The Standup with ThePrimeagen
Original podcast
The very first episode. You’ll like it :) Also, ssh terminal.shop ;)
8,636 words · 4 speakers
Alright, so we are gonna try something completely new. We are going to do the first ever... We don't... this is just a working name right now, okay? It's called "The Name Is..." "The Show." Okay, we got people on here right now. We got TrashDev down in the bottom. We got Tej in the middle. And we got Casey Moratori up at the top. By the way, did you know that Tej streams? And we are gonna be discussing some pretty hot topics here in the tech world. One, is cheating on leaked code actually good? Or is it bad? Or cheating on, uh, technical interviews using the latest and greatest software that you've been seeing, where people do cheat and to be able to get these AmZon internships? And then two, we're gonna discuss a little bit about the Firefox privacy side of things. So we got some hot discussions.
We're gonna have some hot takes. And I have a good feeling like by the end of this, TJ will in fact be cancelled. So I think everybody should be pretty excited about the outcome to this.
I'm excited for my one and only episode. Thank you for inviting me.
Yeah, this one episode's gonna be amazing, TJ.
Yeah.
I have strong, I have strong confidence that you can do this. Alright, so our first topic to discuss is gonna be the leaked code cheating story. I'm not sure if you guys have seen this, but the leaked code cheating story comes with this beautiful website that says "Fuck leaked code." And then it has this whole tool so you can do invisible cheating live on interviews. And the creator, Roy, decided to livestream him, or effectively record a YouTube of him going through an Amazon interview, effectively passing, getting a job offer, and then LOL KekW like, "I'm gonna take a job with you guys," all over the internet. And then released a tool that showed that he could cheat his way through without actually doing anything. And he pretended to be dumb and did the entire copying and pasting from the LLM to get the exact answer.
Of course, Amazon and all of their brilliance did ask the, uh, the two media, uh, the two, or the median, the running median question, so the two-heap question, which is just like the oldest... That thing was in "Cracking the Coding Interview" like maybe 15 years ago. So, I mean, it's also on them for asking such a stupid question. But at the end of the day, is using leaked code to cheat on an, or, uh, using an LLM to cheat on an interview, is that okay? TJ, I know you got a lot of thoughts on this, so why don't you just lay it down on us right now?
This is probably why it's gonna be my first and only appearance on the, on the show. Uh, which is that I'm gonna go with the controversial take and say cheating is just bad. And I know, I know it's not popular. I know that this is the last time I'm gonna be here, but I'm gonna go cheating is just bad. That's my official position. And in honor of Casey, I actually brought a whiteboard. Is that okay?
Yeah, that's okay. Yeah. If you want to use that antiquated whiteboard technology, I mean, you're, that's very, very interesting. The 1700s will be very interested to.
I know. I don't have the cool, really cool setup. This is the best that I, this is the best that I've got going on here. So here is my reasoning why I don't think that you should do this. If you think of you, here's like, here's a person. Okay. And here's your brain for trash, maybe this, but like on average, maybe here. Okay. Every time you make a decision, you fill in a little part of your brain. You draw those neural pathways. You have a decision matrix inside of your brain that helps you make decisions. Every time you make one, when you don't bring back the extra change that you got at the store and say, this wasn't mine, you're filling it in. Stealing's okay. When you lie to someone, you're saying it's
good to lie. When you cheat, it doesn't matter if Amazon's a scummy company. It doesn't matter what they've done. It doesn't matter. Whatever. I'm saying in your own brain, in your brain. Okay. You're filling in cheating equals good. Okay. And for the rest of your life, you'll know you cheated to get your job. And it doesn't matter. It doesn't matter for the rest of your life. Okay. You're working there and you're getting money, providing for yourself. Okay. You cheated and you're filling in your own brain. You're filling in your own conscious and your own decision matrix towards bad things. So that's why I think you shouldn't cheat, even if you think Amazon sucks and lead code is bad.
Okay, so if I'm hearing you correctly, what you're effectively saying is that all of these people tuned in to this show because they wanted to see a 1990's after-school special by like the Disney Corporation. That's where the moral of the story is, look, you cheated on this Amazon interview, and then like 15 years later, you're addicted to crack in some back alley somewhere, and like it's all over, and it was that one thing, that decision to use an LLM on the Amazon interview, that is why you are having this life right now, you know? Yeah, okay.
Can I clarify?
Yes.
I only think that outcome is likely if you're using Grok for your advice.
Okay.
A little wild, a little wild.
Alright, got it.
But, okay, so I mean, to be fair, there is some sort of moral culpability that you have to have, that you have to be like, I am okay with cheating, if you're willing to cheat, even if you think the person that you're going against is bad, right? You still have to say, no, it's okay to cheat. Let's just pretend that Amazon didn't just cheat in general, but specifically cheated against you. I could even cite the old classic "eye for an eye makes the whole world blind", despite me not liking that phrase. But nonetheless, you are saying that vengeance is okay. If you cheat me, I should be able to cheat you in any other way as well. And I'm not sure if that actually makes a better world in any sort of sense. And I know, TJ, I'm probably gonna get canceled along with you, this is also my last stream, but I am actually on the side that I do not think cheating is good by principle, not in specific.
And so in specific, maybe there's some good reasons why you can kind of justify your choice of cheating, but I would say in principle, it's never okay to walk away from a store with some extra change. You should go back and you should give it, even with the inconvenience of four kids.
And I do want to clarify, I'm not saying using LLMs in interviews is evil. I'm saying lying about it and doing it to cheat, right? Like clearly the setup for like the Amazon interview and everything is they put you in some dumb web browser text editor wannabe situation, right? And they want to see you think and talk to them about the problem. Right? So like, if they said, "Use whatever tools you want," I would be like, "Good idea. Use whatever that interview coder website is to get the perfect answer. That's awesome." But like, you did, that's obviously not what they're doing. But like, that's clearly not the case.
I mean, to be fair, you're just to play devil's advocate, the interviewer cheated because he obviously didn't make up this question himself. He got it out of an old book, or asked an LLM, like, "What kind of question should I ask as an interviewer?" Like, "Hey, I've got an idea." ChatGPT in its infinitesimal wisdom comes back. It's like, "How about using two heaps to find a median continuously?" Something that literally no one has ever had to do in the history of computing. It's like, it's like, nobody does this. It's not a thing, right? It's like, it's like, you could go find, you'd have to find the .000001% of all computing problems where you actually want a continuous median, and for some reason, you're gonna write
it in Python. It's like, you're super concerned about the speed, but Python was on the table for some reason. Like, that's in the mix. N log N sorting the thing when we go to get the median, that's, oh, that's, that's gonna be too much to get the median. We're not gonna wanna do that, or we're not gonna get, wanna do O N time by just doing an actual, you can, you can do median in O N time, right? That's an actual thing, unlike, unlike sorting. None of that's okay. We gotta have it continuous, really reporting the median with no extra actual work, but we fired up Python. The programming language that takes 100 cycles to execute an ad, right? So, it, I, I don't know. I mean, is, is the cheating bad? Is it, should you have cheated on this interview? I'm, I'm inclined to agree with both of you that you shouldn't have cheated,
because that's just kind of bad for you. It's bad morally. It's bad, like, to that amazing whiteboard diagram. Can we get that back up on the, here? Is that still available? I don't know. It's like, yeah, I, I think that does kind of say it all from the perspective of the actual person, right?
It's worth a thousand words, at least.
I mean, that's, if, if that doesn't, that is your brain on drugs, is, is I think what he's trying to say here. And I, and I think, we, we can all agree, you don't wanna look like, like whatever that was on the whiteboard. But, all of that aside, the, the suck, like, if we wanted to talk about who is responsible for the percentage of suck, in this case, the percentage of suck is definitely Amazon for whatever that thing was supposed to be. I have to plead ignorance here. I've never seen one of these modern tech interviews, but wow. Like, I watched that video that you, that you sent. I don't even know what to say. So, so yeah, that's my, that's my take.
So I have a quick take. I think, obviously, we talked a lot. I think the obvious answer to this question is no cheating is bad. But I do think there's a responsibility on the interviewer's side to just not ask such dumb questions that can be easily cheated. So we've been doing this forever, right? Leak code interviews have existed. We've been sharing test banks online. People have been given a list of questions to study and effectively just memorize Leak code. So like, is that cheating in a sense? If you see a test bank where someone's like, they ask, they ask a question around graphs and you, and you study graphs. Is that considered cheating? Or are you just well-prepared? Right? So.
Well-prepared would be the answer.
Yeah, I would say that's fair. But like, honestly, I'm kind of glad like, this is happening because I think it's going to actually like, cause companies to like, think about this. And I know people complain about Leak code all the time. Like, I play the game. If I'm going to interview and I think it's going to happen, I'm going to, I'm going to practice, right? But I think it's going to like, result in some like, pretty creative interviews. So that's what I'm looking forward to.
I actually agree with that. I do think that the interview process is not a great interview. But obviously, there is something to the secret handshake, as I like to call it. You went off and memorized 50 hard problems. I ask you one of these 50 hard problems. Ah, you've done it, right? If you've read the book, The Wheel of Time, you have to perform 100 weaves to become an Isidae. And so that's just like a requirement. And so it doesn't make you into a great Isidae. You just have to know those 100 weaves to be able to be a powerful witch in the world of, of a Wheel of Time. And so it's like, there is something to having some sort of barrier. I don't think it's a good, I don't think this is an ideal one by any case. But I, it is something. I know that if someone can do that, they can at least program a function
that can do something complicated. Can they actually build real software where I'm going to find out in three months and either fire them or continue to go, oh, this person's actually really, really good. But I at least know they can program. But I think we can probably come up with much better ways at this point. All right, TJ, I did cut you off.
I'll just give my next hot take, which will definitely get me canceled, which is that, I actually think asking a question like the median thing or whatever is like.
It's actually kind of good as a first filter, assuming that you're not just literally copy pasting from the LLM, in the sense that lots of people lie about their skills and their requirement or like their abilities and just literally like a filter where you say, can you write any code and talk to a human? Like you like, Amazon gets so many applications that they kind of do have to do that. Now, could they do something better? Could they do something? I'm not going to argue that there's no better option, but I'm just going to throw it out there that like, people apply to stuff that they're wildly unable to do. I know I've graduated with people that I was like, how did you pass any of our CS classes?
Right? I've worked with people where I'm like, I don't know what you do here. I've never seen code of yours that runs. So I, I don't know, I'm not saying I love lead codes, but like asking an easy lead code is kind of like, okay, if everyone thinks it's so easy, then just like you should do it. That seems good.
I don't, I don't think it's easy or hard. That's the problem. It's mismatched for the task, right? It's kind of like as if you were going to say like, well, we're, you know, we're, we're screening for this NFL football team. And so what we decided to have people do is like recite Shakespeare and see how good they, you know, are, are they up to, you know, Patrick Stewart levels of, of presence here on the field? And you're just like, what, what is that measuring? Look, if Amazon shipped really performance software and that was their thing, right? They're all about like, wow, like we have to scale really, really big. And the main kind of problems we face is that our programmers just are using N squared algorithms for median, then it's like, okay, maybe this kind of thing makes sense.
But the reality is nobody there is doing that. Like that is just not a thing that's happening realistically. And most of what you're doing at Amazon, I'm just going to go ahead and assume is like, just like shoveling manure, right? It's like, you've got these legacy, these large legacy systems, you've got to deploy new stuff, the, the amount of time you're going to spend thinking about. Something like a dual heap median, which by the way, like I defy someone to find the case where you should be doing that. Like that's to me actually sounds like just a bad idea. And the fact that that's on an interview already is like a weird, I'm like, okay, describe to me exactly the actual scenario in practice.
Where you think that this Python dual heap median is something that someone should have pressed the like, you know, I, what, what, what is the thing that Devon can't do? Push to master or whatever, right? You know, describe that to me. It's very synthetic. It's, it's thinking in a way that you will not be doing 90% of the time when you're programming. And I find it especially disingenuous when these companies ask these sort of questions that, you know, they, they belie two things. One, it's like pretending that the company cares about performance when I know they don't, right? I know most of their time, they don't spend any time. But two, also, if you cared about performance, nothing would look like that. You're not programming in Python.
You're not thinking about using heaps, like heaps are not a very performant thing 99% of the time. You're not giving the interview candidate, like, let's say I was going to take this interview. I would be asking so many questions, right? Like, I'm going to be asking like, okay, wait a minute, what, what is the scenario where this actually makes sense to be thinking about? Like, what, why are you focusing on this aspect of the problem? Are you sure that this is something that you're actually doing? How did you measure how many queries you were doing of the median and all this sort of stuff, right? And maybe the idea behind the interview is to see if the person asked those questions, okay, sort of. But that still feels like a bit of a trick. It feels like you're asking me to answer this question just to see if I'm smart enough to know I shouldn't be answering the question. So I've, I find this kind of interview just sucks.
Figure out an interview format where you can actually interrogate the person about their experience and ability to write code that they are comfortable with, that they're used to writing. See how it goes, right? Don't pick some weird problem that's faux about performance when the interviewer himself probably doesn't even know about the performance of that thing. It just feels off.
Hey, to be fair, in 2011, I got a call from Google and they asked me this question. And at that time, I didn't know how to solve it because I hadn't read Cracking the Coding Interview. And you know, and so I asked that exact question, well, how often are we getting the median? And they said, oh, you know, pretty regularly. I'm like, well, how many elements are we talking about?
You know, is it like small? Can I just store it in an array? And they're like, oh, no, it's going to be like hundreds of millions. And I was just like, well, my laptop can't have hundreds of millions inside of a like, I don't even actually-
So is this out of core now, right?
Like, what are we doing?
Are we having multiple machines? I don't even know how to do this. Like, what are you talking about? I don't even have this kind of memory. And they couldn't even answer the question to make any sort of practical sense. And so I actually did exactly what you said, Casey. I did ask those questions and they promptly said I was not allowed to interview a Google again for a year.
I mean, it's so bad, right?
I believe I answered it correctly. I said, no, this is actually a really bad problem because it doesn't even make any sense. Like, I'm not going to be doing any of this. Like, at least ask me to do like binary search because I've actually done that maybe six times professionally. But I have never once done that kind of nonsense where it's this. Like, I totally get the sentiment, but that does not detract from the fact that it's still wrong to cheat. No matter how stupid the hurdle is, it doesn't make anything different, right? I'd rather take the, I think the bigger thing is that if you really hate a company's hiring policy, maybe that's a good sign that you don't actually want to work at that company. Maybe that's the bigger takeaway you should be having right now is that you don't want to work at Amazon. Because they are going to be like, that is the barrier they think is a good mark of an engineer. And maybe that's not the people you want to work with.
Well, it depends on what kind of flexibility you have, right? I mean, if you're somebody who really needs a job, then you don't necessarily have the flexibility to say like, well, I'm going to be choosy about what company I work for, right? So I can totally get that. And, you know, so I don't know. I, again, I'm sort of playing devil's advocate because obviously I don't think cheating really benefits anybody in the long run. But at the same time, it's kind of like, yeah, it's- The only thing I want to make sure isn't done is like focusing on the cheater as if they're the only crappy part of this. And it's like, yeah, I mean, they're, they're an additional crappy part. But the, but the wider crap is, is a, is a much stinkier,
more expansive crap, if you will. A real bowl clogger, if you know what I'm talking about.
I'd be interested, what's your favorite like technical interview either you've like had or given, Casey? Or like trash, or prime, I guess, too?
I prefer the Chris Hecker style interview. I always have. I think it's the best kind.
For us laymen that don't know what that means, can you enlighten us who Chris Hecker is and why does his interviews rock?
Sure, sure. So, so Chris Hecker, very old school. He is the person who first, first tried to get games on Windows. He made a thing called WinG, which was the precursor to DirectX. He made a game called-
Oh, I hate him.
Okay, all right. I'm sorry. You shouldn't. I mean, it was a good development. It's why we have Steam at all, right? Like-
Oh, we love him. We love him. Never mind.
So, yeah, and, and he made a game called Spy Party, which is, you know, a game you can play on Steam. Oh, and he did the creature creation stuff for Spore, for the Electronic Arts game Spore. Yeah. So his interview process was the best one, and it's the one I always have used and would use again if I was ever hiring someone. And it's very simple. It's basically a drill down interview where you take something that the person has done. So basically, you ask them for their history, and you kind of, you randomly pick kind of something you want to drill down on, and you drill down on it continuously until you get to like the very minutia of that problem. And this is such a great interview because it allows the person to talk about things they are comfortable with. So you're not
springing this like dual heap media nonsense on them, which is something no one's thought about, right? It's like, it's a, it's a vague CS thing that would almost never have come up in 99% of all programmers' lives, and they wouldn't have spent much time thinking about. This allows you to focus on something that you know you, they have thought about. And you get to see, like, can they remember the stuff they do? How far through it did they think? Are they able- To explain those things that they did and why. And these interviews are very productive in my experience and and typically are really good at at figuring out whether a candidate is actually a decent programmer or not. The only thing they can't do and leaked code interviews also don't do this is figure out how much someone programs.
That can be difficult, but I don't think there's much you can really do in an interview to determine that.
Yeah, you know, that is one thing about leaked code style interviews, especially if you have a keyboard that they can actually do it live. You can actually make a lot of judgments just on that. You know, when you see someone at least write parts of the function or parts of something in like a really fast, obviously, they understand the language really, really well. You can say, OK, this person does program a lot. Like when I was watching that video, my biggest takeaway was he's probably a really talented programmer just by watching him type and do stuff, even though he was feigning to be stupid. You could actually see, oh, yeah, he's actually really talented just just by his pure command of how he's interacting with his computer and all that.
By programs a lot, though, what I mean is, is this a person who's just going to talk all day with other people in the office or like spend all day on social media like that kind of thing? I mean, productivity based, because like there's definitely programmers who are very good if you force them to program for an hour. They're excellent, but they don't work like 90% of the time. Right? So what I mean is that you can't screen out in this type of interview and and I don't know how you do.
I actually wanted to double tap on what Prime just said because I had the exact same observation of watching that video. I was like, anyone who probably didn't know anything about anything could not would have wouldn't have passed an interview. Like it was so back and forth with the questions and I was like staring at his eyes a lot of the time and he wasn't really looking at the prompt at all. Like he was like literally solving it. So like the I wasn't even impressed with the product, really. I was more just like, OK, you already know what you're doing and maybe you'll peek at some syntax every now and then. But immediately after like the first like five minutes, I was like, Kiri knows what he's doing like he's talking so well. So like if you handed that to like my mom to pass the same interview, she would just literally crumble and she'd be probably be trying to type into chat GPT like what did this person just ask me?
I don't know if anyone else feels the same way.
No, that's a very good point. I think it's a very good point.
I was thinking the same thing where I was like, a bunch of people are going to buy this who don't know how to code and they're going to get to the Amazon interview and then it's going to be so incredibly obvious that they are cheating because it'll be like, hmm. I'm thinking we probably should use def function minums heap one heap two. What do you got? What do you think about that?
I unironically had someone do that in an interview. One of my last interviews, someone was typing. And I was just like, hey, here's the problems. They're like, okay, what are the problems exactly? And I'm like, here's the exact problems. I kind of write them in comments. And then I watched them highlight my comment, then their highlight went away. And they're like, I'm gonna start over now. And they erased all of their work and then just restarted us. It's like,
real, what, what? This is crazy. I can't, wow, I can't believe you decided to start over and write almost the same code. Huh. I wonder what they're doing.
Yeah, I do think asking people questions like you're saying, Casey, is much better, especially for, say, like, mid to senior level engineers. My, what, like, one of my favorite interviews was we did a very similar thing to that, except like they told me I could bring a project already that I had worked on. So they didn't have to like pick one randomly, but I could sort of like, prepare to do that. And so like, I just brought telescope and I just like said, here's what I did. Here's how I designed it. Here's the interesting bits. And then we just like chatted about it for 45 minutes. And for like mid to senior level, I think that that's way more revealing and way more useful, especially at like Amazon.
You are right. I think a lot of times it's like you're having meetings that could have been emails. So like talking about stuff and getting alignment is probably like a thousand X more valuable than, can you do two min heaps?
That is fair. All right. I'm gonna let, um, I'm gonna let one of the viewers have the last word on this topic. All right. The broadcast element says if your technical interview can be beat by AI, it's a poor technical interview. I generally agree. All right. Hey, well, we're gonna move on. Let's move on on topics. Okay, we're gonna move on. We're gonna talk now about our next topic, which is going to be about Firefox. Firefox just recently released a new update to all of their policies. They've included an acceptable use side, which actually has forbade. Even looking at any adult content, including downloading or uploading, which is kind of funny. I don't think people are gonna agree with that.
Trash, what are your thoughts?
Yeah Trash, what are your thoughts right now? But it also comes with...
Shut up! Wait, what? I'm sorry, can we pause for a second here? Because even separate from the... They put something in there that said you can't use the browser for adult content. They didn't really.
So, so the exact word says you can use Firefox under the acceptable use policy. And one of the acceptable use policy is you can neither download nor upload obscene... Or, I forget the exact...
Graphic content. Explicit and graphic content.
Graphic sexual content.
And so, I'm not sure exactly what they mean by graphic, like maybe not N64 graphics. I'm not exactly sure what they mean by that.
So you're basically telling me I can't use Firefox to update my OnlyFans anymore? Like, I can't even upload my content anymore, is what you're telling me? They're locking me out.
And you know what? This was actually the least of the internet's worries. Okay, the big worry is this thing right here, which says the following. "You give Mozilla certain rights and permissions." Exact quote. "You give Mozilla the rights necessary to operate Firefox. This includes processing your data, as we describe in the Firefox Privacy Notice. It also includes a non-exclusive, royalty-free, worldwide license for the purpose of doing as you request with the content you input in Firefox." This does not give Mozilla any ownership in that content. And so, people are kind of wigged out about that, because having the royalty-free access to your content, in some sense, might mean training AIs on it to give you a better experience.
The ability to be able to use your data to sell for them to make money with the recent ruling with Google and not allowing it as a default browser anymore. There's a lot of different reasons for this. And they've also recently, in the last year, purchased a company called Anonym? Anonym? Which I think is called Anonym? Which is an anonymous advertising-style platform.
Privacy-preserving is the word.
Privacy-preserving, that's right. They no longer say "privacy first," they're now "privacy-preserving advertisement company." So they are kind of gearing or changing gears into an advertising company. Now Casey, I know you love advertising, so why don't you tell me all about... How you feel about this recent Firefox change?
Oh, I'm the leadoff batter here.
Yeah, why not?
I was actually gonna do TrashDev, but I forgot literally mid-talk. And so I was just like, "Oh, I already selected Casey."
I know, it makes perfect sense. We have three web programming people and one non-web programming person. So the change in the browser, it's like, "Let's go to the guy who doesn't do web for a living." That makes perfect sense.
Yep.
I guess what I would say is, the problem with a lot of this stuff is that, you know, that, all of that, what, the entire update, right, is gonna be made by some legal team who's just like, they, you know, they're just trying to cover all possible bases. They're like, they don't wanna get sued in California under the, you know, Reverse-Gromulus Act that says that you can't do this or that with information. And they don't wanna get sued in Texas under the, like, Gromulus's Must Be Covered Act, which says that you can't have obscene material or whatever. So they're just gonna have this long thing that's just every last thing they can think of that they might get sued for, they just wanna make sure you get sued for that instead, right? Or that you can't sue them for it.
And so I think one of the problems with any of these agreements is it's like, you have no idea what they intend to do, right? And, and I think it's, it's bad in both ways. It's bad in the sense that you don't really want the agreement to just be so open-ended and say, "Well, whatever we do, it's fine, don't worry about it." Uh, but at the same time, I don't know how much trust people should put in the agreements when they don't say those things, because honestly, the companies seem to do them anyway. Uh, and then they just pay some class action settlement as, you know, to get out of it. And then they'll update it then, right? So, uh, I guess what I would say is, is this good? Probably not. Do I have a tremendous amount of concern about it? Probably not. Do I use Firefox browser?
No, I don't. Uh, that's about all I've got. So hopefully the web devs now can, can actually have an informed opinion. Uh, so take it away.
All right, TrashDev, Mr. Leadoff. Why don't you lead us off, because Casey's leadoff was amazing. He teed you up right now. Let's go.
So let me preface this. I'm not a big privacy person, so let me tell you a quick blurb about me. I use one password, literally.
And whenever I have to change it, I
You should not admit that, dude.
No, it's okay. I'm gonna tell you.
You don't mean one password to service. You mean one password.
What is going on right now?
I'm gonna tell you why I'm okay there.
Hold on, I'm gonna get there. You guys, you guys can get this. So I've been using one password for like literally the entire existence of the internet. Alright, I have a really small brain as
Is it dolphins?
Yeah, yeah. So and I also use Chrome. So that already tells you a lot about me that privacy is not like the biggest thing. If you went on the dark web, it would probably be a convenience store with all my information in it because it has been, my identity has been stolen like three times and like three weeks ago, someone's tried to steal it again. It's like insane. So I've already given up on my privacy because I know it's just splattered all over the internet. So like at this point, I just like, don't even care.
Are you saying you're not that into privacy because you use one password and everyone already logs into your accounts? Is that, is that what you're saying?
You know, I'm not gonna go any further into my privacy practices, okay?
If only there was a way to solve this trash tab. You could have your privacy back.
So you saw the recommendation from the security expert that was like, you should use a service like OnePassword and you remembered the OnePassword part and forgot the service part, I guess.
Why would I trust OnePassword, you know? I don't know what they're doing.
That's actually a good point.
Actually a good point.
That's what I'm saying.
Good answer.
Totally fair.
Well, do you want me to go trust someone else with my password? No, I trust whatever's in here that I often forget like all the time.
Singular. Like I'm gonna trust them with my one password.
To be fair, my one password is pretty long. It's a pretty big password. So I don't know if you can force it.
You're telling yourself that.
It has a special character. It has an uppercase, lowercase, and has some numbers.
Oh, wow.
That's pretty good. I feel like I'm pretty protected.
That's probably, you're not gonna be able to break that. That's, that's unbreakable right there.
It would probably take like a couple of years, maybe. I don't know. But I feel good.
Do you and your wife share the same password?
No, but she actually is even worse. She doesn't even know her password. She had a lot of her email. We couldn't get back at her email because she didn't even know the passwords. So we were like, just like, what do we do? It's it's it's it's like this household is, uh, we we don't even have curtains really. So people just like look in our windows.
No privacy. The password is written on the wall in big letters in case anyone forgets it. You can just look up there.
I'm not gonna lie, there are some passwords written in places where you could probably see it, but it's okay. It's just us. But anyways, back to Firefox, I can empathize. I feel that someone would care about privacy would be a pretty upset, right? If I read the terms of use, everything was worded so vaguely, they like backtracks like the worst part you could do is like release something and then just like backtrack and like reword it again into something else. And like, they're talking about how like cell means different things and like different states. And I'm just like, dude, what is this? Like, all I got from it was they worded so vaguely that they can ultimately do whatever they want. Because that's just what companies do. I'm not saying Mozilla is like evil corp, but doesn't mean they won't become, right?
At some point, they need money, right? We were talking about earlier, a lot of people know this, that the company needs money to run. The users are your product, right? So how are you going to get that? I guess potentially selling selling data or selling quote unquote data to third parties? That would attract advertisers or something like that. But me, obviously, you know, I don't care about privacy, but I think the dozens of people that use Firefox are pretty scared.
Okay, but hey, to be to be fair, I think I also think that this has nothing to do with the Chrome settlement and income. I've been thinking about this a lot lately, which is that people who are concerned about their privacy in the in the first place was likely going to change their default browser to something else. They don't want Chrome anyways, the people that are already up in arms. And so, yeah, you can no longer have a default search. And so, of course, what's going to happen? All the people who already would use Google will just continue to use Google. It's not like all these people on Firefox going, dang, I wish there was a way to use DuckDuckGo because that's the one I really want to use. But I guess I'm stuck here with with darn Google, right? Those are all the people already motivated to fix it. So my assumption is that they're like, they're actually going to have virtually no change in revenue.
I think this has been a long planned play to get into the advertisement space because they're sick of making a couple hundred million dollars. They want to make a couple billion dollars. And that's the big difference here.
Correct me if I'm wrong, there's just an open source fork of this anyway, right? Like, you can go install LibreWolfe if you don't want to use Firefox, is that, am I correct there?
Yeah, it's a fork of it, but I would assume that, you know, the fork is gonna probably be difficult as time goes on, especially if there's like weird stuff that gets put in. I assume forks become more and more difficult as time goes on.
Alright, TJ, you haven't spoken up at all.
Sorry, I was literally making a meme for trash, can you put it on the screen?
Alright, there you go, this is the meme.
Anyways, um, I think this seems like a really bad blunder by them, primarily because, like you're saying, this is just like saying, "Your whole pitch to people is 'We're the Privacy Browser', over and over, having to delete 'we never have and we never will' sell your data in a public GitHub repo." You can see the red of the diff getting deleted of "we never have and we never will", and now it just says-
That's a promise!
It does say "that's a promise", right, yes. And now you don't have to do that, I mean, in my mind, it's like you may as well just like, pack it up, restart as a new organization, and like, try again. Like, that would actually seem better, like, they maintain two different forks, or they do something else, because you're literally throwing away your super fans. Because the only people using Firefox are super fans, right? To be like, if Neovim was like, "And by the way, now it's scripting in JavaScript, and we're gonna do telemetry, and also you have to log in." Everyone would be like, "Why?" The whole, part of the whole reason we're using this is, we are tired of those things, right?
So, if C added a garbage collector, like, it's, it's like, none of these things are why the audience uses it. So I would have done, you know, I don't know, I'm not in charge, I've never been CEO of a really big company who mostly runs off Google's goodwill, but like, uh, I just think that trying to do literally anything else besides going down that route would have been much- Like smarter for them to exist in five years. Maybe I'll be wrong, but.
Now, correct me if I'm wrong, because again, this is not my area of expertise, but when they say promise, is it possible they just mean a JavaScript promise and that that promise has now completed? Because my understanding is that that would be fine, right? It's like it's for a limited time anyway.
Oh, wait till 2025.
Yeah, yeah, it's been rejected, I believe, is the technical term that you're looking for. The promise was rejected, unfortunately.
Okay, so, so what's the problem? I don't see the problem. This is a web company.
It's the web.
They're, they're just talking, they were, they were using a very technical term and maybe the audience didn't understand that. It was that kind of promise. And so, you know, that, that's what you would expect.
Promises get rejected all the time, boys. You know, no big deal.
I think the thing is that most people probably need to come to the conclusion of is if you use a free product, you're the product, right? And so since this free products product can no longer make the money they may not make, or they're actually wanting, my hypothesis, actually wanting to make much, much more money, they're going to lean into an alternative stream. And so you should never be surprised when this happens. Like, just like when Axe changed its privacy or its terms of use policies explicitly stating they can use all of your content to train on LLMs. And the rich part, of course, is everyone goes to Blue Sky and which is completely open and every single LLM trains off that. Kind of the hilarious leave, like, I'm gonna stick it to the man by training on 10 different LLMs now. You know, it's like, okay, that's very, very cool.
Probably Grok still trains off Blue Sky, too.
Grok is loving Blue Sky just as much. But nonetheless, like, when they did that, like, you knew that was coming. That should be no surprise because you're using a free product. Like, that is, that is what's going to happen. And so, I don't know, I guess I'm surprised that Firefox made it this long, more than anything else.
Well, they couldn't, they sort of couldn't not make it this long because they just, I mean, they're, they're as close to a state subsidized thing as we've got, right? Like, they are just sort of handed the money from Google and no, no matter what they did, they were just getting it. So it's kind of like, if they actually had to compete in some way and their, you know, ability to survive was based off of usage metrics, they would already. Be gone, but it's not right. That's not what they're actually that that isn't what determines their survival. And so, yeah, to your point, I assume that this whole thing is roughly motivated by the fact that they are worried that that gravy train.
Will now drive up, dry up, but I guess I don't really know. I mean, you gotta come up with your millions of dollars to pay your CEO who doesn't do anything somehow. And if if Google isn't going to do that for you, then you're gonna have to, I guess you're gonna have to get it from advertisers. I don't know.
All right, Trash Dev, I'm gonna give you the final word here.
Don't trust password managers. Trust what's in here. Use one password.
Trash, just a quick question.
Of course, the question by God is my judge. It's 95. What's your password, Trash?
It's too long. Starts with a P, ends with a number.
Oh, no. You just reduced the search phase.
There's at least like 40 things in between there. Okay? Good luck, buddy.
He lied about the 40.
Good luck. Dude, I swear if something happens to my stuff today, I'm gonna be very sad.
It's been 15.
But here's the thing is like, if you only have one password, then someone's gonna leak that password. Some service is gonna be starting in plain text.
Every time I log into something, it tells me that my password's leaked and I should change it and I just like ignore it.
This just keeps getting better.
It was like your password was like leaked in the breach. I'm like, yeah, whatever, man.
Yeah, don't talk to me about no breach. I don't want to hear about some breach. It's not not relevant to me.
You don't know where I'm from. I'm not scared of those warnings.
This password's stronger than any breach or whatever.
Oh, my.
This is not online banking, right? You have to go in person.
I do the multi-factor stuff. That helps, right?
It does help. Cut the MFA.
Pretty good. Feels like it's pretty safe. I don't know.
That's good.
I still have my, I mean, not gonna lie. I've had my like savings like wiped like three times.
But they gave it back. They gave it back. So we're okay.
But you just left the front door continued open. They're like, dude, you've been breached. What are you gonna do? I'll just leave the bank password the same.
Real scenario, I was on the phone with one of like the fraud people and he was like, would you like to change your password? I was like, can we just keep it the same? And it was just like 10 seconds of dead silence. And I was like, just make it the same, man.
Not even kidding. Cause I will forget and then I would just be left at it anyway. So.
Oh my gosh. What did you tell him? You're like, look, we.
I had a small brain. I just keep it the same.
As we know, we're able to recover it here. You just said you can get the money back. So I'll just call you if it gets wiped again. It's cool.
The account number and the routing numbers. Let's try it again.
Just hit them with only one in 20,000 people or whatever have identity theft. Like it's not gonna happen to me again.
I think once every couple of years is acceptable.
Oh my gosh. The reason why the reason why that stat is that way is that trash dev represents a huge portion of it.
Very large number of them.
This one guy's been breached 9,000 times. Therefore it's one in 20,000, everybody. One in 20,000.
It's like the whale. He's the whale of breaches. It's like, you know, there's this, this very bimodal graph where like if the heat trash is up at the top, occupying like 50% of all breaches is his one single password.
Should I just switch my browser? If I change my browser, will it be better?
This is such the wrong audience to tell. Like every time Prime opens a website, we get a million DDoS. Like you've made a terrible error. Go change your passwords now.
When this was a topic, I was like, I gotta tell them how I just, how I roll. I'm like the worst person to talk about privacy.
Well, I love how you equate the password with privacy. Like because it's so bad, anyone can access your stuff.
Oh my gosh.
It is what it is. Well, that was fantastic. I think it's time to wrap this up. Hey, everybody. Thank you for joining us in the inaugural episode. I still don't, I'm not convinced that I love the name. The name is the show, but hey, that is the show's name. If you like it, make comments. If you feel that TJ needs to be canceled for his extremely moral take on things. Oh, he's bringing out the whiteboard one more time. Then let people know. Thank you for subscribing. Thank you for being here.
Please, please post your guesses at Trash's password as well, so we can see.
That's what I want to know. What starts the P and ends with a letter?
Hey, you guys only use one password. Post that password in the comments, too.
What's your favorite one password?
Who's your favorite password? Post it in the comments below.
Page updated November 3, 2026 · transcript and summary are machine-generated