Yes, yes,
  62. 1:37yes.
  63. 1:37I'm I'm a very big learning guy. I mean,
  64. 1:39I always liked reading as a kid. I would
  65. 1:41just read encyclopedias all the time. I
  66. 1:43would love watching documentaries and
  67. 1:45learn more about the world. Like I love
  68. 1:47studying for classes.
  69. 1:48>> Yeah.
  70. 1:48>> Um just I really like lecture notes. I
  71. 1:51like textbooks.
  72. 1:53Um that's what I like to do the best.
  73. 1:55>> Then, is it okay to ask what grade are
  74. 1:57you? Oh, I'm a junior. Junior. Then are
  75. 1:59you preparing for getting a job now?
  76. 2:01Yeah, so I have to actually prepare for
  77. 2:04interviews for next summer
  78. 2:07like job offers this summer so
  79. 2:10so like different different careers have
  80. 2:13different like hiring times. So you have
  81. 2:15private equity investment banking which
  82. 2:17are the earliest they start before so
  83. 2:19actually you you you interview for those
  84. 2:22internships for next summer before you
  85. 2:24start the internship for the current
  86. 2:25summer. So those interviews are very
  87. 2:27early and then tech typically starts
  88. 2:30like 9 months before the summer so it's
  89. 2:32starting in September you have to
  90. 2:33prepare for those interviews.
  91. 2:35But then if you want to do something in
  92. 2:37the startup space
  93. 2:38startups start or typically hire most
  94. 2:41around like March before the summer. Do
  95. 2:43you have your dream job? Uh no I don't I
  96. 2:45don't really have a dream job. I like
  97. 2:47it's uh
  98. 2:49I like if if if I had to work in a
  99. 2:51vacuum from the world like if I did if
  100. 2:54what I did didn't impact the world at
  101. 2:56all
  102. 2:57or like you know if money wasn't like a
  103. 2:59concern I would work in like academia. I
  104. 3:01think research is something I I really I
  105. 3:03think I I would enjoy the most but I
  106. 3:06feel in order to establish a career in
  107. 3:08academia there's a huge focus on you
  108. 3:10know papers and putting yourself forward
  109. 3:13as a career like in terms of your career
  110. 3:15rather than like focusing on like
  111. 3:17learning and just reading. So
  112. 3:20I don't know. I I don't think that
  113. 3:22there's a dream career which exists for
  114. 3:23me but uh there's that. There's
  115. 3:26different research opportunities that I
  116. 3:27would enjoy doing.
  117. 3:29And sometimes when I met someone who is
  118. 3:31interested in math they wanted you to go
  119. 3:33to like master or PhD kind of that but
  120. 3:36don't you ever think like that?
  121. 3:38Yeah, so um in amazing we have the
  122. 3:40masters of engineering program. So if
  123. 3:42you're in the computer science
  124. 3:43department especially you can do a one
  125. 3:45year's masters at MIT. And so I actually
  126. 3:47plan on doing this. I'll do my bachelors
  127. 3:49and then I'll do my masters one year
  128. 3:51later as well. And so, I'm actually
  129. 3:53already most of my class I've taken
  130. 3:55right now are graduate level classes
  131. 3:56even though I'm an undergrad. And so,
  132. 3:58these are to fulfill my master
  133. 3:59requirements later. PhD is like it's
  134. 4:02still on the table if I don't really get
  135. 4:04an internship or a job I really like in
  136. 4:06the near future, then I will you know
  137. 4:08consider going [screaming] into PhD cuz
  138. 4:10a lot of my experience is anyways in
  139. 4:11research. But the thing with PhD in like
  140. 4:13this like current age is things are
  141. 4:15changing so fast. With PhD you're like
  142. 4:18doing 4 years in a university sort of
  143. 4:20separate [music] from like what's going
  144. 4:22on in the world. And so,
  145. 4:24what you do a PhD in, by the time you
  146. 4:27apply and get done with it, it might be
  147. 4:29completely different world coming out
  148. 4:30and your skills that you could get in
  149. 4:31the PhD might not be as useful. So,
  150. 4:33that's why I'm scared to do a PhD now as
  151. 4:35well. But yeah. Then it's just my love
  152. 4:38guess, but when I met someone who is
  153. 4:40majoring computer science and math, they
  154. 4:42are always not always but most of them
  155. 4:44are like interesting in quant investing.
  156. 4:47And
  157. 4:48we are so interested like interested in
  158. 4:50quant investing. Uh yeah, so there's a
  159. 4:52field of quantitative finance which MIT
  160. 4:55does recruit a lot from. These are
  161. 4:58majors which which include software
  162. 5:00engineers, includes infrastructure
  163. 5:02people, includes traders, and includes
  164. 5:04researchers. And so, depending on like
  165. 5:07what you want to do, there's different
  166. 5:08requirements. But most often you need a
  167. 5:10good understanding of like probability,
  168. 5:12some basic understanding of how to code,
  169. 5:14algorithms, and just like a
  170. 5:16good ability and good ability to work
  171. 5:19with like into this math. And so, yeah
  172. 5:22this is a a career that a lot of people
  173. 5:24like MIT is sort of one of like the if
  174. 5:26you were to sort of rank careers in
  175. 5:28terms of like the most sought-after,
  176. 5:30this would be one of the top ones.
  177. 5:32Although they're changing now, I feel a
  178. 5:34lot of the people who I know got like
  179. 5:35quant offers summer are actually going
  180. 5:37to go into like open AI
  181. 5:39because open AI research is seen as like
  182. 5:41very prestigious thing
  183. 5:43especially in any of like you know the
  184. 5:45AI frontier labs and the Anthropic
  185. 5:46fellowship. Um, all these sorts of
  186. 5:48things are very sought after.
  187. 5:50>> Then there's some more philosophical
  188. 5:52questions. Yeah. And is it okay to your
  189. 5:54time?
  190. 5:54>> Yeah, sure. Sure.
  191. 5:55>> Then do you think also like AI
  192. 5:57evolution, like the technology
  193. 5:58evolution, makes world better like all
  194. 6:01over the world?
  195. 6:02>> Oh, it'll make the world better. No,
  196. 6:04yeah. No?
  197. 6:04>> No, I don't think so.
  198. 6:05>> Why?
  199. 6:06>> Um, I think for a lot of reasons. I
  200. 6:08think more in this like current day and
  201. 6:11age with like limited regulation on AI
  202. 6:13models
  203. 6:14>> Yeah.
  204. 6:14>> and AI companies, what will happen is
  205. 6:17people will continue investing in the
  206. 6:18future that AI models will be able to
  207. 6:20like take over like all sorts of work.
  208. 6:22>> Yeah.
  209. 6:22>> And so you're going to get more and more
  210. 6:24wealth accumulation in these like fewer
  211. 6:26companies and they will be able to, you
  212. 6:29know, build more data centers, destroy
  213. 6:31the environment more, or like uh, steal
  214. 6:34people's like in- the information to
  215. 6:36train their models. And, you know, use
  216. 6:38like the labor and minerals in other
  217. 6:40like countries to sort of build their
  218. 6:43technology further and further more.
  219. 6:45It's just one of those things that like,
  220. 6:46you know, when you have a technology
  221. 6:48which limits the amount of people
  222. 6:50>> Yeah.
  223. 6:50>> that have access to it, right? Like
  224. 6:52these AI models, like right now, a lot
  225. 6:54of people can use them, right? Like sure
  226. 6:56there's like like plans that are like,
  227. 6:58you know, $20 a month, which is like not
  228. 7:00too expensive, you know, for the average
  229. 7:02like American, for example.
  230. 7:04>> Yeah.
  231. 7:04>> But these like companies are right now
  232. 7:06non-profits. They're doing this at a
  233. 7:08loss. So, in the future when these
  234. 7:10companies do want to make a profit, when
  235. 7:12they have to, you know, make money,
  236. 7:14these things are going to get more and
  237. 7:16more expensive and are not going to be
  238. 7:17accessible
  239. 7:18>> Yeah.
  240. 7:18>> to normal people.
  241. 7:19And so the people who have access to
  242. 7:21technology are going to get smaller and
  243. 7:22smaller and smaller. And the people who
  244. 7:24have jobs are going to get smaller and
  245. 7:25smaller and smaller. So, the majority of
  246. 7:27society is [music] going to be left
  247. 7:28behind.
  248. 7:29>> Mhm.
  249. 7:29>> Right? You need regulation. You need
  250. 7:31things that are able to protect like the
  251. 7:33majority of people
  252. 7:35>> Yeah.
  253. 7:35>> to make sure the world doesn't become a
  254. 7:36worse place with these sorts of things.
  255. 7:37I do think there is a good like there is
  256. 7:39a a world that can exist with AI, which
  257. 7:41is which is better than our world right
  258. 7:42now.
  259. 7:43But, with assistance in place, I feel we
  260. 7:45are like very, very unprepared for this
  261. 7:47type of life situation
  262. 7:49and this type of life Yeah, society.
  263. 7:51>> Do you think that it's possible to
  264. 7:52evaluate the technology with the
  265. 7:54ethically?
  266. 7:55>> Um yes, I I do think like we can make
  267. 7:58ethical AI technology, but we need first
  268. 8:00first we have to make sure we know
  269. 8:02what's going on with these models.
  270. 8:03First, we have to like be able to
  271. 8:04understand them, make sure that they're
  272. 8:06capable of making the [music] right
  273. 8:07decisions. Cuz right now, a lot of AI is
  274. 8:09a black box. What we need is like, you
  275. 8:11know, actual research and like AI safety
  276. 8:14research mech and term, which will help
  277. 8:16us better understand how these like
  278. 8:17models operate. Second, we need to like
  279. 8:19guidelines and policy within the
  280. 8:21government to ensure that these
  281. 8:22companies are held accountable for the
  282. 8:24models and that the the productivity
  283. 8:27that is increased by these models go
  284. 8:29back to the people which helped make it
  285. 8:31in the first place. Surely like this to
  286. 8:32didn't help to improve society in
  287. 8:34general.
  288. 8:35Right? We need we need mechanisms to to
  289. 8:37make these things better for society.
  290. 8:39And yeah, I mean we just need like less
  291. 8:40profit-based thinking in general. I
  292. 8:42think
  293. 8:43with like the like abundance that like
  294. 8:45AI could potentially provide us, we need
  295. 8:48to think about, you know, what type of
  296. 8:49society we want to become. I think now
  297. 8:52more than ever, like we don't have to
  298. 8:54rely on the like the free market to sort
  299. 8:56of
  300. 8:57run our society. There's so many signals
  301. 8:59like we're we're running like in like
  302. 9:01There's so much information available in
  303. 9:03the world that transactions like markets
  304. 9:06are just going all over the place, you
  305. 9:08know, we have like these like poly
  306. 9:09markets, policy things, betting
  307. 9:11platforms where where people are just
  308. 9:13like, you know, like these markets are
  309. 9:14huge. More and more people are just like
  310. 9:16putting their money in like a losing
  311. 9:18system.
  312. 9:19And I I think it's come to a point where
  313. 9:21like the the markets are exploiting
  314. 9:22people rather than benefiting them. So,
  315. 9:24I think we need to see like a a big like
  316. 9:25economic change in the world. And I
  317. 9:27think this has to come from mass
  318. 9:29movements within people, not just uh
  319. 9:31yeah, not not just through elected
  320. 9:32officials.
  321. 9:33>> That's interesting.
  322. 9:34>> Yeah.
  323. 9:34>> You really like think a lot of things
  324. 9:36with this topic.
  325. 9:37>> Well, yeah. I think I think MIT is a
  326. 9:39great place to talk to people about
  327. 9:41these sorts of things. Everyone's
  328. 9:41[music] concerned about
  329. 9:43you know, how these things will be
  330. 9:45applied in the real world. You know,
  331. 9:46there's also clubs on campus such as
  332. 9:48Maya, which is MIT AI Alignment, which
  333. 9:50talks about safety in regards to AI
  334. 9:52models. There's Impact at MIT, which is
  335. 9:55more of an effective altruist
  336. 9:56organization which talks about you know,
  337. 9:57the ethics of different things in the
  338. 9:59world and how we as students can
  339. 10:00contribute to these things.
  340. 10:02>> Mhm.
  341. 10:02>> And so I think MIT does allow these
  342. 10:04sorts of conversations to be held in
  343. 10:05like
  344. 10:07we'll stay up late at night just talking
  345. 10:08about these sorts of things as well. So,
  346. 10:10yeah. Yeah.
  347. 10:11>> Then sometimes when I start like search
  348. 10:14some media, a bit of a media said like,
  349. 10:16"This is the last chance to be a leech
  350. 10:18because AI will break up a lot of
  351. 10:21capitalism thing." And do you agree
  352. 10:23about that opinion, too?
  353. 10:25>> Will AI break up capitalism? I don't
  354. 10:28think so.
  355. 10:28>> You don't think so?
  356. 10:29>> I I I um I think currently AI will
  357. 10:33benefit the technocrats.
  358. 10:35>> Yeah.
  359. 10:35>> Right? The people The people who own the
  360. 10:37AI companies. They will be in the sort
  361. 10:40of in control of the markets. They'll be
  362. 10:42like they'll
  363. 10:43the as the wealth
  364. 10:46concentration
  365. 10:47deepens further in the direction of
  366. 10:48these technocrats, like as they gain
  367. 10:50more and more money with all of the rest
  368. 10:51of people, they will control the markets
  369. 10:53more and more.
  370. 10:53>> Mhm.
  371. 10:54>> And I feel this whole rather I guess it
  372. 10:57will break capitalism in the fact that
  373. 10:58like the markets won't be free to
  374. 11:01everyone. It won't be your typical, you
  375. 11:02know, like like the the the free market
  376. 11:04capitalism won't work cuz you'll just
  377. 11:05have like monopolies of people. And like
  378. 11:07these these individuals or these
  379. 11:09corporations will basically be
  380. 11:10controlling the the the entire the
  381. 11:12entire economy. And I think this will
  382. 11:15just become like I have I can't separate
  383. 11:16the politics from like the economic
  384. 11:18principles here. I I think this will
  385. 11:20just become like a technocratic society
  386. 11:21where you know, the technocrats are in
  387. 11:23charge of the government. They sort of
  388. 11:24make the policy decisions. And
  389. 11:27yeah.
  390. 11:28>> Yeah.
  391. 11:28>> But I do think it will be
  392. 11:30if AI continues advancing and there
  393. 11:33isn't regulation, there isn't resistance
  394. 11:35to this movement, like average day
  395. 11:37people won't be able to contribute to
  396. 11:38the economy. Like we will be shut off
  397. 11:40from the market, we will be shut off
  398. 11:41from the government, we will be shut off
  399. 11:42from society and our contribution to it
  400. 11:44cuz our labor won't mean anything. Our
  401. 11:48our like voices can be suppressed by
  402. 11:50like propaganda and misinformation.
  403. 11:53And so yeah, now is like like yes, like
  404. 11:55now is sort of the last time where we
  405. 11:57can sort of make a difference and you
  406. 11:59know, like make a change in the world I
  407. 12:00think. Because yeah, these types of
  408. 12:01opportunities are just [music] not going
  409. 12:02to exist in the future if things
  410. 12:04continue on the same trajectory they are
  411. 12:05now.
  412. 12:05>> What can I do
  413. 12:06for changing the world better?
  414. 12:09>> Great question. I think I think the best
  415. 12:11the best change currently is through
  416. 12:13politics. Um
  417. 12:15I think we are changing a lot as a
  418. 12:16society right now in terms of
  419. 12:18technology, in terms of culture,
  420. 12:21media. And
  421. 12:23with these types of changes, the only
  422. 12:25way to address things at the systemic
  423. 12:27level is through like under like looking
  424. 12:29at political structures, looking at
  425. 12:31ideologies and how like do we
  426. 12:34implement them in the current society. I
  427. 12:36think like I believe in like people
  428. 12:37power, I believe in like mass movements,
  429. 12:40I believe in organizations which
  430. 12:41advocate for everyone. And so I would,
  431. 12:46you know, support joining these types of
  432. 12:47organizations. And so I think using
  433. 12:50labor right now while we still have, you
  434. 12:53know, the power of labor against these
  435. 12:54corporations, I think is really
  436. 12:56necessary. Cuz in the future this this
  437. 12:58type of power may not exist. So yeah, I
  438. 13:01would say yeah, just join join
  439. 13:03organizations, see what you can do to
  440. 13:05make movements in society. I still have
  441. 13:07to figure this out honestly. I'm not I
  442. 13:09want to be more of a political person. I
  443. 13:11want to like sort of have these ideas. I
  444. 13:12need to get more into like political
  445. 13:13literature and learn from the past to
  446. 13:16sort of
  447. 13:16see what kind of
  448. 13:18what kind of directions I can explore in
  449. 13:19the future to make a society a better
  450. 13:21place. I really want to. And I think
  451. 13:22like now is the time. Now is like the
  452. 13:24last chance we have.
  453. 13:26>> Then if if we really about the world,
  454. 13:29what's the most underrated value in this
  455. 13:31modern society do you think?
  456. 13:32>> Underrated value?
  457. 13:33>> Yeah.
  458. 13:34>> I would say in general, yeah, I guess I
  459. 13:36guess talking to people like in person
  460. 13:38and you know, making like real life
  461. 13:40connections is probably going to be the
  462. 13:42most important thing we can do as a
  463. 13:43society to stay connected.
  464. 13:45>> Yeah.
  465. 13:45>> You know, as it becomes easier to like
  466. 13:47talk online or get get like sort of
  467. 13:48brainwashed by the the amount of media
  468. 13:51that exists online, there's so much you
  469. 13:53know, content already online and with AI
  470. 13:55generated content it's going to get more
  471. 13:57and more worse. There's going to be so
  472. 13:57much noise. You won't be able to tell
  473. 13:59what's real or what's fake on what's
  474. 14:01online. So I think, you know, talking to
  475. 14:03people in your life, understanding
  476. 14:05what's going on in their lives, getting
  477. 14:07involved in your community,
  478. 14:08understanding how you can make a
  479. 14:09difference there, I think it's really
  480. 14:11important. This is like how we sort of
  481. 14:13connect to others, how we, you know,
  482. 14:14stay stay like human as a society.
  483. 14:17Otherwise, we'll just we won't know what
  484. 14:19we are, you know? Like if you look at a
  485. 14:21lot of people nowadays, when they like
  486. 14:23only look at like, you know, one news
  487. 14:24channel
  488. 14:25>> Yeah.
  489. 14:25>> or only look at like content coming from
  490. 14:27like specific sources, they become
  491. 14:29blindsided to a lot of things inside and
  492. 14:31and they live in a world which isn't
  493. 14:32real, right? They they live in a world
  494. 14:34that's a make up on their own and
  495. 14:36they're disillusioned by things and
  496. 14:38yeah, I I think you really need to like
  497. 14:39make sure you're you're you're living in
  498. 14:42the real world if you want to make a
  499. 14:43real world impact.
  500. 14:56>> [music]

About this transcript

This page contains the full transcript of MIT 물리학&컴공생이 체감중인 AI가 바꾸고 있는 부의 미래 | 우연히 만난 사람들 ep.09 by 희야기 | HeeChan, generated from the public captions YouTube serves with the video. The transcript has 3,143 words across 500 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.