NeetCode Ties 90% of Coding Interview Success to Memorization

NeetCode：代码面试成功九成靠记忆模式 技术面试考核的是模式记忆而非工程智力，求职者只需掌握三分之一的Medium难度题目，就能通过重复练习和概率优势斩获大厂Offer。 NeetCode回顾了自己在亚马逊入职仅两个月便因严重的文化冲突和倦怠而仓促离职的经历，随后他花了一年时间制作算法教学视频，并最终借此敲开了Google软件工程师职位的大门。他揭开了技术招聘的内幕，坚称约90%的面试表现取决于模式记忆而非天生智商，并指出能解出大约三分之一的LeetCode Medium题往往就足以通过技术初筛。在对比两位前雇主时，他批评Google开发节奏迟缓且内部官僚作风过重，这与亚马逊极度注重交付的风格形成了鲜明对比。最后，他谈到了全职运营NeetCode的转变过程、从不露脸博主走向公众视角的经历，并解释了为何有抱负的工程师应当深耕自己真正感兴趣的方向，而非盲目跟风追逐热点。 通过科技大厂的代码面试所需的顶尖技能远低于应聘者的想象：能解出约三分之一的LeetCode Medium题就具备竞争机会，且约90%的临场表现来自模式记忆而非纯粹的现场推导。 (9:00) 由于严重的文化冲突以及团队深陷倦怠（其经理第一天就警告他不要问太多问题），NeetCode在没有后路的情况下，入职亚马逊仅两个月便辞去了第一份工作。 (0:37) OpenAI和Anthropic等前沿AI实验室目前设置了市面上难度最高的技术面试，它们不再局限于孤立的算法谜题，而是在禁止使用AI助手的情况下，重点考核多文件代码库和系统推理能力。 (24:40) 尽管Google提供了与亚马逊残酷竞争截然相反的协作文化，但其僵化的内部流程和缺乏紧迫感严重阻碍了高效交付，最终促使NeetCode走向全职创业。 (29:38)

Coding educator NeetCode argues that technical interview success is largely driven by structural pattern memorization and variance rather than raw intelligence, contrasting his experiences navigating corporate cultures at Amazon and Google with running an independent educational platform. Interview Realities: Memorization and Variance (9:00) NeetCode dismisses the idea that LeetCode measures innate human intelligence or that candidates must master Hard-tier problems to receive offers. Instead, he estimates that roughly 90% of interview performance stems from pattern memorization and structural intuition, noting that deriving complex algorithms from first principles within a 45-minute window is practically impossible. Because interviewing functions as a numbers game influenced by question selection, solving approximately one-third of Medium problems can be enough to secure a job. He differentiates typical interview preparation from competitive programming platforms like Codeforces, where the skill distribution is far wider. To illustrate the gap, NeetCode recalls a collegiate competition where an elite team of standard software engineering candidates failed to solve nearly any competitive programming questions, while a Caltech team completed all 13. High performance, in his view, is heavily domain-specific and driven primarily by prolonged consistency and genuine interest rather than genius. Cultural Polarity Between Amazon and Google (0:37) Reflecting on his time in Big Tech, NeetCode outlines contrasting operational dysfunctions across major employers. At Amazon, which he abruptly left after two months as an broke 22-year-old, he encountered an intense, cutthroat atmosphere characterized by team burnout and high turnover. He candidly notes that he passed the behavioral interview by intentionally misrepresenting himself to match the firm's high-agency Leadership Principles despite being inherently risk-averse. In contrast, Google offered a collaborative, supportive environment with attractive perks, but suffered from an agonizing lack of urgency. NeetCode found that Google prioritized strict internal process and employee sentiment over execution velocity, making it frustrating for engineers attempting to ship quickly. He characterizes the two companies as philosophical opposites: Amazon drives rapid delivery at the expense of employee burnout, while Google preserves work-life balance at the cost of momentum. Transition to Full-Time Content Creation (18:00) After nearly a year of recording walkthrough videos while unemployed, NeetCode joined Google, but eventually left corporate software engineering to pursue content creation full-time. He explains that building out algorithms tutorials provided intrinsic satisfaction that routine corporate maintenance lacked. While he operated anonymously for his first three years, revealing his face allowed him to build relationships with peers, though stepping into managerial duties for his platform presented a steep learning curve. Addressing his controversial video titled 'Maybe Some People Should Just Give Up', NeetCode clarifies that he was not advising students to quit, but expressing frustration akin to a disappointed teacher when viewers asked basic questions answered directly in the material. While he acknowledges that candidates can validly choose alternative paths like networking instead of solving hundreds of problems, he maintains that almost anyone who genuinely enjoys programming has the capacity to build LeetCode proficiency. Shifting Frontiers in Systems and AI Interviews (24:40) Analyzing current hiring trends, NeetCode observes that frontier AI labs like OpenAI and Anthropic maintain higher technical bars than traditional Big Tech, driven largely by talent saturation and low headcount. Rather than isolated algorithmic puzzles, these labs frequently test candidates on multi-file codebases, systems reasoning, and applied debugging, while still restricting the use of AI tools during the interview process. For broader technical growth, NeetCode recommends Martin Kleppmann's 'Designing Data-Intensive Applications', highlighting how bottom-up architectural study reframes systems-level thinking. Reflecting on his career, his primary advice to engineers is to pursue niche domains they find intrinsically engaging, such as databases or systems internals, rather than defaulting to popular tracks like web development out of perceived career safety. NeetCode views algorithmic interviewing as an imperfect, domain-specific filter that rewards targeted preparation over intellectual superiority, though how technical assessments will permanently adapt to modern AI tooling remains an unresolved industry experiment.

[0:00] NeetCode: I'll stop there before I like do more reputational damage. [0:37] NeetCode: I couldn't afford to quit, but I still did it anyway [0:37] NeetCode's Manager: Okay, like the team's a little burnt out, try not to ask too many questions. [4:08] NeetCode: I don't know if I should say this, but it's been like 7 years at this point [4:08] NeetCode: I think it's important to be at least a little bit honest with that because I think like I lied for every single question for that [6:01] NeetCode: Probably not. And I don't think I was ever actually that good. [6:01] NeetCode: I don't know how people are going to feel about this [18:00] NeetCode: I don't recommend anybody to give up. I really don't. I swear. [18:00] NeetCode: I know you're better than this... you're too smart to like give up that early. [26:51] Fan / Viewer: Hey, like I didn't know like you're the guy who did that [27:47] NeetCode: every person out there needs to quit their job or do their own side hustle.

[0:00] Anyway, [laughter] I'll stop. I'll stop [0:01] there before I like do more reputational [0:03] damage. This is Node. He's the most [0:06] well-known creator teaching coding [0:08] interview prep. And I asked them all [0:10] about Le Code and his career story at [0:12] Google and Amazon. Where does that place [0:15] you in the percentile of all ELOs? [0:18] >> I don't know how people are going to [0:19] feel about this. [0:20] >> How's your experience quitting Google [0:22] and doing your own thing full-time? [0:24] >> I don't know if I should say this, but [0:25] it's been like 7 years at this point. Do [0:28] you ever miss Google a lot? Like I miss [0:31] Here's the full episode. [0:37] >> You ended up unemployed. What What was [0:39] going on at the time? [0:40] >> Yeah, it's a funny story, I guess, now [0:42] uh in hindsight, but basically um like I [0:46] started working I got my full my first [0:48] full-time offer at Amazon. I'd heard it [0:50] was like, you know, like a pretty tough [0:52] environment from like the online stuff. [0:55] the things that you see on Reddit are [0:56] blind. Um, but for some reason I just [0:59] wasn't like nervous. I just thought [1:00] like, okay, it'll be fine. And like as [1:03] much as I would use like the word [1:05] intense, I think it was more so that it [1:07] was like a culture shock for me, which [1:08] is like I was very like introverted, [1:10] didn't do like too many activities or [1:12] extracurriculars. And then so being in a [1:15] place where it's not really academic, [1:18] it's like there's a lot that goes into [1:19] like having like a full-time job and [1:20] like working with people and like all [1:22] those other things, that was really [1:24] really tough for me. And so um among [1:27] like other reasons, like Amazon was not [1:29] like the the ideal environment. In [1:31] hindsight, I probably could have [1:32] survived if I had to redo it, but I just [1:34] didn't really know things at that point. [1:37] And so for whatever reason, I just [1:39] decided to quit after like a couple [1:41] months. And it was just a really dumb [1:42] decision because it's not like I had [1:43] like a backup plan or I was like like I [1:46] was broke like I couldn't afford to quit [1:48] but I still did it anyway. So it was [1:50] just uh I'm glad it worked out but I [1:52] wouldn't probably recommend it to most [1:54] people. [1:54] >> What did you see at Amazon that made you [1:56] want to quit so quickly? [1:58] >> Yeah. Um so it was like I said a very [2:01] intense environment but I think many [2:03] people have dealt with that. I think [2:05] there's intense environments that are [2:06] like people are still happy. people are [2:09] like trying to perform like at the top [2:11] of their game and it's like kind of like [2:12] a almost like a sport I guess where [2:15] people are super competitive and they [2:16] want to do the best that they can. But [2:17] there are other environments I think [2:19] that are like super cutthroat and it's [2:23] not people trying to do their best. It's [2:25] like people you can tell people are not [2:27] like rooting for each other. Like people [2:29] are I wouldn't say they're trying to [2:30] like screw each other over. I wasn't [2:31] really there long enough to judge that [2:33] but you can tell people are not [2:35] friendly. Like there were some like [2:37] meetings like ice breakers that we had [2:40] and like nobody [laughter] like all it [2:43] was like 10 people like five of us were [2:44] like new grads literally we all had [2:46] joined within the first month. The other [2:48] half were like super burnt out, [2:50] experienced people like uh like [2:53] mid-level and seniors and stuff and they [2:55] were just like very unhappy and that [2:59] like spreads pretty quickly to the rest [3:01] of the team and like I said like I I [3:03] think every other new on the team [3:04] probably survived that experience like I [3:06] didn't really keep in touch with them. I [3:07] don't think they probably got pipped or [3:09] whatever, but it was clearly like most [3:13] people left that team I think like [3:14] within a year. Like I have them all [3:15] added on LinkedIn. So they were not [3:17] happy there either, I don't think. [3:19] >> What gave you the confidence to just [3:21] quit your job so quickly? [3:23] >> Yeah. I think it was like a very [3:25] uncharacteristic thing of me. I think I [3:27] was like 22 and like it's so weird [3:31] because right before I had started that [3:32] job, I was like I'm on top of the world. [3:34] Like I'm like everything is going so [3:36] well. Like I have like I was I was broke [3:38] and now I have like a job and like okay [3:39] like it it might be stressful but like [3:42] I'll have it for the rest of my life or [3:44] maybe I'll switch to another [laughter] [3:45] job. Like it's so crazy like how quickly [3:48] like I was like literally on the moon [3:50] and then it all just like plummeted [3:52] overnight. Like it was like my first day [3:54] my my manager is like okay like the [3:56] team's a little burnt out trying not to [3:58] ask too many questions and so I'm just [3:59] like okay like whoa like what's [4:01] [laughter] going on? like it it's like a [4:03] horror movie like it keeps getting worse [4:04] and worse. But yeah, anyways, [4:08] >> did you have to do le code to get that [4:10] job? [4:11] >> Yeah, you know the funny thing is I [4:12] don't know if I should say this but it's [4:14] been like seven years at this point. Um [4:16] the the le code questions for Amazon [4:18] were actually like surprisingly easy. [4:20] People always say that you can cheat [4:22] now. At that point I I don't remember [4:24] the exact question. It was like some [4:25] variation of twosome and questions that [4:28] really weren't that difficult. Um, they [4:31] really grill you on the leadership [4:32] principles though. And I think it's [4:35] important to be at least a little bit [4:37] honest with that because [laughter] I [4:38] think like I lied for every single [4:40] question for that. Like because it's [4:41] like a personality test. It's like, [4:43] okay, if you're in this situation, what [4:44] would you do? And I know what they're [4:46] looking for. Like they're looking for [4:47] like the personality type, who has like [4:50] high agency, like they're not really [4:51] scared. They're going to like they're [4:52] going to take risks and they're going to [4:54] just, you know, they're going to show up [4:55] and stuff. And so I gave answers like [4:57] that, but that's really not my [4:59] personality. [laughter] [5:00] It's like very like the opposite where [5:02] like I'm very riskaverse and like I try [5:04] not to like mess things up and like I [5:06] take initiative but at Amazon like they [5:08] want like really high initiative where [5:10] like you're on your first day you're [5:11] just like making decisions. So it took [5:14] some like getting used to. [5:16] >> Well I ask cuz when I when I was doing [5:19] recruitment I remember this is a long [5:21] time ago. Amazon, they did some weird [5:24] thing where they had like this group SAT [5:27] at my college where they sat down [5:30] hundreds of us and gave us these it's a [5:32] sequence of logic puzzles basically. It [5:35] would be something like I mean this [5:37] makes it this makes it sound really [5:39] trivial but it it'd be like multiple [5:41] choice and be like you know A B C D [5:45] colon and then it'd be like you know uh [5:49] E FG blank. what's the remaining is just [5:54] you know continue the pattern kind of [5:55] stuff. [5:56] >> I think I had something like that [5:57] actually. It was like a online thing. It [5:58] was like one section of it though. [6:01] >> You said you weren't good at leak code [6:02] to begin with. And I mean today like [6:05] let's say I don't know how practiced you [6:08] are right now. If I gave you the hardest [6:10] leak code hard, could you solve it? No [6:13] problem. [6:13] >> Probably not. And I don't I don't think [6:14] I was ever actually that good. Like [6:17] because we code is like such a big [6:18] spectrum where it's like as you get [6:21] better most people can't really tell the [6:23] difference between like a good engineer [6:25] and like a like a 10x engineer and [6:27] that's like the level like that most [6:29] people I think see like leak code where [6:31] like somebody who's really really good [6:32] at leak code hards like they're just [6:33] indistinguishable but there's like [6:35] there's other platforms like code forces [6:37] I think is like the other big one where [6:39] it's like the distribution is really [6:41] like known because you you'll be like on [6:44] the completely other end of like [6:45] somebody else who's better than you, but [6:47] you might be the same uh like on the [6:49] leak code spectrum as they are, but like [6:52] on a different platform, it's much [6:53] wider. But I I guess to answer your [6:54] question, um yeah, I'm not I'm like [6:58] decent still, surprisingly. Like if you [7:00] do something for years, it's actually [7:01] hard to forget it. And I just kept [7:02] making videos for years, even when I [7:04] didn't need to. Um so like I'm pretty [7:06] like practiced with like the core [7:08] patterns. I could look at a problem and [7:09] like tell you like what the solution [7:11] might be, but it'd be harder to like [7:12] actually like handcode it up, especially [7:14] now that I feel like most people aren't [7:16] like hand typing code anymore, including [7:18] myself. So, yeah, it it'd probably be [7:21] tough. [7:22] >> What percent of like le code mediums [7:24] could you do right now if I just tested [7:26] you right now, say, you know, do this [7:28] one or whatever? [7:29] >> I mean, I'd be curious, but I I don't [7:32] not all of them for sure. I I think for [7:35] the vast majority of them I could [7:37] probably at least tell you like what the [7:38] solution would be or like roughly like [7:40] the pattern. Um do you want to do a leak [7:42] code? [7:43] >> Yeah, we can try it. [7:44] >> So I was genuinely curious to see how [7:46] good he was at le code and so we [7:48] actually did do an impromptu leak code [7:50] whiteboarding session where I asked him [7:52] nothing but leak code hards and he had [7:54] no time to prepare. If you're curious to [7:56] see how he did, I uploaded it to my [7:58] YouTube channel separately because it [8:00] didn't fit in the podcast. Anyways, back [8:03] to the original conversation. [8:04] >> You mentioned the code forces. I think [8:06] they have an ELO system in that. Right. [8:08] >> Right. [8:09] >> This is my my mental model is there's [8:12] there's leak code is actually the just [8:14] the beginning and leak code easys you're [8:16] you're just barely doing it. Le code [8:18] mediums, you get jobs. Le code hards, [8:21] you're you're a bit better. But there's [8:23] you can go past that many standard [8:25] deviations into code forces. And so [8:29] where where do you think roughly [8:32] uh someone who can pass interviews [8:35] decently well, let's say they can do le [8:36] code mediums pretty well, where does [8:38] that place you in the percentile of all [8:41] ELOs, would you say? [8:43] >> Yeah. So that's really [8:46] I guess kind of funny and I don't know [8:49] how people are going to feel about this [8:50] because I've I I talked to a lot of [8:52] people and I've spoken to a lot of [8:53] people over the last few years and some [8:55] people have the mentality that like if [8:57] you want to get a job at big tech today [9:00] most of them are still doing some [9:01] variation of like leak code. They think [9:03] that they need to like be able to do the [9:05] hardest questions like super hards and [9:07] if they failed their interview it's just [9:08] because they weren't able to do all the [9:10] leak code hards and somebody else did. [9:12] Unfortunately, that's not what usually [9:14] happens. The people who are passing [9:16] their interviews, it's not that like [9:18] they're dumb, like they they are [9:19] competent and they're just as good as [9:21] like anybody else, but most likely they [9:23] were just asked a question that they [9:25] were able to answer and the other person [9:27] just wasn't. And um [9:31] I mean that's provided you meet like [9:33] some bar of like okay, you kind of know [9:35] like the core patterns like you'd be [9:36] able to solve a question you've already [9:38] seen before and stuff like that. Um, so [9:40] I think unfortunately like the answer is [9:43] like it's wider than you think. There [9:44] isn't like an average. I think you could [9:46] be pretty like low on that end and still [9:48] pass interviews. You could be pretty [9:50] high and fail some interviews, but if [9:51] you're like really high like probably [9:53] you won't fail most of them because it's [9:54] like a numbers game at that point, but I [9:56] think the average is lower than a lot of [9:59] people think. I think if you can solve [10:01] like some leak code mediums like a third [10:05] of leak code mediums you have a pretty [10:07] good chance I think of like passing in [10:09] an interview. One story that really [10:11] showed me how big the gaps are is uh [10:16] this was in college. Google hosted some [10:18] event where they invited all these [10:20] colleges to bring their you know a a [10:23] team to go and compete in this set of [10:26] puzzles and one of the rounds was just [10:30] pure competitive programming. The other [10:32] ones were like word association and some [10:34] other things. But then the the [10:38] competitive programming one, my team was [10:41] full of people that were all stacked in [10:43] terms of getting jobs. Like le code was [10:46] no problem for them, but these problems [10:49] were so hard there. I think there was [10:51] like 13 and and no one got any of them. [10:54] And then one of the guys on our team was [10:56] he was a little bit he he was much [10:58] better than all of us and he carried us [11:02] and he did like three problems out of [11:03] like the 13. But the crazy thing is [11:06] there's another college there was [11:08] Caltech. They they did all 13 and they [11:12] just destroyed [laughter] us. [11:13] >> Damn. [11:14] >> So like you know there are [11:15] >> there are levels to this stuff. It's [11:17] it's really crazy. [11:19] >> So I I interviewed someone who's a [11:21] complexity theorist. [11:22] >> I want to know. So clearly competitive [11:24] programmers that's above lead coding. [11:27] Where is complexity theory in all and [11:29] all this stuff, right? I feel like to do [11:32] something and to like learn it or to [11:34] just do anything, the the best way to do [11:37] it is to like actually do it. And so [11:39] like if you're training like another [11:40] skill, it doesn't really work it out [11:43] like the same way. I guess to use like [11:44] an analogy, even with like muscles, like [11:46] I feel like I could be wrong about this, [11:48] but if you're doing like a leg press [11:49] machine, it doesn't actually help you [11:51] that much for your squat. like you think [11:54] it would because it's like the exact [11:55] same muscle, but for some reason it [11:56] actually doesn't. It'll make like a very [11:58] small difference. And I think that's how [11:59] a lot of life works. If you're like [12:01] preparing like for competitive [12:03] programming specifically, you'll get [12:05] pretty good at that, but maybe like your [12:07] communication will suck during like the [12:09] actual interview or or you'll miss [12:10] something else. In my case, I did a lot [12:12] of videos and I think that actually [12:15] doesn't make you that much better at [12:17] solving the problems at a certain point [12:19] because it's more of a distraction [12:20] because it's like if you if you take [12:21] like four hours to make a video, most of [12:23] that is spent making the video and [12:25] trying to understand like somebody [12:26] else's perspective rather than making [12:28] yourself better. So, if I had to guess, [12:31] I don't think like the complexity [12:32] theorists are probably like as good [12:34] under like time constraints and like [12:36] typing as fast as they can. But maybe [12:38] for like really like hard problems, they [12:40] could probably like I don't know, [12:41] they've probably just seen way more [12:43] stuff. So they can kind of like recall [12:44] something like okay like maybe this this [12:46] like algorithm or this technique will [12:48] work on this problem or something [12:49] because a lot of them get like really [12:50] really elaborate at that level where you [12:52] like need like some really crazy obscure [12:54] algorithm you couldn't have possibly [12:56] heard of. There's this idea I think this [13:00] was true throughout when I was learning [13:02] math and you know various things in [13:04] school is there's this idea of like [13:08] deriving something when you need it and [13:10] then memorization in this hierarchy [13:13] >> deriving it is the true intelligence. [13:16] That's where you you really you know you [13:18] you come up into the math. You don't [13:20] know the formulas but you're you're good [13:22] with math and you can figure it out on [13:24] the fly. What percent of performance in [13:28] leak code do you think is is just people [13:31] memorizing the solutions versus [13:34] you know deriving things when they hear [13:36] the the question? If I had to guess, and [13:39] even just from like speaking with [13:41] people, I'd say like at least 90% it's [13:43] people just kind of memorizing one way [13:45] or another, which I think is fine [13:47] because if you look at like the entire [13:48] problem of like interview preparation, [13:50] there's a lot to cover and you just [13:53] can't like possibly do it yourself. But [13:56] the other thing I will say is that you [13:58] can't just memorize. You do have to like [14:02] and this is what makes it really hard [14:03] because you have to cover a lot of stuff [14:05] but at the same time you have to like [14:06] build at least some of the skills to [14:08] like understand what's going on so [14:10] you're not just like blindly memorizing [14:12] or you're not like asking yourself okay [14:14] should I do like DFS or BFS like you [14:16] should still have like some like highle [14:18] understanding and that's really hard I [14:20] think um like actually like covering [14:23] things in like a smart way [14:25] >> I guess since we're talking about this [14:26] memorization versus deriving and you've [14:28] taught so many people leak code. [14:30] >> Is leak code an accurate assessment of [14:34] intelligence? [14:36] >> I think definitely not. I think there's [14:39] components to it. I think like is the [14:41] SAT like a sign of intelligence? I think [14:43] not really, but it's a sign of like a [14:45] lot of things that correlate with it and [14:47] it like some subset of it surely isn't [14:50] intelligence because it's not like a dog [14:52] could like pass those tests or like do a [14:54] leak code problem or something. So like [14:56] yeah, but I think it's [14:58] >> I don't know. I think people get like [14:59] caught up on that because then they've [15:01] like they see some people doing well and [15:03] some people aren't and they're like well [15:05] it just must be intelligence or [15:07] something else like maybe partially but [15:10] I think that's it's not just that like [15:12] I'm just saying this just to say it but [15:13] I actually believe it now like after [15:15] meeting so many people and like seeing [15:17] so much stuff and I'm sure you've seen [15:18] so many like high performing people and [15:20] you've spoken to a lot of like guests [15:21] and stuff. I feel like the the number [15:23] one thing the the number one bottleneck [15:25] is usually like some level of enjoyment [15:28] or passion or work ethic or just [15:30] something of that nature where like [15:32] somebody keeps doing something because I [15:33] don't think I really did anything that [15:35] special. All I did was just made a lot [15:38] of leak code videos when most people [15:40] wouldn't have been willing to do that or [15:41] just not had the time to do that. And I [15:43] think that's what most things boil down [15:44] to. like a lot of the like really smart [15:46] people that you have uh on the podcast, [15:48] I think they just enjoyed what they did [15:50] and then just kept doing it and then [15:52] built their life around that and um then [15:56] it just becomes that. And I think the [15:57] people usually in terms of leak code, [15:58] they're usually people that either did a [16:00] lot of math or programming or some kind [16:02] of like math club when they were kids. [16:04] Uh like to get to a certain level like [16:05] the code forces level and stuff, you [16:08] just have to have done that since you [16:09] were a kid. Like if you're a chess [16:10] player, like you have to just start [16:11] early. What kept you making these Le [16:15] code videos for so long? Because you [16:17] have hundreds of them. The one thing I [16:19] guess I didn't like about school and for [16:21] the most part I liked school but I [16:22] always thought that like the computer [16:24] science classes especially especially [16:26] like every time we would get to a part [16:29] where it's like starting to get really [16:30] interesting like whether it's databases [16:32] or like ML or something else like you [16:33] get to the end of the class and you do [16:35] like your random project or whatever and [16:37] then you kind of know like the core [16:39] stuff of that that class and then that's [16:41] the part where I want to like go 10 [16:43] times deeper. I just want to like do one [16:44] thing and then do it really really well. [16:46] I think that's why I wasn't like in [16:49] terms of like being the best developer I [16:50] could be. I think I'm better at like [16:52] other things than than that. And um [16:56] I I I think I think when I got to leak [16:59] code, I just like enjoyed that and I I [17:01] felt productive as I was doing it even [17:03] though like it it really wasn't on [17:05] paper. I was like making a video getting [17:07] like 50 views and I was enjoying that [17:10] but it it wasn't like I was ever going [17:12] to like make a living off of this or [17:14] anything. So, I just kept doing it just [17:16] cuz I was like stubborn and I enjoyed it [17:18] and I was I thought I was like helping [17:19] people as well. So, I just stuck with [17:22] it. [17:23] >> You had this one video I remember it was [17:25] like maybe some people should just give [17:28] up or what was that the that was the [17:30] title? Yeah, it was like maybe some [17:32] people should just give up and yeah, I [17:34] guess my question to you is do you think [17:36] that anyone could be good at le code or [17:38] are some people inherently [17:41] not able to do it? I think the vast [17:43] majority of people especially anybody [17:45] who like is a programmer and likes [17:46] programming I think they could get to [17:49] that point. I think the I mean they [17:51] might feel like they're not smart enough [17:53] and like they might feel that like many [17:54] steps along the way and I think that'll [17:56] probably like burn them out before they [17:58] can like get to that point because it's [18:00] not like any everyone wants to do [18:02] everything like if they if they want to [18:04] like pass their interviews or for [18:05] whatever reason they need to I think [18:06] they could probably do that but um I [18:09] think most people will find like that [18:11] there are other like most people decide [18:13] for whatever reason there are better [18:15] ways to spend their time like most [18:16] people will start the leak code grind [18:18] and then give up. up. They're like, [18:19] "Okay, screw this." Like, "I'm going to [18:20] get a job another way. I'm going to [18:21] network my way. Like, I'm not I'm not [18:23] about to like do 300 leak code [18:25] problems." And that's perfectly fine, I [18:27] think. But yeah. [18:29] >> Well, so when you say that maybe some [18:31] people should just give up, what's your [18:33] your take on that? [18:35] >> Okay. Um [18:37] Yeah. I I don't like I don't recommend [18:39] anybody to give up. I really don't. I [18:42] swear. I like it for me like I want the [18:46] opposite. I really do. Like I want [18:48] people to like push themselves and to be [18:49] like the best that they can be. And I [18:53] don't know. I guess it's just that like [18:55] the reason I made that video I guess is [18:57] like there's a certain attitude that I [19:01] feel like it just really like turns you [19:02] off. Like sometimes you'll see it in [19:04] school, especially like early school [19:05] where like half the class like is paying [19:07] attention and half the class just [19:08] doesn't really care. And that's fine. Um [19:10] but it's like I don't know. It's like [19:11] the more seriously you take something [19:13] and you spend like a bunch of time like [19:16] making this video explaining all these [19:18] parts and somebody asks you a question [19:20] [laughter] that like you know they could [19:21] have answered. It just makes me like [19:23] disappointed. It just makes me think [19:24] like like I know you're better than [19:26] this. Like like you can like [laughter] [19:28] you you can figure it out yourself. Like [19:30] you can like you're smarter like you're [19:33] too smart to like give up that early. [19:35] Like that's like how I feel. But [19:37] obviously like the words that I said and [19:38] like when I [laughter] like say it a [19:40] certain way like I recognize that people [19:42] aren't going to interpret it that way. [19:44] So I have to live with that. But I think [19:45] I had a lot of people I'd say like 95% [19:48] of people actually like liked that video [19:50] like clicked like on that video which [19:52] was surprising to me. Most people that [19:54] like reached out or commented they said [19:55] that hey like I needed to hear this. [19:57] This was like a wakeup call. And I had [19:59] somebody like message me before their [20:00] like Microsoft intern interview and [20:03] they're like, "Hey, like I really needed [20:04] this. Like I'm going to lock in for the [20:06] next like couple weeks." And I was like, [20:07] "Okay, like I guess I'll leave that [20:09] video up because at least somebody got [20:11] something out of it." But yeah, [20:12] >> I see. So the the energy in that is um [20:16] it's like a disappointed teacher or like [20:19] disappointed. Yeah. [20:20] >> Yeah. because why would I make videos [20:22] that are teaching people and then on the [20:25] same side like tell people that they [20:26] should give up and not do this like it [20:29] doesn't make sense and yeah so like I [20:31] said I can live with like what I said it [20:33] wasn't the nicest thing if I could do it [20:34] over again probably would have I think [20:37] like the only thing I said in that that [20:39] video though I think that was actually [20:40] offensive is that I can't believe [20:42] somebody would ask me this dumb and I [20:44] said another word after that and I said [20:46] question [laughter] that's that's the [20:48] literally the worst thing I said I [20:49] didn't make it personal with that [20:50] person. I didn't call anybody like not [20:52] smart enough or anything, but yeah. [20:54] Anyway, [laughter] I'll stop I'll stop [20:56] there before I like [gasps] [20:58] do more reputational damage. [21:00] >> When when you when you were recording [21:03] this uh you know video cuz I I like the [21:08] video by the way, but when I saw the [21:10] packaging and you know it was like this [21:13] unhinged thing that you [laughter] [21:15] draw. Were you were you worried at all [21:17] on your side? cuz you you got to record [21:19] it, watch it over and click send, right? [21:22] So, [21:23] >> yeah, I watched that video a couple [21:24] times actually because I was like, am I [21:26] sure? But like sometimes I'm just like I [21:28] think when I made that video, if I could [21:30] like somehow find a way to harness that [21:32] energy, I'd have so many viral videos by [21:34] now, but I really can't. Like I'll just [21:35] like every like few months I'll just get [21:37] like mad about something and just like [21:38] make a rant video, but I really wasn't [21:42] planning it. Like that video like when I [21:43] when I look back at it, it was just a [21:45] really long like ramble where I went off [21:47] on like five different tangents to make [21:49] my point. I'm like, you know, if I was [21:51] trying to make a viral video, this could [21:52] be a little bit like better. Open AAI, [21:55] Enthropic, Cursor, and Verscell all use [21:59] this product to make their lives better. [22:01] And the problem it solves is when you're [22:03] building SAS or an AI product and you [22:06] want to sell to other companies, there's [22:07] all these requirements you need to meet. [22:09] There's SSO, there's skim, there's [22:12] arbback, there's audit logs. These are [22:14] all things that take time to integrate [22:16] but aren't the main focus of your app. [22:18] Work OS is an API layer that lets you [22:21] meet all of these requirements in just a [22:23] few lines of code. So, let's say you [22:25] have a new SAS product and you want to [22:27] sell to other companies. Work OS will [22:29] solve all of these critical feature gaps [22:31] for you. You can check them out at [22:33] workos.com to learn more and get [22:36] started. and I appreciate them for [22:38] supporting my work and sponsoring this [22:40] podcast. [22:40] >> In your career path, I mean, you you [22:44] went to Amazon, you you instantly quit [22:46] and you started making these videos, [22:48] right? You you keep creating these [22:50] videos and later you get your, you know, [22:52] your quote unquote dream job. Yeah. So, [22:54] when I was making the videos, I was [22:56] hoping that somebody would like get me a [22:59] job interview at some point and I'd like [23:01] ace it. It took a long time. time. It [23:03] took like almost a year and then it just [23:05] happened to be Google and I was like, [23:07] you know, I kind of would have actually [23:08] preferred somebody else so that the [23:09] interview would be a little bit easier. [23:11] But I was like, okay, well, like I guess [23:13] like here goes. And the interview [23:15] itself, I wasn't very nervous for for [23:17] whatever reason. I guess I just prepared [23:18] so much. But um and then getting into [23:20] Google pretty much, like you said, it [23:22] was like my dream job. I think a lot of [23:24] people like young college students, they [23:26] see like a tech company like that and [23:28] like once you're in, like you feel like, [23:30] you know, you've made it. And that's [23:31] definitely how I felt. And I was making [23:33] like pretty good money. I was up in uh [23:35] here in Seattle and um like everything [23:38] was going well. Like my team was really [23:40] nice. It was kind of the opposite of [23:41] Amazon where like everybody is like [23:43] super friendly and like willing to help [23:45] and the work was I would say fine. Like [23:49] it was technical at least but it wasn't [23:50] the most interesting. It was a lot of [23:52] like processes and bureaucracies. So [23:55] that was kind of boring. Um but um then [23:59] I got to a point where I felt like I [24:02] could actually like go full-time into [24:03] like the YouTube neat code thing and it [24:07] um that was pretty riskaverse. Like I [24:09] probably could have done it actually a [24:10] lot sooner and um yeah I really enjoyed [24:14] my time at Google but it was just [24:16] something that I I felt like I enjoyed [24:18] like algorithms and stuff more and it [24:19] was an opportunity that wasn't going to [24:20] like be there forever so I just wanted [24:22] to take it. But I could have definitely [24:24] seen myself working in big tech my whole [24:26] life. A lot of people I talk to, they [24:28] don't feel that way. A lot of people are [24:30] like, I'm just here like until I don't [24:32] have to be here anymore for how many [24:34] ever years that is. Um, but for me, I [24:36] was pretty happy just like doing [24:38] something technical my whole life. You [24:40] mentioned that some, you know, some like [24:42] Google's interviews maybe were certain [24:45] level of hardness and le code. If you [24:47] were to come up with a tier list of like [24:49] the just purely le code, so purely uh [24:53] which companies ask the hardest leak [24:55] code questions and if you pass that bar, [24:58] you must be pretty great at le code. [25:00] What's your tier list? I think the labs [25:03] ask pretty like tough questions these [25:05] days like OpenAI and anthropic, but [25:07] they're not like exactly like the leak [25:09] codes traditional leak code format. [25:11] They're a lot different. But yeah, I [25:14] think like the it usually just comes [25:15] down to like the supply and demand, like [25:17] how many people are trying to get into [25:18] that company and like how like how many [25:20] roles does that company actually try to [25:22] fill? I I feel like the smaller [25:23] companies are usually harder. [25:26] >> You said OpenAI Enthropic's interview [25:28] process a little different. How does [25:30] OpenAI and Enthropic do their coding [25:32] interviews? [25:33] >> Yeah, I've heard a lot of mixed things. [25:35] Like recently what I've heard is that [25:38] definitely like knowledge of data [25:40] structures and algorithms will be pretty [25:42] helpful. Like it's pretty hard to get [25:45] through it without that. But it's it's [25:47] definitely not like a leak code [25:49] interview is what they would say. It's [25:50] that like there might be components of [25:52] like an algorithm or like multiple files [25:55] or like you're just given like a problem [25:56] and you kind of have to like reason [25:58] through it. Some of them mention that [26:00] like they have AI access but I I think [26:02] for anthropic it's not. And I think the [26:04] most recent person I asked for OpenAI [26:06] also said that they don't have AI [26:08] access, which is really interesting to [26:09] me. I think like interviews are just [26:11] kind of really random lately. I'm trying [26:13] to figure out like where this is going, [26:15] but it's really impossible. Like I [26:17] thought it was going to be AI assisted, [26:19] but like all these companies are still [26:21] like figuring it out as they go. So, [26:23] it's it's been interesting to watch. [26:25] >> You went to Google and you liked it, but [26:28] sounds like you you left because you [26:31] wanted to pursue full-time. like it [26:33] wasn't like you were running from [26:34] something. You were going towards Le [26:36] Code and content creation. I I remember [26:39] back in the day you were faceless. Were [26:43] you nervous at all to post your your [26:45] face? [26:46] >> Yeah. No, it's it's kind of funny, I [26:47] guess, cuz it's been six years now since [26:49] I like made the YouTube channel. And for [26:51] the first three years, I think I was [26:52] faceless. And like roughly three years [26:54] ago, this month I think is when I did [26:56] it. And so now I've had like half and [26:58] half. And I think um I think it's been [27:01] great because I've gotten to like [27:02] actually be like a person. And after I [27:05] did that, I had a lot of people like [27:06] reach out to me. They're like, "Hey, [27:07] like I didn't know like you're the guy [27:09] who did that." Like, "Hey, like let me [27:10] know if you're ever in town, blah blah [27:12] blah." And um I guess it was cool to [27:14] like actually like meet people and I [27:17] don't know, just meet people I guess [27:18] that you can like relate to like content [27:20] creators like we met as well. Yeah. I [27:22] mean, so there's a part of me, I guess, [27:23] that kind of liked the faceless version [27:25] as well, cuz you just get to be like [27:26] really invert introverted and just like [27:28] not have to like ever deal with people, [27:31] but people are pretty nice to me and I I [27:33] don't have any regrets about that. Once [27:34] in a while, like I'll be out and [27:36] somebody will like recognize me and if [27:37] I'm with a friend, it's like you can [27:39] kind of like show off or act like you're [27:41] a big deal. But for the most part, I [27:43] don't know. I just Yeah, it's been fun [27:45] meeting like other nerds and stuff. [27:47] How's your experience quitting Google [27:50] and doing your own thing full-time? I [27:52] >> I think in my case, um I really enjoyed [27:55] what I was doing, so it was a lot easier [27:57] for me. I have a lot of people reach out [27:59] that like they don't really enjoy their [28:01] jobs, but they don't really know what [28:02] else they want to do either. I think [28:03] that's a lot harder. Um for me, it was [28:06] just that I liked algorithms and doing [28:09] things around that, whether it was like [28:11] building my site or making YouTube [28:12] videos. and I just kind of leaned into [28:14] that and that's been working out for the [28:17] past few years and more recently I've [28:19] been trying to like expand into other [28:21] areas and stuff. But I guess what I've [28:24] learned is that like I I guess passion [28:26] really matters and it's like if you're [28:28] passionate about one thing that doesn't [28:29] necessarily mean you'll be passionate [28:30] about everything. But um yeah, I think [28:33] it's I've definitely had no regrets. I [28:36] think it's been a success for the most [28:38] part. Um, but yeah, it's not it's not [28:42] like perfect. I wouldn't re necessarily [28:44] recommend like every person out there [28:46] needs to quit their job or do their own [28:47] side hustle or whatever. I don't like [28:49] buy into a lot of that stuff. [28:51] >> Do you ever miss Google [28:53] >> a lot? Like I miss like I I can afford [28:56] food now, but I still miss the free food [28:58] for some reason. I don't know if you had [28:59] that experience because they really like [29:01] make it like in a way that you'd miss it [29:03] if you weren't there. Like especially [29:05] the free coffee and stuff because I'd [29:06] always get like a latte. Um but [29:09] [laughter] [29:10] anyways, but um yeah, I miss um I mean [29:13] the one thing I miss is the culture [29:15] because I could like at my neat code and [29:18] like I could hire a bunch of people and [29:19] kind of create my own culture and it's [29:22] really nice that you have like the [29:24] freedom to do that, but I'm not really [29:25] like a culture builder. I'm more of like [29:28] a person that just does things or solves [29:30] problems or whatever. Like I don't [29:32] consider myself a manager, but I've had [29:33] to do that a lot the last couple years. [29:36] And it's it's been a learning [29:37] experience. [29:38] >> You mentioned what you miss about [29:40] Google, but what is it that you hated [29:42] most about working at Google? [29:44] >> That's a good question. I think it's [29:46] just really slow. Like urgency and like [29:50] speed doesn't seem to matter, which is [29:52] fine. I guess the reason that's a [29:53] problem because some people would see [29:54] that and think like, oh, it's like a [29:56] work life balance paradise. you can kind [29:57] of just chill and do whatever you want, [29:59] which is which is true because there [30:00] there's that side to it. But the other [30:02] side is that like if you're ambitious or [30:05] you're trying to like get your project [30:06] delivered sooner or you're trying to get [30:07] promoted or just anything, you're going [30:10] to be fighting against like all that [30:12] like work life balance culture because [30:14] you're trying to get somebody else to do [30:15] something and like you don't want to [30:17] like put too much pressure on them. You [30:18] don't want to like mess with their work [30:20] life balance cuz then they're going to [30:21] mess with yours and like yeah, it was [30:23] just really slow. If you're a technical [30:26] person, you you want to be there because [30:27] they're not going to probably at Google [30:28] like cut too many corners, but sometimes [30:31] you wish they would because it just goes [30:32] so slow and like like they prioritize [30:36] processes over actually like making a [30:38] like an impact for like a customer or [30:40] like getting a project delivered. It's [30:42] just like yeah, they care too much I [30:44] think about like what the engineers [30:45] think and like what they feel rather [30:46] than like actually trying to serve like [30:48] customers. So it's like the opposite of [30:50] Amazon I'd say in that respect. [30:52] What's the top technical book [30:54] recommendation for software engineers? [30:57] Do you have any thoughts on that? [30:59] >> Yeah, I guess I really liked um DDIA, [31:01] designing data intensive applications. I [31:03] I just spoke with somebody a month ago [31:05] and they were telling me that they they [31:08] did read through it, but for them like [31:10] most of it didn't really stick. And I [31:12] think that's fair actually. I think for [31:13] most people it wouldn't stick. But I [31:15] think my brain is just kind of like [31:16] wired a different way where like rather [31:18] than like I think most people learn by [31:19] like getting an introduction to like a [31:21] highle topic and then like digging [31:23] deeper into that. But for me I like [31:25] understanding like the actual like [31:26] building blocks of something and that'll [31:28] like help me understand like the higher [31:30] level stuff. So like for example like [31:32] why did like data warehouses use like a [31:34] completely different query engine than [31:36] like a actual transactional database? I [31:38] mean most people don't need to know and [31:40] it doesn't really matter most of the [31:41] time. So you probably won't remember [31:43] like even if you did learn it but like [31:45] there happens to be like the way the [31:47] data is actually assembled and for [31:49] certain query patterns blah blah blah [31:50] like these things do matter and like the [31:53] reason I say it's a good book uh based [31:55] on that isn't because like you'll [31:57] necessarily take those exact details and [31:59] then apply apply them to like some [32:01] problem but it like changes the way your [32:03] brain thinks about things like the same [32:05] way learning algorithms will help you [32:06] think about things differently [32:08] understanding like just general like [32:10] computer science problems and [32:11] distributed systems and all that stuff. [32:14] Uh for me at least like I notice a lot [32:16] of like parallels and stuff like when [32:18] I'm reading DDIA practically every [32:20] single chapter I can like find at least [32:22] like a few leak code problems that like [32:24] can like help explain like certain [32:26] things or like help me understand [32:27] certain things. And so in that way it's [32:30] been useful for me but I wouldn't [32:32] necessarily say for everybody it would [32:33] be. Last question is, if you could go [32:36] back to your younger self when you just [32:39] graduated college and give yourself some [32:41] advice, what would you say? Definitely [32:44] focus a little bit more on what you [32:46] enjoy because my priority at that point [32:49] was just trying to get like some job [32:51] that I would reasonably enjoy. Um, and I [32:54] thought like anything technical will be [32:55] fine. And then so like as I was doing [32:57] computer science, I started learning a [32:58] lot about like web development because I [33:00] saw people talking about that online and [33:02] learning React. But I think probably my [33:04] career would have been just as well if [33:06] not better had I done something like [33:08] really random and obscure like operating [33:10] systems or databases or something which [33:12] I would have enjoyed a lot more. And uh [33:16] I think I probably would have been [33:17] better at it than I was at web [33:19] development. Even though people say like [33:20] webdev or just like building apps is [33:22] like the easy thing. I feel like it's [33:24] not really easy if it's like boring to [33:25] you and to me it was pretty boring. So [33:28] um yeah people say the same thing about [33:30] leak code. like it's hard, but for me it [33:32] was I guess easy because I just enjoyed [33:34] it. So yeah, I'd say yeah, I think a lot [33:37] of people are asking that question now. [33:38] And so I would just say that like try to [33:40] enjoy it at least a little bit. [33:42] >> Awesome. Well, thanks so much for your [33:44] time, Nate. Really appreciate it. [33:45] >> Yeah, appreciate it. [33:46] >> Yeah. [33:46] >> Hey, thank you for watching this [33:47] podcast. If you liked it and you want to [33:49] see the show grow, please support with a [33:51] comment or a like. Also, if you have any [33:54] recommendations for people you want me [33:56] to bring on, please drop a comment. [33:59] guests like Barbara Liskoff, Mike [34:01] Stonerebreaker, Mark Brooker. These were [34:03] all people that I brought on because [34:05] someone left a comment. On another note, [34:07] aside from the podcast, I'm working on [34:09] building the ergonomic keyboard that I [34:11] wish existed. Here's a glance at the [34:13] prototype. It's a split keyboard, so [34:16] there's two sides. Um, this is in the [34:18] case, but yeah, we launched on [34:19] Kickstarter and we hit our goal within [34:21] eight hours of launching. I really [34:23] appreciate it if you were one of the [34:24] people who grabbed one of the early [34:26] units. Um, we're now working on the long [34:28] journey of building the tooling now. And [34:30] so if you still want to pick one up, [34:32] I've left the late pledges open on [34:34] Kickstarter, so you can grab one there. [34:36] I'll put a link in the description. [34:38] Thank you again for watching the podcast [34:40] and I'll see you in the next