Stanford's youngest instructor on InfoSec, AI, catching cheaters - Rachel Fernandez [Podcast #217]

[00:00] Welcome back to the freecodecamp [00:01] podcast. I'm Quincy Larson, teacher and [00:03] founder of freecodecamp.org and today [00:05] I'm interviewing Stanford's youngest [00:07] instructor, Rachel Fernandez, a [00:09] 19-year-old student who also teaches [00:12] their C++ course. First, some quick [00:14] community news. Freecodecamp just [00:15] published an automation for beginners [00:17] course. You'll learn how to automate [00:19] your daily routine tasks by piping [00:21] together triggers and actions and by the [00:24] end of this course, you'll have your own [00:26] model tech context protocol MCP server [00:29] that you can share info with between [00:32] your productivity apps and your agents. [00:34] That's a 4-hour course on the [00:35] freecodecamp YouTube channel. Link in [00:37] the description. Watch it after you [00:39] finish listening to this podcast. [00:41] Freecodecamp also published a [00:42] full-length handbook on data quality. [00:44] You'll learn the most common ways that [00:46] bad data enters the system and how to [00:48] prevent them. You'll get exposure to the [00:50] different layers where data validation [00:52] needs to happen, front end, back end, [00:55] database, business logic, data [00:57] ingestion. This handbook will also walk [00:59] you through testing strategies to keep [01:01] bad data out of your projects. [01:04] AI governance may sound like something [01:06] that only managers need to worry about, [01:07] but in practice it's us developers who [01:10] actually have to build the responsible [01:11] AI systems. You can bookmark this new [01:14] freecodecamp handbook and code along [01:16] with four hands-on Python projects, a [01:18] model card generator, a bias detection [01:20] pipeline, an audit trail logger, and a [01:22] human in the loop escalation system. [01:24] This is a full handbook. Link in the [01:26] description. [01:27] This week's song of the week is Dans le [01:29] noir by L'Impératrice by a French disco [01:31] band L'Impératrice. [01:34] Sorry, I'm not a French speaker. I hope [01:35] I didn't totally butcher the name. This [01:38] 2024 banger features a heavily [01:40] syncopated bassline that I think you'll [01:42] love and the singer subtly alternates [01:44] between French and Italian. The music [01:46] video is unique as well and has super [01:48] good vibes. Think you're going to love [01:50] it. Support for this podcast comes from [01:52] the 10,118 [01:54] kind folks who donate to our charity [01:55] each month. Join them and support our [01:57] mission at donate.freecodecamp.org. [02:02] You can also pick up a shirt like the [02:03] one I'm wearing. shop.freecodecamp.org. [02:06] $20 with free shipping anywhere in the [02:08] US. [02:09] Now for the interview. I'm hyped to talk [02:12] with her, Rachel Fernandez. She grew up [02:14] in Westminster, a small California town [02:17] with a largely Mexican and Vietnamese [02:18] population. 70% of the students at her [02:21] high school had family income so low [02:23] that they qualified for free school [02:25] lunches. And Rachel was the first [02:26] student from there to get into Stanford [02:29] in years. We're going to talk about the [02:31] state of computer science education in [02:33] 2026, Rachel's thoughts on C++, a [02:36] language that she teaches at Stanford, [02:38] and why she thinks it's still important, [02:40] and her tips for how devs should go [02:42] about using AI tools without deskilling [02:45] themselves. Rachel and Fernandez, [02:47] welcome to the free code camp podcast. [02:49] Thank you so much. I'm so excited to be [02:51] here, Quincy. [02:52] Yeah. Well, I'm excited to ask you a [02:55] whole bunch of questions. First, I want [02:56] to talk a little bit about C++. [02:58] You teach Stanford's C++ course. How [03:01] important is C++ to the software [03:03] ecosystem right now? [03:06] Well, I think it's it's incredibly still [03:08] important. I think there's a lot of push [03:09] right now to go to programming languages [03:12] like Python, right? These have become [03:13] very popular recently. Um a lot of the [03:16] the startups and companies that I talk [03:17] to are are writing in things like Go and [03:19] Python. Um I still think C++ is [03:21] incredibly fundamental to be honest [03:23] though. I mean, a lot a lot of really [03:26] really big things have been written in [03:27] C++. And so, I mean, in terms of uh [03:30] low-level capabilities, in terms of [03:32] systems, and and infrastructure, I think [03:34] C++ still has a major part in in the [03:36] industry. Um [03:38] uh I can think of things like uh I think [03:40] the next honestly big thing um outside [03:42] of like the AI boom is is robotics. And [03:44] I think uh particularly like C++, like [03:46] low-levels low-level stuff like this in [03:48] in systems things, kind of like [03:50] robotics, still need programming [03:52] languages like C++. [03:54] And there are still tons of issues and [03:56] bugs with it. Don't get me wrong, but [03:58] it's still incredibly important just [03:59] because so many of the things that our [04:01] industry is built upon, the foundations [04:03] of it is C++. Yeah, so like Windows for [04:06] example runs C++. Like Python and [04:09] JavaScript even if you're programming in [04:11] those, you're under the hood C++ powers [04:14] like their engines and stuff like that. [04:16] Yeah. [04:17] So right now there's a crisis in C++ [04:21] around memory safety, memory corruption, [04:23] timing errors, and other exploitable [04:25] issues within the language that have [04:28] just been there for, you know, 50 years [04:29] or so, 40 45 years, however long C++ has [04:32] been around. And you actually had the [04:34] C++ creator, the creator of the [04:36] language, visit Stanford and speak at [04:40] your class. And one of the things he [04:42] said in the past is he's imagining type [04:44] safety, resource safety, memory safety, [04:46] range safety, arithmetic safety, that [04:48] they could all be standardized. How [04:51] confident should people be that C++ will [04:53] continue to be a secure language worth [04:56] learning and building with? And I know [04:57] that you work at a startup that's [04:59] focused on security. So this seemed like [05:01] the perfect question to be asking you. [05:03] Yeah, absolutely. Um Bjarne was awesome [05:06] when he came to come visit us. Uh [05:08] the yes, yes. I think there are a lot of [05:11] security issues with C++. And honestly, [05:13] this is like one of my favorite parts of [05:15] the language is that you're kind of just [05:16] given like full reign of the programming [05:19] language and in terms of like memory [05:20] access and and stuff like that. Um but I [05:23] I think people should still be fairly [05:24] confident that the language itself is [05:26] going to be safe and it's going to be [05:28] continuedly used within within industry. [05:31] Um the fact that like Bjarne is still [05:33] talking and still making improvements to [05:34] C++. It makes me feel fairly confident [05:37] that I mean, I think he knows the [05:39] gravity of the situation of just like so [05:41] many things are built in C++. And I [05:44] there's been kind of a push recently to [05:46] like use other programming languages [05:47] like Rust because of I think they have [05:50] higher features and and more security um [05:52] implemented within it and that you're I [05:54] guess given less free rein in terms of [05:56] like memory access and things like this. [05:58] But I mean there have been pushes made [06:00] to make C++ more secure, things that [06:02] you'd mentioned in that in that snippet [06:04] that he had said, but things like [06:05] profilers and then modules and static [06:07] analysis, like all these things are [06:09] components that are being built up right [06:10] now for the programming language. And [06:12] the fact that there's a push for this [06:14] and the fact that Bjarne is so vocal [06:15] about it, I think it makes me feel [06:17] better and it makes me feel safe that [06:18] this this language is going to be [06:19] continuously used. Um and then the [06:21] context of security, I mean a lot of the [06:24] a lot of the code that my the the [06:25] companies like I work at, um they're [06:27] either doing pen testing or they're [06:28] looking at like LM generated code. Um [06:31] the security is not really looking at [06:33] the low-level stuff, but I think that [06:35] fact is like kind of scary upon itself. [06:38] Like all the languages that that are all [06:40] the code that, you know, we're looking [06:41] at in terms of like AI generated code, [06:43] right? It's all built on the [06:44] fundamentals of of [06:46] C++ code, right? And so if we are [06:49] looking at all this, you know, AI [06:50] generated code and it's built on [06:52] something else, let's make sure that [06:54] thing that it's actually built on is [06:55] secure. And I feel fairly confident that [06:58] a lot of people within the industry [06:59] recognize this and realize this and [07:01] there's been a greater push to focus on [07:02] security, um especially within the past [07:04] few years just because of all the AI [07:06] generated code. Um people are just [07:08] outputting a lot of code that they [07:09] honestly genuinely don't understand. And [07:11] so I think people are more cognizant of [07:13] the fact that that the code that they [07:14] have to write has to be secure. Um or [07:16] else a lot of big things like [07:17] infrastructure are going to go down. [07:19] Yeah. [07:20] And a lot of people like blame the uh [07:24] you know, AI assis- [07:25] assistants code generation. They'll [07:28] they'll they'll take like the code base, [07:29] but [07:30] would you say that it sounds like maybe [07:32] some people should actually look at the [07:33] underlying software and maybe the [07:35] software was always vulnerable in [07:37] certain ways that are just now being [07:38] discovered and you can't necessarily [07:40] just blame LLMs for introducing so much [07:43] insecurity. [07:44] Yes, absolutely. Absolutely. I think [07:46] wait, I mean there's there's a lot of [07:49] there's a lot of bugs within C++ code. [07:51] There's a lot of things that need to be [07:52] fixed and a lot of things that need to [07:53] be changed. Um I think if you're if [07:55] you're writing like web dev stuff, [07:56] there's not a need for you to go down to [07:58] the systems e-type of uh material, but [08:01] no, absolutely. I think um there's [08:03] there's a lot of issues with a lot of [08:04] code being written right now. Um but [08:06] systems code I think is always just [08:07] first off fun to look at. Um and yeah, [08:10] there there tons of bugs and memory [08:12] leaks and things like this um within the [08:14] foundations of a lot of code that is [08:15] being written. And so um [08:18] yeah, I I do think it's still important [08:20] for for developers to be looking through [08:22] that type of code. Yeah. Well, let's [08:24] talk about TreeHacks, this massive [08:27] hackathon at Stanford, which I was [08:29] there. I made a documentary about it. [08:30] You interviewed me. It was a lot of fun. [08:32] You interviewed Sam Altman from OpenAI. [08:35] Uh you got to put together this entire [08:38] event. Uh it had a million dollars in [08:40] prizes, a thousand [08:42] uh devs selected from 15,000 applicants. [08:46] Uh and [08:47] I mean it sounds like a massive [08:49] undertaking. Can you fill us in on the [08:51] history of the hackathon and how you got [08:53] involved? [08:54] Absolutely. So, this is this is the 12th [08:56] year of TreeHacks, which is really [08:58] exciting. I feel like it has grown so so [09:00] much um from where we first began just [09:03] not even like I guess a little bit more [09:05] than a decade ago. [09:06] Um it started off as as this thing that [09:09] I think uh these these three friends [09:11] decided, you know, we should have a [09:12] hackathon at Stanford. Like let's just [09:13] build one. Um and the founders come back [09:16] every year and they talk about how much [09:17] the event has changed. Uh but yeah, I [09:20] think I think hackathons have become an [09:21] an even bigger thing just as of [09:23] recently, especially with AI boom. Um [09:25] but yeah, this is our 12th year of [09:26] TreeHacks, and I think it was the best [09:28] one that we've had so far. Um [09:30] and in how I personally got involved in [09:32] it, I uh I actually I did a lot of [09:34] logistics stuff in high school, so I was [09:36] the logistics director this year for [09:37] TreeHacks, and I organized the event [09:39] last year, too. Um but I took more of a [09:41] leadership role this year. Uh so, when I [09:43] first came to Stanford, I I heard about [09:46] this club, TreeHacks, and it's quite [09:47] popular here. A lot of people want to [09:49] help organize the event just because of [09:51] the grand scale of it. Um and the [09:53] community is is really great as well. [09:55] And [09:56] I I was I was around club fair, and I I [09:58] saw this booth, and I just like I I was [10:01] too scared to go up there. There were [10:02] too many people asking questions. I was [10:04] I was being with my math problem sets, [10:06] and and someone had told me about the [10:07] logistics behind the event. Um and so, I [10:10] I went to a very I went to a very poor [10:11] high school in that I remember for our [10:13] robotics team, like we would we would [10:15] struggle to raise like even $100 like [10:17] doing a bake sale outside like an [10:18] Albertsons or a Vons. And so, I I'd [10:21] heard about the grand scale of the [10:22] event, right? It's like a big budget. [10:24] And I think that really drew me to [10:25] TreeHacks, the fact that like [10:27] any idea that I wanted to have, any [10:30] thing that I wanted to see implemented [10:31] at TreeHacks, and any big scale impact [10:33] that I wanted to have could be done with [10:35] TreeHacks and the and just the scale of [10:36] the budget and the size of the event, [10:38] right? [10:38] >> budget is phenomenal. Like you said, [10:40] okay, for people outside the US, [10:42] Albertsons and Vons are like grocery [10:43] stores. So, you would like bake a bunch [10:45] of goods, like cookies and stuff, then [10:47] you go, you set up a table in front of [10:49] this grocery store, and you try to get [10:50] people to buy your cookies, and use that [10:52] to raise money to buy like Arduinos or [10:55] other robotic stuff, right? Like Yeah. [10:57] And to go from doing that to Stanford's [11:00] budget, which I don't know if we can [11:01] publicly reveal the budget for [11:03] TreeHacks. Uh I know what it is. It's [11:05] massive. [11:06] Uh [11:07] >> Yes. I like [11:08] I mean, you talk about doing logistics. [11:10] Maybe you can compare doing logistics in [11:12] high school for robotics club to doing [11:14] the logistics of TreeHacks. Thousand [11:17] people, like tons of security, tons of [11:21] uh you know, just like the buildings, [11:23] like renting the actual [11:25] auditorium where you had to interview [11:27] Sam Altman. Like that was not free. Like [11:29] there there was a lot involved in this [11:31] process, and you had to manage a lot of [11:33] money and a lot of people along the way. [11:36] Maybe you can just take a moment to [11:37] compare and contrast your high school [11:39] experience versus Stanford. [11:41] I think in in high school a big part of [11:43] it is you have to sell people on what [11:45] whatever you're doing, right? For for [11:47] the robotics team that I worked on in [11:48] high school, it was like you you want to [11:50] support us, right? Like we're like these [11:52] low-income kids that are like building [11:53] robots and I think that's great and we [11:55] have so much passion and energy and so [11:58] much of the time that we had spent [11:59] fundraising was convincing people you [12:01] want to invest in us. And I think that's [12:03] really hard. Like why would companies [12:05] that are struggling, right? That I mean [12:07] it's it's been hard I think financially [12:08] for a lot of people as of recently. Um [12:10] why why would you want to spend your [12:12] money, right? That your company makes to [12:14] come sponsor this tiny robotics team [12:16] that that hasn't won any big awards or [12:19] anything like that, right? Um and then I [12:20] came to Stanford and it was just like [12:22] let me give you money. Like let me [12:24] sponsor this event. I mean we have so [12:25] many really big sponsors um part of the [12:28] event and it's wild like I I [12:31] came into this position and it was it [12:33] was suddenly so much more freeing. Like [12:35] it was we have this huge budget. [12:37] Literally anything you can dream of you [12:38] can make it happen at TreeHacks. And so [12:40] I remember last year I just I threw out [12:42] the idea of uh can we bring llamas to [12:44] TreeHacks? It was a complete joke. I I [12:46] like didn't think anything of it. I was [12:48] like oh it's going to be expensive, [12:49] right? And the the guy that came with [12:51] the llamas was from I think Central [12:53] California so he'd have to drive all [12:54] this way up. And someone was just like [12:56] yeah, we can make it happen. Like let's [12:58] do it. Um and so I think llamas has kind [13:01] of become a tradition at at other [13:02] hackathons as well and we had llamas [13:03] this year at TreeHacks. Um [13:05] and [13:06] yeah, I mean [13:08] running this this huge budget and this [13:10] huge event has been I think one of the [13:11] greatest things I've ever done in my [13:13] lifetime and in the TreeHacks is [13:14] honestly one of my favorite weekends I [13:16] think of my whole life. Um [13:18] it's incredible to be able to build this [13:20] event that make that makes other people [13:22] really happy. Um but yeah, it's it's a [13:24] lot lot work in terms of logistics. Just [13:27] I don't think I [13:29] This is like the first event I've ever [13:30] planned at this scale and I think [13:32] there's just so many tiny things that go [13:34] behind making an event like this that [13:35] like you just like you never recognize [13:37] until you're actually the one doing all [13:38] the planning. [13:40] Um but I feel very lucky in that I think [13:41] a lot of hackers really enjoyed the [13:43] event and a lot of them actually noticed [13:44] a lot of those tiny things. And so I [13:46] think that means a lot to me in terms of [13:48] of planning this event and and all of [13:49] that was totally and completely worth it [13:51] and I think it was one of the best [13:52] things that I've done here at Stanford. [13:54] Yeah. Well, I had a close encounter with [13:57] one of the llamas and I guess I have to [13:59] thank you to thank for that. That's [14:00] going to be in the documentary. Took a [14:01] big ball like fistful of feed. It just [14:05] went crazy for it. They are wild [14:07] animals. Let's put it that way. And I [14:09] got to brush one too. [14:10] Uh majestic animals really. [14:13] So um [14:14] like [14:15] So high school, it sounds like you were [14:17] just very resource starved and [14:20] I want to get into kind of like your [14:22] your background because it is novel like [14:24] it's it's not common. [14:26] Uh you were the first person from your [14:28] high school to go to Stanford in years. [14:31] And I think you said like out of your [14:32] graduating class you had one other [14:33] person went to like a prestigious school [14:35] that would be like Yale. [14:36] Uh but but like that's not where most of [14:40] the people at your high school were [14:41] aiming. They [14:43] uh they were I guess less ambitious than [14:45] you were to an extent. Like like can you [14:48] talk about um what it was like at your [14:51] high school in [14:53] um [14:53] Westminster, California which is in [14:56] Southern California, population of [14:57] 90,000 people. Kind of a small uh town. [15:01] Is Is it on I don't know a lot about it. [15:02] Is it on the coast or is it like a [15:03] valley type town? Where is it exactly? [15:05] Um if you know around like the beachy [15:07] areas of Southern California kind of [15:08] like Newport Beach, Huntington Beach, [15:10] it's about like 15-20 minutes away from [15:11] there. Okay. [15:12] >> Um yeah. Yeah. [15:14] Uh yeah, Westminster I I I owe a lot to [15:17] I think where I grew up and and that [15:19] being Westminster. [15:20] Um it's a very immigrant heavy town so [15:23] there's a lot of like uh Mexican and [15:25] Vietnamese immigrants there. Um, [15:27] my high school in particular was [15:30] it was it was full of a lot of students [15:32] that I think came from backgrounds of [15:34] poverty and a lot of families that [15:37] couldn't really afford to be living in [15:39] Southern California to be honest. Um, [15:42] and I think for me going to a high [15:43] school like that really transformed the [15:45] way that I view education that I value [15:47] it so much more. Um, I think when I when [15:49] I entered high school it was I mean a [15:51] big part of it was I grew up with people [15:53] like this and I saw that if you took the [15:56] opportunity in terms of education to [15:58] really really push yourself that you [16:00] could escape this level of poverty that [16:01] I saw so many of my classmates um, grow [16:04] up in. And so I think for me it was it [16:07] was all the way throughout high school. [16:08] I would always push my classmates to be [16:10] like, if you could just try a little bit [16:11] harder in this class, right? Maybe your [16:13] GPA could be a bit higher. Maybe you [16:14] could get into a better school. Um, but [16:17] a lot of the students at my high school [16:18] were just not motivated to be honest. I [16:20] think there was a [16:21] small group of students that were [16:23] interested within STEM. That was maybe [16:25] like a hundred of us and so because my [16:27] high school was was low income our [16:29] district had allocated us more money in [16:30] terms of like a STEM program. Um, [16:33] and [16:34] that was a great opportunity. I think it [16:35] ended up leading me to come here to [16:37] Stanford and and really taught me that I [16:38] like to computer science, but so many of [16:41] my classmates just had absolutely no [16:44] motivation to to do well within school [16:46] and I I think it's a it's a cultural [16:47] thing and that that's just something [16:49] that they weren't pressured to do as [16:50] kids, but it really frustrated me that [16:53] like [16:54] there were so many opportunities for [16:56] them to grab on to that even if they [16:58] took one thing, maybe they could get [17:00] into a better college, right? Like [17:01] secure a better future. But I think more [17:03] so than that, I saw that a lot of my [17:05] classmates just had no passion for [17:07] anything in particular. Like a lot of [17:09] what they did just [17:11] it was just kind of hanging out after [17:12] school and like I think there's nothing [17:14] wrong with that, but [17:16] it just really killed me that there were [17:17] a lot of my my classmates that could [17:19] have done fantastic and incredible [17:21] things and were truly, really talented [17:23] and smart and just had no internal [17:25] motivation to do anything within school [17:27] and find anything that they were like [17:29] particularly passionate about. And I [17:30] think that shaped a lot of [17:32] my experiences in terms of [17:34] being very grateful for any opportunity [17:35] that was that I was given, especially at [17:37] Stanford. I feel like I've tried to take [17:39] on anything that was given to me just [17:41] because I don't know, I kind of FOMO in [17:43] terms of I don't want to [17:45] I don't want to fall not fall, but like [17:48] I want to keep that internal motivation [17:49] going on within me. Yeah. Yeah. So, [17:52] FOMO, fear of missing out. Um Yeah. You [17:54] don't want [17:56] to [17:57] be caught slipping, so to speak, where [17:58] where you're like resting on your [17:59] laurels. Like, "Oh, okay, I got into [18:01] Stanford. I ran this hackathon or I [18:05] accomplished this or that." [18:06] >> Yeah. Uh [18:07] that's cool. But, I guess one question I [18:10] have is like there there's like the the [18:13] parable of Icarus flies too close to the [18:15] sun and his wings melt and he falls into [18:17] the sea. And then his his dad, Daedalus, [18:20] "Oh, I should have like helped rein in [18:21] some of his ambition, I guess." Like [18:24] you are in like the most ambitious [18:27] perhaps part of the entire world, all [18:29] right? Like [18:30] Palo Alto, San Francisco, uh Bay Area, [18:33] right? Like you have like [18:35] ambition is specifically something that [18:37] is selected for. [18:39] Uh if you read the books of like the the [18:41] various like [18:42] you know, I guess air quotes luminaries. [18:44] Again, I don't want to sound overly [18:45] critical, but but I mean the people out [18:47] there that are, you know, taking all the [18:49] money and giving it to the kids coming [18:51] out of Stanford and getting them to [18:54] you know, ram into the wall as hard as [18:55] they can with their idea to see who can [18:58] break through, right? Yeah. Um [19:01] Like, how do you kind of [19:02] strike a balance between having [19:05] that level of ambition that has gotten [19:07] you where you are today without [19:09] maybe going over where overboard and [19:11] like burning yourself out cuz I know [19:13] that that happens. Had lots of friends [19:15] who've kind of like cracked so to speak [19:18] and just [19:20] you know like gone to farm goats. [19:22] All right, as they say about developers [19:24] when they when they have [19:25] you know edited a few too many lines of [19:28] Java they go farm goats, right? Or maybe [19:31] or maybe llamas. [19:32] How do you strike that balance yourself? [19:35] Absolutely. I think I think in high [19:37] school I I definitely was very burnt [19:39] out. I think there was this expectation [19:41] that I put with on myself that like [19:43] all of this has to be worth for [19:45] something great. And so I just kept [19:46] pushing myself and pushing myself and [19:47] pushing myself until I came to Stanford. [19:49] And I think when I when I received my [19:51] acceptance letter at Stanford, I find [19:53] that was like my first sigh of relief [19:55] that like all of this was worth it, [19:56] right? Like I pushed myself to this [19:58] point and I can finally breathe, right? [20:01] And I think that I haven't felt that [20:03] level of burnout that I did from high [20:04] school to be honest and for quite a [20:06] while. I I don't think I've even [20:07] experienced it here at Stanford. And I [20:09] think if you talk to a lot of Stanford [20:11] students, to be honest, I think a lot of [20:12] them will tell you that Stanford's in [20:14] some ways feels easier than high school [20:17] and I think it's also that like we all [20:18] just went through so much in high school [20:20] that like obviously we look back and and [20:21] we don't think it was as bad as it [20:23] actually was. [20:24] Um but I feel like the classes here are [20:26] very manageable. But there's definitely [20:27] an expectation at Stanford in [20:29] particular. I feel like compared to [20:31] other big schools like MIT or Caltech is [20:33] that you have to be doing everything. I [20:34] think if you go to talk to other people [20:37] at more like researchy heavy schools, a [20:38] lot of what they're just doing is [20:40] classes and research. And that's great [20:42] and that's like awesome for them. Um [20:45] but at Stanford there's definitely a [20:46] culture of you you have to be like [20:48] founding something. Like you have to be [20:50] talking to venture capitalists. Like you [20:52] you have to be also like having a 4.0 [20:54] GPA in all your CS classes, right? Um [20:56] you also have to be doing research. And [20:58] I think that that expectation that you [21:00] have to be doing all these things is [21:01] what drives a lot of people to burn out [21:03] especially like especially at a school [21:05] like Stanford. [21:06] Um [21:07] And so I think for me in particular [21:10] the burnout [21:11] is kind of fun in that like I really [21:13] really just enjoy making myself busy and [21:16] pouring myself into things and people [21:17] that I love. And so I I don't think I've [21:20] experienced that burnout just because of [21:21] the community of the people here. Like I [21:24] think everyone here is working very hard [21:26] towards something that they're [21:27] passionate and really ambitious about [21:29] and I think that's very inspiring. And [21:31] whenever I feel sense of I'm getting [21:33] overwhelmed, I can fully and completely [21:35] just talk to other people about this, [21:36] right? But I think [21:38] like pouring myself into things that I [21:40] that I truly love has helped prevent the [21:42] burnout in that I feel like I'm not [21:44] doing things for the sake of like [21:45] getting into college like I kind of was [21:46] in high school. It's more just like this [21:49] is like for example TreeHacks like [21:51] that was I think one of the busiest and [21:53] most stressful times of my entire life. [21:55] For the past like 8 months we've been [21:56] planning this hackathon and at least [21:58] like the 2 months leading up to it I [22:00] feel like I was working on it at least [22:01] like 7 8 hours a day. And the week the [22:03] week [22:03] >> That's a lot. On top of all your [22:04] studies. Like like you had this weekend [22:07] and it's not like you're you got like [22:08] some pass to not have to study for exams [22:10] or something, right? [22:11] >> Yes. I'm still It was on top of your [22:13] entire academic and of course you [22:15] mentioned you're working at it like a [22:16] security company out there. Yeah. Um I'm [22:19] also part of like the security teams [22:20] here and and I I lecture on the side as [22:22] well. It's a lot of work, right? Um but [22:24] I I just I like love I love doing these [22:27] things. Like if I think if you talk to [22:29] any of the organizers from this year I [22:30] think if any if you ask any of them like [22:32] you saw Rachel being very stressed out [22:34] and very like [22:35] she was she was just kind of losing it [22:37] at some points. I think all of them [22:38] would say like it's just because she's [22:39] so passionate about this. Like there's I [22:41] would never ever take back the time that [22:43] I took doing TreeHacks and and all the [22:45] stress that I went through because [22:47] number one I got great stories out of it [22:48] and I feel like I grew a lot as a [22:50] person, but I just genuinely adored it [22:52] and I when the hackers were coming into [22:55] the the building that we hosted [22:56] TreeHacks in and just getting to meet [22:58] all these people that I feel like I've [23:00] worked so hard for. Like I I always knew [23:02] that a thousand people are going to come [23:03] to this event and then finally actually [23:05] like meeting them and being like I put [23:07] all this work and effort into this event [23:09] for someone like you who's also worked [23:12] really hard to get here. It's the most [23:13] fulfilling and heartwarming thing ever. [23:15] And and even when I teach at Stanford, [23:16] it's also a lot of work, right? Like I [23:18] lecture my own class. [23:19] Um I handle a team of graders. I make my [23:21] own assignments and and lectures and [23:23] that takes up a lot of time and effort, [23:25] but I just adore teaching. I love that. [23:27] And so I think for people that are [23:28] struggling with burnout especially at [23:30] school, um it's very possible for you to [23:32] pour yourself into things and still get [23:34] burnt out. I think I felt a little bit [23:35] about that after TreeHacks, but finding [23:38] things that you're genuinely passionate [23:39] about and then finding community within [23:41] that, I think is so important to [23:43] maintaining kind of like your [23:45] self-composure and ensuring that when [23:47] the stress gets there, you have people [23:48] to talk to. And the stress comes from I [23:50] guess kind of a good spot because you [23:52] you genuinely just love doing the things [23:54] that you're choosing to do. [23:56] Yeah. Well, I have a question about that [23:59] because [24:00] you know, you mentioned that this, you [24:02] know, high school run-up to college was [24:04] an extremely burnout [24:06] inducing experience for you and that [24:09] once you got to university it's actually [24:10] a lot easier. [24:12] Yeah. [24:12] >> like you could kind of take a breath and [24:14] it wasn't quite as uh crunched. Yeah. [24:17] >> But um and I've heard that from a lot of [24:19] people in a lot of different cultures. [24:21] Like people from India, people from [24:22] China, uh people from Europe. The the [24:25] actual process of getting into [24:26] university is often way harder than [24:28] actually going to university and [24:29] succeeding in an academic setting like [24:31] that. Uh [24:33] Stanford has a very low acceptance rate [24:35] and it's very difficult to get in. Uh [24:38] can you like there are probably some [24:39] parents listening to this, there might [24:41] be some high schoolers listening to this [24:42] that would love to go to a school like [24:44] Stanford that opens so many doors uh [24:46] industry and and offers like [24:49] awesome experiences like TreeHacks to [24:51] people who attend. Uh [24:53] what can you walk us through like that [24:55] process of getting into Stanford and uh [24:58] like all the different things that you [24:59] had to do along the way and like [25:01] theories of why they accepted you as [25:04] opposed to other people like was it [25:06] based on your robotics club work things [25:07] like that like anything you can give us [25:10] that would be useful to people who are [25:12] thinking about either having like [25:13] attending Yeah. you know a selective [25:16] institution like Stanford or [25:18] um and and you mentioned some aspects [25:20] where Stanford is even like [25:23] more hardcore or more um [25:26] than like MIT and Caltech and other like [25:29] hardcore engineering schools. [25:30] >> Yeah. Um [25:32] particularly because of like the [25:33] connection to Silicon Valley and like [25:35] the venture capital ecosystem and like [25:37] all these giant tech companies that are [25:39] based [25:40] uh you know within a few miles of where [25:42] Stanford is located. [25:43] Uh like people who don't already know [25:45] this like Google came out of Stanford. [25:47] Um you know like like pretty much all [25:49] the big tech companies [25:51] like Yahoo was like right in the area [25:53] like [25:54] I Sun Microsystems was right in the area [25:56] like a lot of the important companies [25:58] throughout kind of tech history have [26:00] been like in that general vicinity. [26:03] So [26:04] again to to stop giving you context and [26:06] just to to ask the question again [26:09] how do people get into a school like [26:10] Stanford and what was your experience? [26:13] Absolutely. I mean I I get asked this a [26:14] lot even from where I am from home just [26:17] because not a lot of people are exposed [26:19] to big schools like Stanford and and [26:21] it's not very common for someone to get [26:22] into a school like Stanford. [26:24] Um [26:27] There are a lot of different reasons as [26:28] to why I think I could have been [26:29] accepted. I think Stanford used to They [26:31] actually used to let you read your [26:32] application and see what your your [26:33] admission officer would would say about [26:35] your application. Um they unfortunately [26:37] removed that which I was I was very [26:38] devastated to hear about. [26:40] I think honestly a big big part of it is [26:42] passion and I think that's like such a [26:43] generic thing that people will say is [26:45] like oh you have to be passionate to [26:46] like get into college. I think it's true [26:48] and honestly I think it is a bit [26:49] unfortunate in that like [26:51] I think I was very lucky to find very [26:53] early on that I like to computer science [26:54] my freshman year of high school is when [26:56] I first discovered teaching and I first [26:57] discovered cybersecurity and that's when [26:59] I knew that I I loved CS and I wanted to [27:01] keep pursuing it. Um but I have a lot of [27:04] classmates that like didn't they didn't [27:05] know what they wanted to do until like [27:07] their senior year of high school. And [27:08] for some of them like they still don't [27:09] know what they want to do, right? [27:10] They're at community college and they're [27:11] trying a bunch of different things. And [27:13] I think that's fantastic. I think [27:14] everyone takes college at their own pace [27:16] and for some people this environment is [27:18] not what they need to grow as human [27:20] beings. Um but quite honestly I I do [27:23] think passion plays a very very large [27:24] part of it. I think uh when I was when I [27:27] was approaching this whole college [27:28] applications thing, I think a big part [27:30] of it was should I focus on one [27:31] particular thing throughout high school [27:33] or should I be like a generalist and [27:34] trying a bunch of different things? Um I [27:37] think for students in high school, they [27:39] should just try whatever they're [27:40] interested in. The way that I first [27:43] approached high school was I did not [27:44] want to go to Stanford. It was actually [27:46] I wanted to go to like a small Cal State [27:48] in California when I first came into [27:49] school because this is kind of what my [27:51] parents had impressed on me um when I [27:53] was younger, but when I came to high [27:55] school I knew I will never have the [27:56] opportunity again in my life to just try [27:58] a bunch of different things for free and [28:00] just see what I like and what I don't [28:02] like. And at the time I like didn't [28:03] really know what I wanted to do in terms [28:05] of like career or what I wanted to major [28:06] in in college. And so I tried a bunch of [28:08] different things. And that's how I ended [28:10] up finding about CS. Um [28:12] I think that's the right way to approach [28:13] high school. I think there's a lot of [28:15] societal pressure and a lot of external [28:17] pressure to get into schools like [28:18] Stanford, but honestly just try whatever [28:20] you like and see what you're passionate [28:21] about. And once you find what you're [28:23] really truly passionate about, just go [28:25] deep into it. And I think the question [28:27] of should I go into one niche specific [28:29] thing to get into college or should I um [28:32] you know, should I be more of a [28:33] generalist? I don't think that's really [28:35] the question that people should be [28:36] asking. I think it's more what are you [28:38] passionate about and what have you been [28:39] doing to work on this? Um I think a lot [28:41] of people make things for the sake of [28:43] doing things for college and I would [28:46] very much recommend not doing that. Um [28:48] for the first reason is that if you pour [28:50] your time into something in high school, [28:52] uh for example, a lot of people make [28:54] like non-profit education things in in [28:56] high school. Um and there's absolutely [28:58] nothing wrong with that if that's what [28:59] you're passionate about. But a lot of [29:01] people that I have talked to are just [29:02] making them for the sake of saying that [29:03] like they tutored X people in X amount [29:06] of things, right? If you are making [29:08] things for the sake of making things for [29:09] college, let's say you don't get into [29:11] your dream school. That time that you [29:12] would spent towards something that you [29:14] weren't truly passionate about is time [29:16] that in my opinion was kind of wasted, [29:18] right? And I think high school is such a [29:19] great opportunity to meet awesome people [29:21] and try new things. Um [29:23] and so I think you should honestly [29:25] genuinely find something that you're [29:26] passionate about and just dive fully and [29:28] deeply into it. Um [29:30] and when you're writing your college [29:31] application essays, I think I always [29:33] encourage people to just like write how [29:35] they speak. Um I'm someone that just [29:37] like speaks, I think, very like [29:39] passionately, and I feel like it's it's [29:40] very easy for me to talk about things [29:42] that I'm excited about. But I think a [29:43] lot of people get like scooped up into [29:45] the it has to sound formal, it has to [29:47] sound all these things. I think when you [29:49] are applying to big schools like [29:51] Stanford, especially like the Ivy [29:52] Leagues, they want like the human part [29:55] of They want like the raw part. Like it [29:56] doesn't have to sound like an interview [29:58] when you write your application. [29:59] Um [30:00] and so [30:01] you really want to just find what you're [30:03] passionate about, and you want to write [30:04] about it, and [30:06] you should write about it in a way that [30:07] like it doesn't sound fake when you talk [30:09] about it because you genuinely are this [30:10] excited about this. Um and if you if you [30:13] haven't found that passion, that's [30:14] totally okay. Just try different things [30:16] and see what you're interested in. Um [30:18] and I think for CS, this is something a [30:20] lot of people want to major in, right? [30:22] Um and so when I was approaching CS, [30:25] right? Everyone applies for CS at [30:27] Stanford. Stanford is I think it's like [30:29] 50% of the students here end up [30:31] switching or came in as CS. And Stanford [30:33] is great in that like you can switch [30:34] your major as many times as you want, [30:36] and they don't It's not impacted. I [30:38] think that's what other universities [30:39] call it. Um [30:41] so when I was writing my Stanford [30:43] application, think this is the biggest [30:44] tip I give to people is you want to find [30:46] what makes you so uniquely you. There [30:48] are just so many other applicants that [30:50] have the same stats as you, right? Like [30:52] they have a high GPA, and they did this [30:54] and this and this, and they did research [30:55] and all these things, and they have all [30:56] these big awards. But, what makes you [30:58] uniquely you? And I I think a lot of [31:00] people on their college applications [31:02] will write, "Oh, I want to be I want to [31:04] be at Stanford because I'm doing like [31:05] they have this program for me, or [31:07] there's this professor I can do research [31:08] with." Um [31:10] That's great, no. And I'm I'm very [31:12] youthful. It's like great that you know [31:13] what you want from the school. But, the [31:15] entire point of your college application [31:17] is you need to sell yourself to the [31:19] school. Why should this school accept [31:20] you over someone else? And this is the [31:22] entire mindset that I had when I was [31:24] writing my Stanford application. Um and [31:26] I think that I think that helped me a [31:27] lot. And so, I thought about what unique [31:30] characteristics do I bring as like a [31:31] human being to Stanford and and to the [31:33] culture and to the community. Um and I [31:35] would recommend writing like very niche [31:37] topics in in your in your application. [31:40] Try to be Try to talk about the most [31:42] unique stories that you have and and [31:44] what makes you so special. And [31:46] especially if you're particularly [31:48] passionate about something, which I I [31:49] think you should be if you're applying [31:50] to a school like Stanford, talk about [31:52] that. Um and I would say try to be as I [31:55] guess like emotionally raw as possible [31:56] without making it too crazy without I [31:59] guess like crashing out over um the any [32:01] application or anything like that. Um [32:03] yeah, I think you just I think [32:06] I admission officers just they want to [32:08] be sold on why they should pick you for [32:09] a school. And [32:11] I guess I can talk about for me in [32:12] particular, I don't know how helpful [32:13] this is. Uh [32:15] a big part of my growing up is that I'm [32:16] mixed. I'm Spanish, I'm Vietnamese, and [32:18] I'm Argentinian. And I'm a girl with a [32:20] computer science, and that's not [32:21] something that people often talk about. [32:22] But, there's other people like me, [32:24] right? And so, when I was approaching [32:26] CS, a big part of what I loved doing in [32:28] high school was teaching. And so, a big [32:30] part of my application was I thought, [32:32] "Okay, every other person that's [32:34] applying to Stanford for CS is going to [32:35] be doing research. They're probably [32:37] going to be interested in one specific [32:38] niche topic within CS. [32:40] Um and I was just thinking of the most [32:42] like stereotypical CS person that I [32:44] could think of, right? Um and I tried to [32:47] recognize I like got a piece of paper [32:48] and I wrote, "Okay, I'm unique in these [32:51] aspects." And I think a big part of me [32:52] is I just love bringing community [32:54] together and I love um [32:56] I love I guess a big part of my [32:58] application was just bringing humanity [32:59] back into computer science. And I think [33:01] that's a topic that not a lot of people [33:02] talk about. Um so in any way that you [33:04] can differentiate yourself from other [33:05] people is an opportunity that you should [33:07] take and you should write about it. And [33:09] um a lot of people that I knew in high [33:11] school did just one extracurricular [33:12] throughout high school. Um this is my [33:14] hot take. I don't think that's I I think [33:17] that's great if that's something you're [33:18] passionate about, but I think it's hard [33:20] to get into college if you do just one [33:21] thing throughout high school, whether [33:22] that's like band or robotics or things [33:24] like this. There were a lot of [33:26] brilliant, amazing people that I knew [33:27] through the robotics program that I did [33:29] in high school that just did robotics [33:31] all throughout their time um in high [33:33] school. And they did incredible things, [33:35] but it's just hard to put onto paper the [33:38] things that you've done for just one [33:39] club. And so I think it's better to try [33:42] a bunch of different things and have a [33:43] bunch of different experiences on your [33:45] application rather than pour your time [33:47] and efforts all into just one single [33:48] extracurricular. Unless you do believe [33:51] and fully feel like that is the one [33:52] thing that you truly love, then I think [33:54] the passion that you that you get out of [33:56] it and when you write about it will come [33:57] through. Um but this is just something [33:59] that I experienced and that like I saw [34:00] so many brilliant people who just poured [34:02] their time into like one extracurricular [34:04] who I think should have ended up at big [34:05] schools like Stanford, but when you [34:08] compare that application to a someone [34:10] else that has done so many other [34:11] different types of things and just got a [34:13] wide range of experiences, it feels more [34:15] tempting to pick the person with a wide [34:17] range of experiences in my opinion. Um [34:19] so yeah, in terms of tips for Stanford, [34:21] just be passionate, be excited, try a [34:23] bunch of different things. Um [34:25] and it it's it's hard to ignore the [34:27] societal pressure to go to big schools [34:28] like Stanford, but I I fully and [34:30] fundamentally believe that like everyone [34:32] who applies to college will end up where [34:34] they need to be. And if that's out of [34:35] school like Stanford, I think you will [34:37] end up there. And if not, that's totally [34:39] and completely okay, too. You'll find [34:40] your own sense of community somewhere. [34:42] Yeah. [34:42] >> Wow. I like I took so many notes while [34:44] you were talking because I I think that [34:46] So, if I had to recap some of the things [34:48] that you said there because I think that [34:49] they are important and I think they're [34:51] overlooked by people. I I mean, there's [34:53] like an entire industry that like [34:55] preparing people to get into [34:57] these programs, especially overseas, [34:59] trying to help people get like into a US [35:01] university. And [35:03] uh there's a lot of kind of conventional [35:05] wisdom that you kind of challenge there. [35:08] Uh but one thing is [35:10] even though specialization is rewarded, [35:12] obviously, in the workplace and and uh [35:14] things like that, like in in terms of [35:16] the application process, you do want to [35:19] have a lot of different things so you [35:21] can be able to not just list that you're [35:23] like you went really hard on robotics [35:25] for like 4 years, but that you tried [35:28] this and this and this. [35:29] >> Yeah. [35:30] >> having like a general theme in your [35:33] application, you said uh bringing [35:35] humanity back to computer science was [35:38] kind of the overarching theme. Yeah, [35:40] yeah. I think a big part of of college [35:43] essay writing I've read a lot of college [35:44] essays and just helping friends from [35:46] back home write write their essays. You [35:48] need to have a story. Like there needs [35:49] to be like someone um I actually this [35:51] this might be helpful. Uh when you when [35:53] I came to Stanford for my uh admit [35:55] weekend, since this is like when I was [35:57] first considering whether or not to take [35:58] Stanford, um they have this thing called [36:00] admit weekend and all of the people that [36:01] were admitted come out and basically [36:03] Stanford tries to sell you on on why you [36:05] should pick their school, right? Um you [36:07] meet with your admission officer in [36:09] outside and under the sunny California [36:11] weather. And my admission officer [36:13] basically told me that the process for [36:14] admissions is [36:15] um someone will read your application [36:17] and if they think that you're solid [36:18] enough, what they will do is I think [36:21] there's there's some sort of process for [36:22] like weeding some people out, but um [36:25] basically like the the end goal is that [36:26] your admission officer wants to be able [36:29] to make a presentation on you and sell [36:31] to other admission officers why you [36:33] should in particular be picked. And that [36:35] like that blew me away when I found out. [36:37] Like someone in the admissions room [36:40] literally made a presentation about me [36:41] and fought for me in front of a council [36:44] of other people that also needed [36:45] students to be admitted from their [36:46] sections, right? I felt so honored and [36:49] so seen when she had told me that, but [36:51] it's also, I think, a very terrifying [36:53] fact that had I known going to [36:54] admissions, I would have I would have [36:56] been so scared to write my application. [36:58] Um [36:58] >> Yeah. [36:59] So, to emphasize what you just said, [37:02] Yeah. [37:02] >> an internal champion essentially [37:04] advocating for you to their peers. You [37:06] need to sell somebody so hard they're [37:08] willing to put their own reputation on [37:10] the line advocating on your behalf. So, [37:12] this is kind of like enterprise sales, [37:14] which I do a little bit of at [37:16] freeCodeCamp. Like, but basically, [37:18] you're trying to find somebody, make a [37:19] connection with them, and convince them [37:21] that like no, like this is definitely [37:23] the right tool for the job. Like, you [37:25] should adopt my, you know, LM [37:28] as opposed to that other financial model [37:30] or whatever, right? Like, and and then [37:32] you should put your own reputation on [37:33] the line to go forth and [37:36] uh [37:36] you know, [37:37] this makes a lot of sense, and it really [37:39] contextualizes the uh job search. [37:42] Um So, one thing that I want to ask is [37:46] uh I just I want to also want to [37:48] recapitulate the fact that you need to [37:49] have a story. [37:51] Yes. [37:51] >> Um it's not just a collection of bullet [37:53] points. A bullet points do not [37:56] do it. They do not win the day, Yeah. [37:58] >> Um but at the same time, it like maybe a [38:00] narrative that like kind of walks over [38:02] the different facets of your high school [38:05] experience and and your aspirations and [38:06] goals. [38:07] Uh making appeals to like, "Oh, I love [38:10] this program so much because this [38:12] professor, I read their book and XYZ." [38:13] That's that's kind of like wasting time. [38:16] Uh wasting space in your application. Uh [38:19] Is is kind of the vibe that I'm getting [38:20] from you. That did you're not trying to [38:23] like kiss up to them and get them to [38:25] think that oh, this is this will be a [38:26] very [38:27] you know, like enthusiastic student. No, [38:29] they're looking for people that are [38:32] I guess um not just sycophantic in the [38:35] application but but rather like [38:37] confident that they would bring [38:38] something novel to the university and [38:40] not just show up and accept the [38:42] offerings of that university. [38:44] Absolutely, yeah. I think I would really [38:46] really avoid the most generic essays [38:48] that people write. Um [38:50] for me, I think there's a question that [38:51] Stanford asks about about curiosity. [38:53] Like what shapes the world in terms of [38:55] what you find curious and and how do you [38:57] develop that skill as a person? [38:59] Um and I I tried to think really hard, [39:02] okay, what would what would everyone [39:03] else write? And I think a lot of people [39:05] would write, oh, like I was part of like [39:07] this like computer science club and like [39:09] I really find out I was interested about [39:11] like this specific thing within CS and [39:13] like I like did research with this [39:14] professor. So many other people are [39:17] writing essays like that. Like what [39:18] makes you unique? Um and so actually I [39:21] think in my entire Stanford application, [39:23] I like barely talked about my academics. [39:25] I talked a lot more about just who I was [39:26] as a human and my personality. And I [39:29] love that the Stanford application I [39:30] think focuses in more on that. But the [39:32] way that I ended up answering that [39:33] question was about my I have a pet [39:36] parrot and she got sick my freshman year [39:37] of high school. And I went on this whole [39:39] ordeal of having to research and I found [39:41] out that a lot of doctors that I was [39:43] going to like they didn't even know what [39:44] they were talking about when it came to [39:45] the health of my bird. And so it kind of [39:47] shaped the world in that I I recognized [39:49] like the professionals didn't have all [39:50] the answers, right? And then I kind of [39:52] had to like do all my own research. And [39:54] I think I was I was so proud of that [39:56] essay because it was just so uniquely [39:58] me. You know what I mean? Like this is a [39:59] unique experience that who else is going [40:01] to talk about their sick bird in their [40:02] in their Stanford application, you know [40:03] what I mean? So it's really selling [40:06] yourself, having a narrative, and [40:07] thinking, okay, what would everyone else [40:09] write? And it's fine if you have to [40:10] write what everyone else writes. Like if [40:12] that's all you have, that's great. But [40:13] if there's something unique and [40:14] something that you can like tie it to [40:16] story to, I just love interesting [40:17] stories. I think you should absolutely [40:19] do that. Yeah, and it sounds like you [40:20] had the confidence to talk about [40:22] something that people might dismiss as [40:24] oh, that's such a trivial thing like [40:26] that like they're going to feel insulted [40:28] that I'm talking about my my parrot. [40:30] >> Yeah. In my college or or you know, or [40:33] but you had the confidence to like no, [40:34] this is relevant to who I am. And so you [40:37] were [40:37] >> Yeah. able to tie it into the story and [40:40] and that I think illustrates a certain [40:42] degree of confidence in one's own [40:44] thinking and reasoning. Um yeah, that's [40:47] really cool. So, [40:49] the process of getting into Stanford, I [40:51] mean you said that it was like a pretty [40:52] exhausting like that final sprint toward [40:55] the finish line in high school. Can you [40:57] talk about like what those final few uh [41:00] months in high school were like like [41:01] during this process cuz you said you got [41:03] really burned out and then it was like a [41:05] huge reprieve once, you know, uh [41:09] we're in and and you could like just [41:11] let that go to an extent and and [41:13] decompress a bit. [41:14] Yeah. [41:15] I think it's honestly a lot of a mental [41:17] game and that [41:19] whenever I was given an opportunity, I [41:21] knew that I had to pour myself fully and [41:23] completely into it and really try my [41:24] best to stand out. And I think it was it [41:27] was that I wanted other people to [41:28] perceive me as as, you know, this person [41:30] that works really hard. But I think also [41:32] internally it was anything that I was [41:34] given, I wanted to take full and [41:36] complete like I just wanted to like put [41:38] all my effort into it, right? Like this [41:40] is something that was given to me, I [41:41] should try my hardest with it. And I [41:43] think that mindset is honestly just [41:44] incredibly exhausting. I think there [41:46] were like at least like 10 different [41:48] things that I was part of in high school [41:49] and I genuinely tried my hardest to pour [41:51] myself into everything that I was [41:53] working on. And also the people that I [41:54] was working with, right? I think the [41:55] community that that that I was in in [41:57] high school was really important to me [41:59] and making sure that my peers felt [42:01] supported too was really important. [42:03] Um [42:04] It was it's just mentally taxing in my [42:07] opinion to be in high school and want to [42:08] be part of want to want to go to a [42:10] school like this. There's this constant [42:13] like notion that like [42:16] any moment that I'm free is a moment [42:19] that I should spend doing something. [42:21] This is a thought that I had for four [42:23] all four years throughout high school. [42:25] Um my weekends were spent doing robotics [42:28] or my weekends were spent doing research [42:29] or doing homework that I caught up on [42:30] because I was doing other [42:31] extracurriculars throughout the week. Um [42:34] and that's like so exhausting. Like I I [42:36] cannot think of like a single week where [42:37] I ever took a break off. And it it hit [42:39] the point my senior of high school where [42:41] like there was a there was a competition [42:42] I was at and I literally just like broke [42:44] down sobbing and I remember my dad was [42:47] trying to comfort me and being like, [42:48] "Oh, when you go to Stanford like you [42:49] will be with other people that have [42:50] worked just as equally as hard as you to [42:52] get to where you are. Um and you're [42:54] going to feel supported and you're going [42:55] to feel seen." But that was something [42:56] that I very much struggled with in high [42:57] school. Um [42:59] that that mental taxation of like you [43:01] always could be doing more. There's [43:02] always someone that's doing more than [43:03] you. And at least when I came to [43:06] Stanford that sigh of relief that I'm [43:07] referring to is I knew that I was going [43:09] to be okay. Like all this was worth for [43:11] something, right? I think in high school [43:13] it's very scary to think about all this [43:14] effort and time that you're putting into [43:15] things might not even be worth it. Like [43:17] you might not even end up at a school [43:18] that like I had friends that worked [43:20] really hard in high school and they [43:21] ended up at community college. And [43:22] there's nothing wrong with community [43:23] college. I think it's honestly great and [43:25] I think they've grown a lot as people, [43:26] but I know they were like severely [43:28] disappointed at the fact that like they [43:29] didn't end up at like at least like a [43:30] UC, right? And I feel really bad for [43:33] them. And I think that's incredibly [43:34] scary and I think that is what I mean [43:36] about the burnout and that there's this [43:38] this mindset like you always have to be [43:40] doing more because there's someone else [43:41] doing more than you. Um that's so scary [43:44] and I can confidently say that now I'm [43:45] at Stanford I have this like sense of [43:48] comfort that like [43:49] I'm going to be okay, right? Like I I [43:51] worked here and I'm at the school and [43:53] I'm going to be okay. And I worked [43:54] really hard to get to this point and I'm [43:56] really great that I I'm really grateful [43:57] that I have that sense of stability in [43:58] my life. Um but it's it's really hard in [44:01] high school to balance it all, to have a [44:02] work-life balance. And I honestly can [44:05] say I I feel like I didn't really have a [44:06] work-life balance. But I wouldn't have [44:08] changed anything. I feel like I grew a [44:09] lot as a person and the people that I [44:11] knew that were also going through this, [44:13] like in that way we had a sense of [44:14] community. Um but yeah, it's it it was [44:17] it was really hard. I was really burnt [44:18] out my senior of high school and I got [44:20] into Stanford early and I didn't I [44:21] didn't apply to any other schools [44:23] besides the UCs and I feel very lucky [44:24] that I got in early and I didn't have to [44:26] write like 20 more college applications [44:28] over my winter break. Um but yeah, I I [44:31] was I was very burnt out after the whole [44:33] process and if I had to if I had to do [44:36] it all over again, I don't know how I [44:37] would do it. I honestly think I worked [44:39] harder in high school than I have at [44:40] Stanford and it's just because of that [44:42] that scary mindset to be honest. [44:44] >> Yeah. [44:44] Yeah. [44:45] >> So uh you have to do more because [44:47] there's always someone doing more than [44:48] you. That's the mindset that is is [44:50] grinding you uh and and burning you out [44:52] essentially. [44:54] Um [44:54] >> Yeah. and I want to talk a little bit [44:55] like you don't know what you're doing [44:57] that is excessive. Like maybe I don't [44:59] need to do as much of this. Maybe I can [45:02] get away with doing less of this and you [45:04] don't know at the end of the day uh you [45:08] could you can probably presume that some [45:10] amount of your effort was squandered, [45:12] but you don't know which effort. So I [45:14] mean, you don't really have a way to [45:16] tease all that. That must be also a [45:17] frustrating aspect of it. Um just the [45:19] opacity of the process. Even with [45:21] Stanford being pretty transparent, I [45:23] guess historically they used to actually [45:24] show your application [45:26] uh and like the comments and stuff, but [45:28] um [45:29] I I imagine that could be quite uh [45:33] tilting for a lot of people that are [45:34] just trying to get into a good school [45:36] and you mentioned some of your [45:37] classmates that were trying really hard [45:38] to get into [45:39] uh you know, more selective schools and [45:41] then end up just having to go to [45:42] community college, which is a fine [45:43] outcome if you're not aiming high. If if [45:45] you just want to go to an inexpensive [45:47] school, California's community college [45:49] system is like the best in the world. [45:51] It's very reasonably priced, excellent [45:53] teaching. [45:54] Uh that is not a bad place to end up by [45:55] any means, but [45:57] >> Yeah. if you were aiming to get into [45:59] like [46:00] UC um you know, LA or or if you were [46:04] aiming to get into Cal or something like [46:06] that, then then that could be kind of [46:08] crushing that you're not able to achieve [46:10] that despite all of your effort. So, [46:13] when there is so much incentive to get [46:16] into these programs, people are going to [46:19] try to cut corners naturally because [46:21] it's really hard and you don't even [46:23] necessarily know if it's worth it [46:25] because you have a you see all these [46:26] other people that are not necessarily [46:27] getting into it despite having it [46:29] spending all the effort. And that can [46:31] lead to a lot of academic dishonesty, [46:34] people cheating, things like that. And [46:36] you have taught at Stanford in addition [46:38] to been a student. So, you've been [46:39] exposed to a lot of academic dishonesty. [46:43] Uh, and I wanted you to if you can like [46:45] maybe talk a little bit about that. Like [46:47] students cheating in high school, [46:49] students cheating at university, and [46:52] like just share any thoughts you have on [46:54] that. Like what is driving that? Like [46:56] what can be done about that? Like what [46:58] you would say to somebody who is [47:00] considering cutting corners, maybe using [47:03] LLMs to write their papers, [47:05] uh, maybe um, not [47:08] necessarily studying but just like [47:10] pulling up cheat sheets and stuff like [47:11] that. [47:12] Yeah. [47:13] Um, I think for for So, I mean AI the [47:16] whole AI boom happened to my senior year [47:19] around spring quarter. Like I remember [47:20] right before I came to Stanford was when [47:22] all this was happening. And it was it [47:23] was it was honestly scary. I I feel so [47:26] tremendously grateful that I did not [47:28] have AI in high school. I think it is so [47:30] tempting to have to use those not to [47:32] have to use those tools. But to have [47:33] them available to you, like it's [47:37] all every if everyone else is using it, [47:39] then sometimes it feels like, "Oh, it's [47:40] okay if I do this, too." I feel so lucky [47:42] that I thugged it out, that I had to [47:43] learn chemistry from the textbook, that [47:45] I had to write all these papers and stay [47:47] up late and and learn how to like [47:48] actually effectively communicate through [47:50] papers. [47:51] I feel very very lucky that I didn't [47:53] have any of this pressure or any of this [47:56] access to these tools. [47:58] Um [47:58] and [48:00] I In some ways I I do wish college was [48:01] still like that. [48:03] Uh [48:03] I have a younger brother who's in high [48:05] school, and he just tells me, I mean, [48:06] everything in my AP class like [48:08] translated to paper. Everyone's doing [48:10] everything on paper. If he has to write [48:11] an essay for a class, it's you have to [48:13] sit down, and you have to write it on [48:14] paper, and you can't you can't use like [48:16] LLMs or anything like that. And I think [48:17] that sucks, and I think that it's awful [48:19] that we have to start putting more rules [48:21] on students that generally wouldn't [48:23] cheat because everyone else is cheating. [48:25] Um [48:26] I think if you think about the mindset [48:28] of someone that is trying to get into a [48:30] big school, like um like the IVs or [48:32] Stanford, [48:34] GPA is just so it's like the bare [48:38] minimum now to get into a big school. [48:40] Like you have to have a perfect GPA to [48:42] even be considered for a school like [48:43] Stanford. And I think GPA inflation is [48:46] something that a lot of people have been [48:47] talking about recently, and that it's [48:48] just it's just so easy to get A+s now. [48:51] There's I mean, even at my high school, [48:53] there was such a push to give everyone [48:54] As and to give everyone high grades so [48:56] that the graduation rate could be [48:57] higher, right? So they wouldn't hold [48:58] people back. Even though very clearly [49:00] there are some people that should have [49:01] been held back. Um There There's some [49:04] people that like would get Fs at my high [49:06] school. Like they just would genuinely [49:07] put no effort. They would come over for [49:10] like a summer program. They would just [49:11] sit and watch videos. Like literally [49:13] they don't have to take quizzes or tests [49:15] on anything. Um and they would get [49:16] credit for that class. Just by like [49:18] staying at school for like 2 hours a day [49:19] like watching videos, which I think is [49:20] absurd. Um but I I I think that's a [49:23] different type of person. For the type [49:25] of person that's like hyper-competitive, [49:27] I think it's very easy to use these [49:28] tools just for the Okay, I just like [49:30] need to get this bare minimum GPA in [49:32] order to even qualify for some of these [49:34] schools on top of Okay, the time that I [49:36] would have spent doing school can be [49:38] time spent doing extracurriculars or [49:40] other things that can boost my [49:41] application. Um [49:44] That mindset is I think it's Yeah. [49:46] >> Yeah. So So just to be clear, [49:48] because there's so much pressure to also [49:50] do all these extracurriculars, but you [49:52] have to check the box of having like a [49:53] perfect GPA. A lot of people are looking [49:55] around and saying, well, if I can just, [49:57] you know, farm out the GPA part to LLMs, [50:01] Yes. [50:02] >> and do the extracurricular stuff, then [50:04] I'll have more time for [50:05] extracurriculars. So, they're they're [50:06] maybe being squeezed. Like, they they [50:09] don't feel they have time to do both. [50:11] Yes. Yes. Absolutely. I think that's the [50:13] perfect way to put it. Um that mindset [50:15] is is is quite scary. Uh even at [50:17] Stanford, I was a I was a TA my my [50:20] freshman year for the intro to CS [50:21] classes, and I mean, I saw so much [50:24] cheating. I think a lot of people always [50:25] have this mindset of like, "Oh, I'll [50:27] just like learn it later. Like, I just [50:28] like I like need to get this done." Um I [50:30] think particularly at Stanford because [50:32] of what I mentioned earlier, everyone is [50:33] trying to do everything right. People [50:35] are trying to make a startup and do [50:36] research and also do the classes on top [50:37] of clubs that they're doing. And it's [50:39] just it's just too much work for I think [50:41] one person. And so, this pressure, it's [50:43] the same thing of, "Oh, my classes are [50:45] like least important." I think [50:46] particularly at Stanford, people have [50:48] this sense of security of like, "Oh, [50:49] because I'm at this school, like I could [50:51] just cheat and like I would still get a [50:53] degree from here. And then like I'll [50:55] have a job and like I'll still be fine." [50:57] But I think it's it's honestly so sad [51:00] that people think that way. Um and [51:02] whenever I had students that would [51:03] cheat, I mean, I couldn't directly talk [51:05] to them, but it was the the the [51:07] narrative that we would tell these [51:08] students is "You're paying all this [51:10] money to go to this school, and what are [51:12] you getting out of it?" Like, the the AI [51:14] knows how to do these things, right? The [51:16] entire point of you being part of school [51:18] is that you can teach yourself these [51:20] things, right? I'm not teaching the I'm [51:21] not teaching the ChatGPT how to do [51:24] proofs or things like this, right? You [51:25] are the one that needs to learn how to [51:27] do proofs. And it's it's really [51:28] unfortunate, a lot of the I think the [51:29] test scores at Stanford have have [51:31] tanked, but GPA inflation is still very [51:34] high here, and it's it's unfortunate. [51:37] And I think it's really hard to motivate [51:38] people that have access to these tools [51:39] and freely and openly use these tools. [51:42] Um when there's just so much growth and [51:43] learning that can be done here. And for [51:45] me particularly, like I'm paying for my [51:46] own college, it makes me really [51:48] appreciate that like I'm paying X amount [51:50] of money to go to school. I need to get [51:51] as much of it out of it as possible, and [51:54] that includes learning from my classes, [51:55] right? But I think a lot of students [51:57] here just don't have that insight or [51:59] that that maturity, to be honest, to [52:01] recognize that, and it's just always [52:03] easier to take the easy way out. And [52:04] it's it's becoming incredibly easy, and [52:06] it's really hard to catch cheating. I [52:08] have like I know people that like use [52:10] tools that humanize certain things that [52:12] they get out of like ChatGPT for essays [52:14] that they would write. And I think [52:16] that's crazy that there are layers upon [52:17] layers upon layers of AI to replicate a [52:20] human's writing. Um [52:22] it's it's just really unfortunate. I [52:23] think that these people are going to get [52:24] hired when they have to do like [52:25] technical interviews for schools. Um but [52:28] yeah, it's it's really unfortunate, and [52:30] I think schools are taking bigger [52:32] measures to deal with it, but I don't [52:34] think they'll they'll ever fully be able [52:36] to [52:37] to handle all the AI tools that people [52:39] are using. They're just they're just [52:40] getting too good, in my opinion. [52:43] Yeah. [52:44] It's kind of arms race, basically. Uh [52:46] the AI tools can't tell they can't [52:49] detect [52:50] the output of [52:52] AI tools, no matter how much, you know, [52:54] companies try to sell these uh [52:56] you know, school systems [52:58] anti-cheating or cheating detection [53:00] software and stuff. Like fundamentally, [53:02] if an LLM can't tell whether it created [53:04] an output, then how are how are you [53:05] going to do that? And I had a a [53:08] guest a while back who was talking about [53:10] like there are certain attributes like [53:11] burstiness of an LLM, and like certain [53:14] like Oh, and a lot of people like [53:16] armchair, you know, [53:18] uh LLM detector people who love to do [53:20] that that proofing, [53:22] uh they they will um say, "Oh, if you [53:24] use a lot of m dashes, you that means [53:26] that you just And And you know what? I [53:29] use tons of m dashes throughout my [53:30] writing. For years and years, all my [53:32] writing was So I have to like rewire how [53:34] I write to not use m dashes cuz I don't [53:36] want people to think that I'm using an [53:37] LLM. So, it's ruined m dashes for me. [53:40] It's ruined m dashes for me, too. I [53:41] Those are were favorite things to use. I [53:43] remember when I unlocked this in high [53:44] school, I was like, oh, I'm unstoppable. [53:46] I know how to use em dashes. And then, [53:48] of course, ChatGPT still uses em dashes. [53:50] It's so unfortunate. Anyone that I know [53:51] uses em dashes, it's it's so [53:53] unfortunate. It's sad. [53:54] >> it's the most like uh it's the Flex Tape [53:56] of writing, basically. You can use it [53:58] for anything. [53:59] Um anytime you're not sure what [54:01] grammatical or like what punctuation to [54:03] use, em dash it, and you can just append [54:05] a clause. [54:06] Um [54:07] so, [54:08] where are you going from here is one of [54:10] the questions that I'm very curious [54:12] about cuz you're pretty early in your [54:13] college career. How many more years do [54:15] you have before you graduate? I'm a [54:17] sophomore right now, so I still have 2 [54:18] more years. Maybe we'll get my master's [54:21] here. We have a 1-year master's program [54:22] at Stanford. Um but, yeah. It's It's [54:24] really scary. It's going by quite [54:25] quickly, but yeah, 2 more years. [54:28] Yeah, and like what are your current [54:29] plans? I know you're working at the [54:31] security company, uh [54:34] and you've been teaching as Stanford. [54:36] Again, huge accomplishment to to be a [54:38] student and also being an instructor. Um [54:41] but, like like where do you see yourself [54:43] heading? Or like how are you going to [54:44] make the most of the next [54:46] 2 years or so before you look at going [54:49] to grad school or look at the job [54:50] market? Yeah, for sure. Um [54:53] I think there's honestly still a lot of [54:54] pressure for me to try a bunch of [54:56] different things here. [54:57] Uh I think [54:59] Stanford [55:01] I I [55:01] I don't know how I will feel in 2 years, [55:03] but as of the moment right now, I love [55:05] the school. And I I love the fact that I [55:07] can literally learn whatever I want at [55:09] any time by the best people in the world [55:12] at this school. And there's so many [55:14] things that I want to try that I just [55:15] like haven't tried yet. And it's scary [55:17] that I feel like I have to cram all [55:18] these things in 2 years. [55:20] Um [55:21] I think a big part of why I work so hard [55:23] is that I want my future to be stable, [55:25] and I want to feel like whatever I put [55:27] my time into at Stanford made me a [55:29] better person, that I grew more as a [55:31] person. And I feel like there's just a [55:32] lot of pressure, at least internally [55:34] within me, that like [55:36] I have to just work really hard here to [55:37] like make use of every opportunity [55:39] possible. And I think that mindset from [55:40] high school is like kind of coming back [55:42] to me. And so I've I've gotten better at [55:43] work-life balance here. But yeah, [55:44] there's this pressure that I feel to to [55:46] do as many things as possible. Um [55:49] So, yeah, right now well, I just wrapped [55:51] up TreeHacks, which was fantastic and I [55:54] think it really taught me a lot about uh [55:57] When I think about potential career [55:58] paths, I think a big part of what I was [56:00] pushed to do as a kid is the dream job [56:03] as a do SWE at Google. Like that is like [56:04] stability, you got money, like it's [56:07] always going to be like a great job, [56:09] right? Um [56:10] And so this was something that that like [56:12] I came into Stanford thinking, "Oh, I [56:13] want to do SWE at Google. I want to do [56:14] SWE at big tech." Yeah, and SWE is [56:16] software engineer. Yes, software [56:18] engineer. [56:18] >> like it's just a short sweet way to say [56:20] it. Yeah, yeah, yeah, [56:22] um [56:24] But I think after after doing TreeHacks, [56:25] it taught me more that like a lot of [56:27] what I really truly love doing is [56:29] connecting with people and building [56:30] community and leadership and just [56:32] putting on things that make other people [56:34] happy. And uh maybe my dream job is not [56:37] just sitting at my desk every day like [56:38] coding. I think this is what TreeHacks [56:40] has taught me. And so I think in all the [56:42] other future opportunities that I'm [56:43] considering, I do want to build my [56:44] technical skills, but also just like [56:46] connecting with people I think is a very [56:47] big passion of mine. [56:49] Um But yeah, so so TreeHacks just [56:51] wrapped up and I'm still thinking a lot [56:53] about what I want to do in terms of a [56:54] career just because I feel like I don't [56:55] have that many years left. [56:57] Which is scary. But yeah, right now I'm [56:59] working at a a security startup in SF [57:01] and I'm also working at a uh security [57:03] company in the summer. It's called [57:05] Armory. They actually just went public. [57:07] Uh but they're working on AI agents that [57:08] do penetration testing, which is [57:10] basically hacking into systems. Um I [57:12] think this space is super cool and it's [57:15] kind of like the first thing that I [57:16] tried here at Stanford. Um I knew that [57:18] like coming into here like I wanted to [57:19] try research in robotics and and [57:21] security and and even some bio stuff. Um [57:24] and so I looked down the list of things [57:25] that I wanted to try and I said, "Okay, [57:27] I'll try security first." I ended up [57:29] joining both the the cybersecurity teams [57:31] here at Stanford and I've I've [57:32] absolutely loved it. And then I was [57:34] like, "Okay, you know what? Like this [57:35] field is stable. This is something that [57:37] I think will always be big in the future [57:39] of AI-generated code. It's always [57:41] insecure, right? So, there's always [57:42] going to be spot here. And I think [57:44] security is super fun because it's like [57:45] very gamified and that like you're given [57:48] like kind of like a puzzle, right? And [57:49] then you have to like get your way into [57:50] it. And I feel very lucky in that [57:52] security I feel like was the first thing [57:53] that I tried here and I genuinely just [57:55] adored it and I loved it and I'm going [57:57] to try a bunch more things here next [57:59] year, but this might just be the thing [58:01] that I end up going into and I I've I've [58:02] I've loved all the startups that I've [58:04] looked at so far. Um and so yeah, I I [58:06] think for me right now it's it's taking [58:08] hard classes, but also I want to venture [58:10] out more into Silicon Valley. Um [58:12] maybe I would consider starting [58:14] something while either in school or [58:16] after I've graduated. I think a lot of [58:18] the best founders make good companies in [58:21] their late 20s and so I I don't know if [58:22] I would feel like experienced enough, [58:25] but honestly I don't know if you ever [58:26] feel experienced enough to make your own [58:27] startup, but yeah, I think for me right [58:29] now it's it's just I really enjoy [58:31] learning more about the business aspect [58:32] of of directions to startups and [58:35] um I don't know. I'm I'm still [58:36] considering a bunch of different [58:37] careers, but it's just still honestly [58:39] trying a bunch of different things [58:41] to see what I really like until I have [58:43] to venture out into the big world and [58:44] and make a decision. [58:46] Yeah, well that that observation that [58:47] we'll probably always need security [58:49] engineers essentially. [58:50] Uh there there's always going to be some [58:53] frontier that's being [58:54] explored and potentially exploited by [58:57] attackers and things like that. Um [58:59] sounds like a good place to focus. [59:01] Uh I I want to talk a little bit about [59:03] um AI and how it's impacting the [59:05] security field if you still have a few [59:07] minutes. Yeah. [59:08] Like [59:10] I can um obvious like there was this [59:12] curl maintainer. I don't know if you're [59:14] familiar with this. He gave a talk at [59:16] like I think it was like PyCon or [59:17] something, but he was talking about how [59:20] the volume of spurious [59:22] security disclosure like people [59:25] basically responsibly disclosing like, [59:27] "Hey, there's this vulnerability Uh and [59:29] opening, you know, GitHub issues and [59:31] things like that that it was taking like [59:32] 6 or 7 hours of his week. [59:34] Every week was just like people using [59:36] LLMs to analyze the curl code base and [59:39] find usually hallucinated security [59:42] issues and that he had to take all these [59:44] very seriously and it was just burning [59:45] him out cuz it's like [59:47] the LLMs apparently are not that great [59:50] at finding vulnerabilities, at least [59:51] they weren't a few months ago when he [59:53] gave this talk. But [59:56] what are some of the ways that security [59:57] is being impacted by AI systems? [60:02] I mean, so the company I'm working for [60:04] right now, basically what their job is [60:06] is they look at PRs, which are like pull [60:08] requests from GitHub. So whenever like [60:10] So basically I was with the way that [60:12] most software engineers work, like most [60:14] companies work, is they have something [60:16] called GitHub, right? And they'll have [60:17] these things called repos and think of [60:19] them as just like code bases that a [60:20] bunch of different people can [60:21] collaborate and work on. [60:23] And so a pull request is basically you [60:25] want to make an edit to something in the [60:26] repository. [60:28] And so you'll write code and then you'll [60:30] push it to the repository. [60:31] So these pull requests are basically [60:33] changes in code. And the the company [60:36] that I'm working at right now, basically [60:37] their entire thing is that they have [60:41] they look at the pull requests and they [60:42] see the code that often times Claude or [60:45] like Code X is writing and they will [60:46] say, [60:47] "You wrote specific things that are [60:49] definitely definitely not safe like in [60:51] any way, right?" [60:53] A lot of LLM generated code does this. [60:55] Like a lot of a lot of Claude code is is [60:58] unsafe. And I think it's getting better. [61:01] >> That would make sense though because [61:02] it's trained off of [61:04] the open web's, you know, code, like all [61:06] the open source repos on GitHub, many of [61:08] which are not secure, all right? Many of [61:10] which were written by people who didn't [61:12] necessarily have like a really firm [61:14] grasp on security or maybe it's older [61:16] code that hasn't been updated as new [61:18] security best practices have come [61:20] online. So it totally makes sense that [61:22] LLMs should be generating insecure code [61:24] left and right. Yeah. Yeah. And so, [61:26] that's what the the company I'm dealing [61:27] with right now does. And they also they [61:30] inject specific prompts into LLMs to [61:33] into the cloud to make sure that like, [61:35] "Hey, this codebase is built for this. [61:37] It probably will help this feature. [61:39] Let's make sure that this feature is [61:40] particularly safe." Um it's it's really [61:43] unfortunate that at least at Stanford, [61:45] there's no push to learn security stuff [61:47] here. Security is a very tiny space [61:50] within [61:51] the undergrad at least population at [61:52] Stanford. The people on my security team [61:54] have asked Stanford publicly, like, [61:56] "Could you Could you at least make a [61:57] security track?" Like, we have different [61:59] tracks that we can um major in at [62:01] Stanford within the CS track or within [62:03] the CS major. Um and there's never been [62:07] ever a security track for undergrads to [62:08] study. There's a grad There's a grad [62:10] track, but um [62:12] a big part of of what people have been [62:14] pushing for at least within my security [62:16] team is like, [62:18] "Stanford is outputting some of the best [62:19] software engineers, some of the best [62:20] engineers in Silicon Valley. These [62:22] people will be doing grand things in [62:23] their life. They need to know at least [62:25] how to write secure code. And a lot of [62:27] my friends are doing their first uh [62:28] software engineering internships. And no [62:30] one is writing code manually anymore. [62:32] Everyone is literally just using Claude [62:34] upon Claude upon Claude, which is it's [62:37] great in terms of fast output, but a lot [62:39] of people I think don't really know the [62:40] code that they're writing. And so, if [62:42] you don't even have this like push for [62:44] people to learn security, these big [62:46] people that are going to be writing code [62:47] that I guess they they don't even fully [62:49] understand and one of them leading big [62:50] companies, [62:52] their codebases are going to be so [62:53] insecure. And I I feel like it's [62:55] unfortunate that like Stanford doesn't I [62:57] guess [62:59] acknowledge this more publicly, but this [63:01] is something that I think more like I [63:02] think this needs to be more fundamental [63:04] within the CS core of any major at any [63:06] school. Like, I think more people need [63:08] to be thinking about security um in the [63:11] context of especially because [63:12] uh just AI is writing everything these [63:14] days. And people actually genuinely need [63:16] to understand the code that is being [63:17] written. Um So, yeah. Yeah. Security's a [63:21] huge issue and I think it's honestly [63:22] just getting bigger and bigger. And I [63:24] think big big companies like Anthropic [63:26] are are working on this right now, but [63:29] it's there's just so much AI-generated [63:31] code that is being written right now and [63:32] I don't know if it can be caught up in [63:34] time to catch all these bugs, to be [63:36] honest. [63:36] >> Yeah. And the rate at which code is [63:38] being created now is astounding and [63:40] >> fast. So, we're going to have to go back [63:42] through all the code that's been being [63:44] generated today once we have the [63:46] advanced stuff and we're going to have [63:47] to essentially have like the band-aid [63:49] solution that goes through and finds the [63:50] stuff. And it sounds like your your uh [63:52] company, the product that they're [63:54] building is um already doing that to an [63:56] extent. Like it's trying to catch it at [63:57] the PR level, uh but [64:00] you know, a lot of stuff will be closed [64:01] source until it kind of like gets opened [64:03] up uh or it'll just remain closed source [64:05] and people will figure out ways of [64:07] identifying these things without even [64:09] looking at the code base, just like [64:10] poking around in the system and they're [64:12] going to find these vulnerabilities. And [64:13] people are going to get pwned. [64:16] Yeah. It honestly feels like a band-aid [64:17] on top of this really really big issue. [64:19] I think if engineers are more pushed to [64:21] learn basic security, at least like in [64:23] their undergrad classes at um at [64:25] whatever school they end up working at, [64:26] right? Or if companies even push for [64:29] their engineers to take security [64:30] courses, this could be quite good. You [64:32] don't even have to know the crazy things [64:34] as long as you know the basics, um I [64:36] think this could really help in terms of [64:37] of people writing software that's that's [64:39] really important to [64:41] like the foundations of a lot of the [64:42] industry, right? Um but there's I just I [64:44] just don't see that push right now and I [64:46] think that's something that should [64:47] happen. Free Code Camp is definitely [64:48] emphasizing security a lot. Uh I mean [64:51] we we definitely want people to [64:53] understand and appreciate like the very [64:56] bare minimum, like the OWASP type stuff, [64:58] uh but but also just have that mindset. [65:02] Like And uh [65:04] I I mean hackathons are great, but like [65:06] are there hackathon like red team blue [65:08] team type hackathons as well [65:11] uh at Stanford? Like you all have other [65:12] events like that that people can [65:14] participate in? Yeah, yeah. [65:16] We don't This is something that we're [65:18] actually working on right now. I don't [65:19] know if you've ever heard of a capture [65:20] the flag before. Yeah. A CTF. Yeah, [65:22] yeah. So, CTFs I think are becoming [65:24] increasingly more popular. And this is [65:26] actually something that I that I This is [65:28] like a tiny bit of a tangent. I think if [65:29] people are interested in interested in [65:31] CS, I think it's very hard to get a CS [65:33] job right now in the market. I think if [65:35] you go to a big school, you're going to [65:36] be fine. But I think if you go to like a [65:37] smaller school, like community college, [65:39] it's really hard in my opinion to get [65:41] like a software engineering degree [65:42] just with all the all the changes that [65:44] are happening within the AI field. I [65:46] think a security is something that like [65:47] you don't have to go to a big school to [65:49] to do. I think a lot of it is like [65:50] practical knowledge. And so, something [65:52] that I did last summer was get like a [65:53] security certification. [65:55] If you feel like you're in the space [65:57] where like you're really worried about [65:58] job stability, I think security is [65:59] something that is very approachable from [66:01] from any lens, right? You don't It's not [66:03] something like research or you have to [66:05] go to like a big school for that. A lot [66:06] of it is just honestly practical [66:07] knowledge. [66:08] But anyways, yes, capture the flags are [66:10] something that have become really [66:11] popular. And basically what a capture [66:12] the flag is is you [66:16] you're given little puzzles and [66:17] challenges that you have to basically [66:19] like hack into systems or either do like [66:21] crypto things or you have to [66:24] um There's like a There's a bunch of [66:25] different things that you do with [66:26] capture the flags, but basically you're [66:28] trying to capture some sort of flag. So, [66:29] some sort of flag is hidden in some sort [66:30] of system and you either have to do [66:32] something security related to get that [66:34] flag out. And so, this is something that [66:36] I recommend to people in their first [66:37] learning security is is to do capture [66:39] the flags. Um There are I I participate [66:43] in red team blue team competitions in [66:45] college. This is something that I think [66:47] most other schools have, not just [66:49] Stanford. And if there isn't one, people [66:51] should start one. But I think it's [66:52] really fun. The pen testing competition [66:54] to the red teaming competition that I'm [66:56] part of [66:57] is it's really fun. We just hack into [66:59] systems and then people [67:01] will try to defend their systems and you [67:03] get rankings and you get points for it [67:05] and you end up going to like big [67:06] competitions. You get to travel. It's [67:07] really fun. Yeah. And I I love security [67:10] for the reason that it just feels so [67:11] gamified, you know? Like it's it's fun [67:12] that you get to hack into this thing and [67:14] uh the competitions that I that I do, [67:16] they're like legit teams of red teamers [67:19] that are trying to get into your system. [67:20] It's It's so It feels cool. Like they're [67:22] actual professionals shaping this event [67:24] and they you get to learn so much about [67:25] what the actual industry is like in [67:27] terms of penetration testing and and [67:29] defense. [67:30] >> I imagine the the feedback loop is a lot [67:31] tighter than like trying to bring a [67:33] product to market and then you see who [67:35] actually care and stuff. Like you're [67:36] you're actually directly trying to break [67:38] into a system and you're either breaking [67:39] into that or you're not, right? So So [67:42] you're like getting like That That's why [67:44] I think uh [67:45] you're you're using this term gamified. [67:46] Like it feels like kind of like you're [67:48] playing a game. Right? Um It actually [67:50] definitely does. Yes. Yeah. Uh so I have [67:53] one last question for you. You mentioned [67:54] like, "Hey, if people [67:56] uh want to [67:58] get like a you know, red team, blue team [68:00] competition or some sort of capture the [68:02] flag competition or even just a [68:03] hackathon going on at their school, at [68:06] their company, if they're like, "Hey, we [68:08] need to raise awareness of security [68:10] uh concepts." or "Hey, we need to raise [68:12] awareness." Like something I'm a big [68:14] advocate of is getting people to [68:15] actually use LM tools and not just write [68:17] them off because I [68:20] you know, used them 2 and 1/2 years ago [68:22] and they weren't very good and I used [68:23] them a year ago and they still weren't [68:24] very good, but if you use them now, [68:26] they're pretty good. [68:28] >> Yes, absolutely. [68:28] >> Yeah, and they're getting better and [68:29] better and I feel like so many people [68:31] haven't given them a chance lately and [68:34] they're doing themselves a disservice [68:35] because they could be getting a lot more [68:36] done. [68:37] Uh shout out to Magesh Mohapatra on the [68:40] free code game team who recently wrote a [68:41] book that's free on AI assistant [68:44] assisted development. Uh I'll put a link [68:46] to that in the description, but like [68:48] let's say hypothetically somebody's [68:50] trying to just get raise awareness [68:53] at their company or at their school, [68:55] trying to get other people to take [68:56] serious security seriously, trying to [68:58] get people to take like LM code gen [69:00] tools seriously, things like that or or [69:02] whatever it is that they're interested [69:03] in, they want to organize some such [69:05] event. Having just organized, along with [69:08] some other awesome people on your team [69:10] who I got to talk to, [69:12] Treehacks, like what advice would you [69:14] give to somebody if they had to start [69:15] from scratch and like try to get a [69:17] hackathon going at their institution? [69:21] Yeah, absolutely. [69:23] I feel very lucky that Treehacks is like [69:25] this massive event and the [69:26] infrastructure and the foundations for [69:27] the event were built on before I even [69:29] came to Stanford, right? So, I I feel [69:31] like I inherited this thing and I got to [69:32] make it slightly better the year after. [69:34] But if it is scary, I think to start off [69:36] with something new. [69:38] I think a a big part of the tips that [69:40] I've been telling other hackathon [69:41] organizers is there's a few foundations [69:42] to a hackathon, right? You want to make [69:44] sure that you have a few prizes, you [69:45] have a few tracks that people can [69:47] compete in, and you really just want to [69:48] give people all the resources and [69:50] support to make whatever that they want [69:51] to build possible. I think it's a big [69:53] part of Treehacks and our mission is [69:54] that [69:55] we fly you out here and we give you like [69:58] unlimited snacks and food and resources [70:00] and sponsors and connections and [70:02] community and you can literally build [70:04] whatever you want within those 36 hours [70:06] that you can that you can code, right? [70:08] I think that is the most pure vision of [70:11] any hackathon and I think that is the [70:13] core the core mission that people should [70:14] stick to. [70:16] And so, if you're fundraising for the [70:17] first time, I mean just reaching out to [70:18] local companies is always really [70:20] important [70:21] and I think a one project I think [70:24] there's a lot of different guides on [70:25] organizing hackathons that that I would [70:27] recommend looking up. [70:29] I think MLH is is Major League Hacking, [70:31] they're a big hackathon organizing [70:33] support. [70:34] I don't know if they're nonprofit or [70:35] not, but they're they're a great company [70:37] to talk about. [70:39] But one thing in particular for me that [70:40] I've learned a lot while organizing [70:41] events like this is [70:43] it's the like human part of organizing [70:47] that is the best part of it. And and [70:48] what I mean by that is yes, I did all [70:50] these great things like we interviewed [70:52] Sam Altman at Treehacks and and we had [70:55] all these really big sponsors come and [70:56] we had robot fighting and llamas and and [70:58] all these great things and we flew out [70:59] all these people, but my favorite part [71:01] of organizing any event is always just [71:04] interacting with people. Like talking [71:05] with the hackers I think is always [71:07] always my favorite part of the event. [71:09] And um before TreeHacks had any of this [71:12] big budget, I mean we also started off I [71:14] feel like pretty small and uh [71:18] for the fun events at TreeHacks, we have [71:20] pretty grand things now, but before we [71:22] used to have no budget. I remember uh [71:24] before we had llamas the year prior we [71:26] had literally no budget for any of the [71:27] fun things. And what we ended up doing [71:29] is we would rally up all the dogs from [71:32] all the like grad students around here [71:34] and we'd bring them to to TreeHacks and [71:36] that was like doggie hour, right? That [71:38] didn't require any money, right? Um and [71:41] that was still something that a lot of [71:42] people got joy out of. A tradition at [71:44] TreeHacks is to do a light saber battle, [71:46] so we have big steps um down into the [71:48] building and what we'll do is we'll buy [71:51] like the most cheap light sabers. Like [71:53] they break basically on impact, but [71:55] they're like a hundred dollars and you [71:57] get so many of them and people don't [71:59] care. They don't care. They just like [72:00] fighting with light sabers, right? And [72:02] so I think whenever people are trying to [72:04] build really grand scale events like [72:05] this, it's the tiny things like that [72:07] that I think matter a lot. And you don't [72:08] need a big budget in order to make these [72:09] things happen. Uh for a long time [72:11] TreeHacks operated on making these [72:14] really fun events happen without really [72:16] any grand scale budget or anything like [72:18] that. And so I think just interacting [72:20] with hackers from like an organizer [72:22] perspective is tremendously important. I [72:24] remember this year [72:26] like we spent a lot of money on food to [72:27] feed all our hackers and staff. Um [72:29] there was uh I think us directors [72:31] ordered pizza right before the our [72:33] keynote with Sam Altman and we had [72:35] leftover cold Domino's pizza. And I [72:37] remember I walked around that night as [72:39] hackers first started funneling into the [72:41] building and I was just handing out [72:42] pizza and their reactions were like, "Oh [72:44] my gosh, Domino's pizza. This is exactly [72:47] what I wanted." That cost us like almost [72:49] nothing to buy this pizza, but just [72:51] interacting with hackers and giving them [72:53] out this cold pizza like literally [72:55] brought them so much joy. And this is [72:57] what I mean, you do not need anything [72:59] big in order to build community at these [73:01] hackathons. I think community is [73:03] absolutely the biggest part. Um, [73:05] you just need to pour your time and your [73:06] effort and your passion, and you really [73:08] just want to build connections. And you [73:10] don't need anything grand scale to make [73:11] that happen. Um, so that that's my tip [73:14] for for organizers. And in terms of um, [73:16] kind of raising awareness about [73:17] security, uh, [73:19] one major part I I think people talk [73:21] about it, but no one really truly [73:22] understands how bad it is unless you [73:24] actually experience it. Um, [73:26] there's [73:27] I mean, being at Stanford, I am very [73:29] much exposed to all the AI stuff that's [73:30] been happening with using AI agents and [73:32] using Claude. This all just comes [73:34] normally to me because this is the [73:35] ecosystem that I'm going to school in. [73:37] But whenever I go back home, no one [73:39] knows what Claude is. No one knows about [73:41] any AI generated uh, code tools. And so [73:44] what I will do is I will just pull up [73:45] Claude and they'll be like I'll tell [73:46] Claude like make this website for me. [73:48] And I I will see their reactions be [73:50] like, [73:51] "This exists. This is crazy." Um, I [73:53] think it's just really cool to just I [73:55] think people are really intimidated by [73:57] these tools and that's very valid. I [73:58] think I was also very intimidated to [74:00] begin with, to be honest. And now I feel [74:01] very very comfortable with them. Um, but [74:03] just showing people the grand scale of [74:05] the things that these tools can help you [74:07] build and improving and and helping your [74:09] life be more efficient is it's it's [74:11] really awesome. I I had a friend at at [74:13] um, [74:14] at another university and he his lab [74:17] like never used any AI tools. And I just [74:20] told him about Claude and Cursor and [74:22] they used it for the first time and they [74:23] bought I think Claude Pro for like every [74:25] single person within that lab. And it's [74:26] just like it [74:27] totally changed their efficiency. Um, so [74:30] yeah, I would I would just recommend [74:31] just just showing people [74:34] just the awesome things that these tools [74:35] can help you build. And now people don't [74:37] have to be like bogged up with all the [74:39] the the basic code writing. I think [74:41] people should still learn the [74:42] fundamentals of programming, but 5 [74:44] coding exists now. Basically, you're you [74:46] can make anything like any idea that you [74:48] have, come to life because of these [74:50] tools. And that is that is so awesome. [74:52] And I love that we're in that that stage [74:54] of CS now. Yeah. [74:56] >> Yeah. Yeah, absolutely. And I just want [74:58] to like kind of coast on what you just [75:00] said, like I I strongly encourage people [75:02] to learn the fundamentals. Recurse Camp [75:04] will teach you the fundamentals. There [75:05] are great books. Uh if you get a chance [75:08] to go to Stanford, uh hit up on a [75:10] raffle's C++ classes. But [75:14] these tools are so incredibly powerful [75:16] and they can save you so much time and [75:18] energy. Like with hackathons, it's [75:20] incredible the projects that come out of [75:22] hackathons now versus a few years ago [75:25] where Yeah. [75:27] the level of polish, the the number of [75:29] the the scale of the ambition that you [75:31] can achieve when you can just have some [75:33] agents write the code for you and you [75:34] can focus on the higher level [75:36] architecture and stuff like that. And [75:37] you know how to fix the agent code. You [75:39] know how to kind of keep it on the like [75:41] from going off into the weeds and doing [75:43] dumb things. Uh you know how to like [75:46] figure out a way to like make it [75:47] relatively secure, which is not really [75:49] that big of a deal with a hackathon cuz [75:50] most of these projects frankly are not [75:51] going to be sustained after Like I guess [75:54] that's a a final quick question. I know [75:56] I said final question. Yeah. How many of [75:58] the projects from uh TreeHacks do you [75:59] think are still [76:01] you know, operating? Like were there any [76:02] that you were like, "Wow, that was [76:03] really impressive. That's going to like [76:05] like the team's going to continue to [76:07] work on it?" [76:08] Oh, absolutely. I think one special [76:10] thing about TreeHacks is well, just [76:11] because we are in Silicon Valley, a lot [76:14] of There's just a lot of startup [76:15] influence here, right? I think a lot of [76:18] When we when we are thinking about [76:19] TreeHacks and and developing it just a [76:21] few months ago, a big question for us is [76:23] do we want more of the startup space to [76:24] have Not control over TreeHacks. Not [76:27] I guess the word control and infiltrate [76:28] sound very negative, but do we want to [76:30] have more influence from from these [76:32] spaces? Do we want to allow venture [76:33] capitalists to have I don't know, to [76:35] talk more with with TreeHacks people? Um [76:37] I think for us this is important because [76:39] I think a lot of people want to build [76:40] companies out of some projects that [76:41] they've built from hackathons, but I [76:43] think the the core part of TreeHacks is [76:44] we just want people to build really cool [76:46] projects and the engineering innovation [76:48] is is so important. And if you want to [76:49] pursue a company afterwards, totally [76:51] cool. We will give you the resources to [76:53] do that, but we just want people to [76:54] genuinely build cool projects and [76:55] innovate cool things. Um [76:58] There there's uh there there are some [77:00] people that I met at the hackathon that [77:02] are actually trying to fundraise right [77:03] now for projects that they were working [77:05] on. And it it's really awesome. We have [77:06] we have a few sponsors that that can [77:08] definitely help out with that. Um [77:10] I mean, we even had like Y Combinator [77:12] office hours was one of the prizes that [77:13] you could win. Um so, I think there [77:15] there at least within TreeHacks, I I [77:17] haven't gone to many other hackathons, [77:18] there's a big push for whatever you end [77:20] up building potentially becoming a [77:21] startup idea. Um but yeah, I know some [77:23] people that have that have started [77:24] already fundraising for projects that I [77:27] mean, TreeHacks was [77:28] it was a literally a month ago and [77:30] people made really awesome projects and [77:32] now they're yeah, they're trying to [77:34] build something real out of it and it's [77:36] awesome. [77:37] That's awesome. Well, it sounds like a [77:39] huge win. Everybody check out my [77:41] documentary on TreeHacks, which will [77:43] have come out by the time this goes [77:44] live. Uh I'll link it in the [77:46] description. Uh Rachel, thank you so [77:47] much for giving us so much of your time. [77:49] I know you're very busy teaching and [77:52] working in security and of course it's [77:53] still in in classes, trying to finish [77:55] school. Uh [77:57] it's very inspiring to talk to somebody [77:58] like you who [78:00] uh just has your level of ambition, but [78:02] is also just like super chill. Um and uh [78:05] yeah, like you're very easy to talk to [78:07] and [78:08] uh I feel like you're you're like super [78:09] down to earth and uh yeah, like [78:13] thanks for all the guidance you provided [78:14] us uh for, you know, ourselves and [78:17] potentially for our kids as they [78:18] navigate, you know, a pretty uh [78:21] competitive future. Things are not easy [78:23] out there, they're not going to get any [78:24] easier, right? Yeah. [78:26] Well, thanks so much for seeing me. I [78:27] like really genuinely enjoyed this. It [78:29] was It was great chatting and I'm I'm so [78:30] happy that I was able to do this. I [78:32] Yeah, anytime, anytime. Thank you. [78:34] >> Awesome. [78:35] Thanks again, everybody. Well, until [78:37] next week, happy coding.