SPEAKER_01: Hey, everybody, we've got a great show for you today.

SPEAKER_00: Happy Monday. We are we're taking all the risks today. We're calling this third rail Monday, you will see why. But first we have another generative AI update for you from over the

SPEAKER_04: weekend. The bubble watch goes on. Yeah, anthropic is raising 300 million at a $5 billion valuation

SPEAKER_06: and a bunch of other AI startups seem to be raising at huge, huge revenue numbers.

SPEAKER_00: Big time. And then as I believe I predicted coding is in fact over Semaphore is reporting that open AI

SPEAKER_09: has hired 1000 contractor workers for data labeling and creating data for its AI to learn how to code. And then China's version of Google Baidu is creating its own version of chat GPT. What could go wrong?

SPEAKER_11: Yeah, what could go wrong? And then we have a conversation about

SPEAKER_14: Mr. Beast latest video where he cures people of blindness. Is it exploitive? Or is it endearing?

SPEAKER_15: Uh, and then we just go super third rail from that project Veritas Pfizer stuff.

SPEAKER_18: There's like vaccine talk. Like I'm just saying if you want to skip ahead, like just stop after Mr. Beast, go to the end and then you'll still like us in the morning.

SPEAKER_20: Yeah. Billy McFarland from fire. Yeah. Yeah, he's back. Just all the horrible stuff at the end of the show. Stick with us. It's amazing. It's gonna be a great show. It's gonna be a great show. Stick with us.

SPEAKER_23: This Week in Startups is brought to you by LinkedIn Jobs. A business is only as strong as its people and every hire matters. Post your first job for free at linkedin.com slash twist. Notion. Notion is one place for notes, docs, projects, and everyday work that goes way beyond a wiki. Get started for free at notion.com slash jason. And Cast AI. If you run software in the cloud and it's been a significant cost driver, listen up. Cast AI automates cloud cost reductions with client savings of an average of over 60%. Twist listeners can get a cloud cost audit with a personal consultation free of charge. Visit cast.ai slash twist to get started.

SPEAKER_30: Hey everybody. It's Monday. We made it. How you doing Molly? It's Monday. I did that gut check with myself today too because I don't know what. Anyway, it's Monday. It's Monday. It is Monday.

SPEAKER_36: Yes, we're back. You're back. And I'm back. Yeah, I'm feeling better. I was in Miami last week. Did a quick interview with the mayor of Miami. Like you do. Fan meet up like one does and had a

SPEAKER_39: little dinner for LPs and I'm back. But I'm feeling better. I'm at 85% I think. All right,

SPEAKER_41: we'll take it. Yeah. As long as you're not worse after travels, that seems like the right trajectory

SPEAKER_36: because travel can set you back. So usually it does, but I was on the tail end of I think

SPEAKER_35: whatever I got, it wasn't COVID. I took tests. Um, but yeah, um, I'm feeling better and yeah,

SPEAKER_36: that's it. Good. I'm better. I think I'm better. I'm hoping I'm better. I got more travel coming up. I'm going to, I got a speaking gig in Salt Lake City and then I got a speaking gig in Japan and I'm going to take, uh, four days to ski in Japan when I'm there. So I got a little mini vacay for Jayco.

SPEAKER_35: I'm doing a solo Niseko trip. I'm going to ski in Japan. So fun. By the way,

SPEAKER_49: we're both available for speaking gigs. So please feel free to reach out. Feel free to reach out.

SPEAKER_36: Yeah. Um, and coming up, uh, February 16th and 17th later, uh, this month, well, it's still January, I guess. But, uh, in a couple of weeks, we're going to have our next founder university and founder university is a two day program. We've done it. Gosh, two dozen times. And, uh, in this two day program, you're going to learn a bunch of important stuff as a founder, customer acquisition, sales, marketing, pitching, investors, hiring, all that kind of good stuff. Uh, I take pictures at it. Molly takes some pictures at it. And it's just a great way, uh,

SPEAKER_39: to level up as a founder, get to know our firm and it's free. Yep. It's free 18 of the founders who

SPEAKER_09: have gone through these two day workshops, by the way, have gone on to be accepted in the launch accelerator. And the one that's coming up February 16th and 17th is for women founders only. So

SPEAKER_00: about once, I think once a year, maybe one of the ones we do, uh, maybe a little more often is for, to really kind of incubate and nurture female founders, which, you know, it's a way for us to

SPEAKER_36: increase the deal flow, uh, from female founders when we created a specific space where women knew they were welcome, uh, then more women showed up. And so we do this, uh, go figure. Uh, and it's been a great success for us. It's really increased the number of female founders that we get to meet and

SPEAKER_39: ultimately invest in. So, uh, go to founder dot university to apply and it's a great, great two day

SPEAKER_61: workshop and free. Hope we see you there. And it's remote. Oh yeah. And it's remote. You don't have to go

SPEAKER_53: anywhere. Even better. And if you don't zoom, get into this one, you know, there will be three more

SPEAKER_36: in 2023, but yes, founder of university. We do basically quarterly and a great two day program. All right. Listen, there's a ton of news. This generative AI, uh, has just taken

SPEAKER_60: the industry by storm. I think the industry needed something to feel good about. And this impressive demo of chat GPT and stable diffusion though. Well, yeah, I think post crypto posts, uh,

SPEAKER_36: you know, the bubble bursting people are like, Hey, what's next? And we still don't have AR. We're waiting on those Apple AR glasses and people pretty much have given up on web three. So what's going to fill that space. I think yeah, chat GPT and stable diffusion have inspired people.

SPEAKER_39: It's a lot of developers are doing a lot of cool stuff with it. They are. I mean,

SPEAKER_00: I'm just going to go ahead and make my plug for all the climate opportunities that still exist. So we still have that. That's it's a double bubble. That bubble's still going. However, yes, all of the, uh, all the super tech bro excitement right now, it's all about generative AI.

SPEAKER_09: And, uh, we have now entered, I think the stage of the generative AI fund raising hype cycle being equally real potentially. Um, according to anonymous New York times sources over the weekend

SPEAKER_00: anthropic, which is quote an AI safety and research company that's working to build reliable, interpretable and steerable AI systems shrug is close to raising $300 million at a $5 billion valuation.

SPEAKER_67: Yeah. Let's let the numbers sink in for a minute.

SPEAKER_36: Yeah, no, I mean, I always like to read how they describe themselves. And I think the key thing there is it's a research company and they obviously, uh, care deeply about AI safety, uh, which is like these

SPEAKER_39: things don't cause damage in society that that seems to be on the minds of everybody doing AI.

SPEAKER_36: Uh, but we're far from having general AI, but there are other types of harm that could be done, like stealing people's content or misinformation. If ChatGPT is right half the time. Well, that's not good enough. Like you don't want a recipe that works half the time in coin flip and

SPEAKER_70: nor do you want medical advice or customer support or legal advice. So as impressive as ChatGPT is,

SPEAKER_36: uh, it can't be right half the time. It's gotta be right. 99.99% of the time. Right. Uh, to take a human out of it.

SPEAKER_41: I also have to think that a big part of this safety conversation does have to do with bias

SPEAKER_09: and abuse. Like, cause we've seen so many AI systems repeat, you know, racist rhetoric and all kinds of things. So the safety could mean a lot of different things here. And we don't know, we just know garbage in, garbage out, right? Garbage in, garbage out the, some of its co-founders, a couple of interesting facts. Anthropic was founded in 2021. Some of its co-founders include former open AI researchers. Its CEO actually is Dario Amodi, a former VP of research at open AI. And the website really, um, emphasizes the safer, more reliable AI systems messaging. Uh, Anthropic last raised, and this is extra interesting, $580 million at a $4 billion valuation. That was an, in April, 2022 series B. Its last round was led by Sam Bankman-Fried

SPEAKER_73: with Carolyn Ellison and other FTX. And executives joining.

SPEAKER_39: Which means the majority of the money or the largest check typically. And the person who sets the price when you say it was led by that's people who don't know. Typically the lead does two things sets the price and puts in the most money compared to other investors. So, you would think of the $580 million. That's a big number. Maybe you put $100 million in or $250 million in. There's no idea how to tell. Uh, but that's interesting because they bought shares in a company. I wonder if they've deployed that capital, Anthropic.

SPEAKER_36: And so then the estate or whoever's doing the bankruptcy of Sam Bankman-Fried, FTX, and trying to collect this money, which was obviously stolen or fairly obvious to me that it was ill-gotten gains unless he made it somewhere else. Um, that would go back to those shareholders. So that's an interesting wrinkle.

SPEAKER_09: Yeah, it really is. It's just sort of like side note. Some huge amount of money could potentially be clawed back, um, by the bankruptcy court. So we don't really know.

SPEAKER_39: Or if this is a round that's going to be at $15 billion or would they say $5 billion this round?

SPEAKER_36: $5 billion. $5 billion. So that's 20%, 25% increase from the last one. So Sam Bankman-Fried's money should have gone up 25%. So if you put $100 and it's worth $125 now, maybe they could sell those shares to new shareholders

SPEAKER_60: as a secondary right now. Maybe, yeah. And get a profit for the people who lost money.

SPEAKER_70: The march to $1 billion continues. LinkedIn is now at 875 million users. Unbelievable. There is no business network on the planet that comes even close to the scale of LinkedIn.

SPEAKER_89: So if you're a small business owner or you manage hiring at your company, you know that your success in 2023 all depends on your team.

SPEAKER_70: The team that you surround yourself is going to define your success. Don't I know it? And that's why you need to check out LinkedIn jobs. With LinkedIn jobs, you're going to hire qualified candidates much more efficiently. Why? How? Well, they match the open roles with people who have the skills, values, and experience that you're looking for out of the pool of 875 million members. And LinkedIn has the most qualified people at every level of employment. So if you have a strong following like me, you know you can add that purple hiring frame to your LinkedIn profile. And that spreads the word that you're hiring. Now you're going to get friends of a friend coming in. You know those high quality referrals. Your inbounds are going to happen much faster. That's why you need to be active on LinkedIn. Write some articles. Do some updates. Keep your profile fresh. So go find those amazing candidates today.

SPEAKER_91: People are looking. They're open to opportunities. And in 2023, you're going to want to fill your team with amazing candidates that you find on LinkedIn jobs. And that's what we've done at launch. LinkedIn jobs helps you find qualified candidates you want to talk to and it helps you find them faster. Post your job for free at linkedin.com slash twist linkedin.com slash TWIST to post your

SPEAKER_92: first job for free terms and conditions apply.

SPEAKER_00: So the other fairly large piece of information that is missing with respect to this now,

SPEAKER_09: potentially $5 billion company is what the product is. We don't know. Like you said, it's a research company. Anthropic appears to be pre revenue. Two articles covering the fundraising, one from 2022 and one from last week. Neither mentioned the word revenue. And the blog post that announced last year's fundraise said, quote, the purpose of this research is to develop the technical components necessary to build large scale models, which have better implicit safeguards and require less after training interventions, as well as to develop the tools necessary to further look

SPEAKER_00: inside these models to be confident that the safeguards actually work.

SPEAKER_09: Okay. I just read a bunch of that, but didn't understand it. So maybe if you were listening, you can explain it.

SPEAKER_99: Well, I mean, I think part of the research marketing,

SPEAKER_36: I'll be totally honest, this idea that the, you know, AI is going to become sentient and cause all these problems in society. It's, you know, it's a bit of science fiction right now. That's not happening anytime soon. That could happen in 5, 10, 15 years if we get to some sort of form of general AI. But I don't think that's, I think it's, people use that as kind of a marketing schtick right now, more than reality. Like what is chat EPT or stable diffusion going to do?

SPEAKER_41: Like they're going to flood the internet with images. And so what do you mean when you say safety?

SPEAKER_09: Like, I guess, you know, it's like, listen, I talk to a lot of companies that are in the R&D phase and they're raising money to commercialize. That's one part of this equation, right? Is do you, is that a, is that an investable and venture appropriate

SPEAKER_106: endeavor one, but two, what do you mean? What do you mean when you say safety?

SPEAKER_36: Yeah.

SPEAKER_14: So I think that it's marketing around this general AI becoming sentient and doing bad things in the world. Like literally people think it's going to escape the computer network and then go do bad things in the world. Is that what you think? Or that somebody will take it and use it for harm.

SPEAKER_70: So like some, this, you know, ephemeral terrorist organization. So Al Qaeda will get their hands on this and then do something with it.

SPEAKER_14: Um, that is really bad. Like, uh, like deep fakes, right? Like, so you make a thing where Putin declares war or says, I'm going to drop a nuke and then somebody reacts to it.

SPEAKER_15: That's kind of the hypothetical situation. I think a lot of these AI people start to convince themselves of. Oh, for sure.

SPEAKER_09: I mean, that's, you know, like, uh, as I reported from seeing Sam Altman speak, you know, he sits in public and says the worst case scenario is lights out for all of us, but I think. I sort of feel like there is a safety conversation to be had around AI. That's a lot more narrow than that. That doesn't have to be sort of like, and that's the part where like, that's what I want to know. Like, is that what they mean? Do you mean AI that is accurate that doesn't plagiarize that doesn't accidentally reinforce racial or gender stereotypes?

SPEAKER_18: Like, I would appreciate if it were me and I was talking about a hundred million dollar check. Yeah. I would want to know a little more about what you mean.

SPEAKER_14: They probably have some demos that they're showing people that have a path to some kind of product. But if you look at what happened with, um, you know, when Google bought DeepMind,

SPEAKER_15: they were, you know, trying to beat video games and do sort of proof of concept with that. And it was Google who looked at it and said, you know what? We can buy this company. I think I bought it for five or 600 million. And we can apply it to our stack of products, Google local, Google flights, Google shopping, Google search writ large, you know, Google ad network. We'll figure out where to apply this and point it. So that's the playbook they're going for here. Obviously build a bunch of talent, build a bunch of proof of concept, show that you have the dexterity to do all kinds of interesting things. And then you have the option to get bought by Amazon, Apple, whoever needs, you know, a couple of hundred AI developers, which are worth,

SPEAKER_36: you know, conservatively one, $2 million each, $3 million each is probably what they get bought for, um, as a group in an acquire.

SPEAKER_60: And then who knows, maybe they have an actual product in there. That's like some sort of chat GPT like, uh, product.

SPEAKER_73: The producers went through the job listings for product focused software engineers.

SPEAKER_09: Yep. Good reporting. Uh, some of the projects listed were developed conversational interfaces that leverage, leverage the language models.

SPEAKER_73: So chat GPT like products, build core components of the API to access large language models.

SPEAKER_09: So an API for others to build on sort of like open AI does. And then things like write billing, email, and other integrations with third party providers suggesting that some version of revenue will eventually be turned on.

SPEAKER_18: So really it sort of sounds like open AI 2.0, but with this kind of vague safety lens.

SPEAKER_36: Yeah, I would discount the whole safety thing as marketing. And it sounds like they're going to build the same exact stuff that open AI is building. And they will make it AWS like.

SPEAKER_15: In other words, you can go take this and use it, um, to, uh, build your startup, build your vertical product. Right? So I think this is kind of like the beginning of a new cloud computing era where there'll be the equivalent of S1 or EC2, whatever they are, you know, CPUs on demand, storage on demand, CDN on demand. Now you'll just have some, um, chat GPT like product on demand.

SPEAKER_14: And if you are, I don't know, intuitive, you have some billing software, you're going to be able to have it automatically figure out, Oh, these expenses are in this category.

SPEAKER_36: And it'll just, you'll use something like, uh, open AI, Azure, something on Amazon's cloud to do that. Uh, the thing that's super interesting, I think is in the short term, what this is going to do to

SPEAKER_15: developers, because it seems like developers are going to be writing better and better code faster and faster. So this is where this augmented human intelligence, I think is what's going to be particularly notable. If developers just as a group become 20, 30% more efficient, then they're going to be training the AI that made them 20 or 30% more efficient, which means that AI is going to become again, 20 or 30% efficient because it's going to see those developers say, Oh yeah, the of these five

SPEAKER_60: snippets of code you gave me, these two were perfect. This one needed a tweak. These two didn't work and I passed on them. So now you're training that AI in real time to write better code. So as developers use this, it's kind of a self fulfilling prophecy. It's just going to get better and better and better. And then developers, you know, are just going to be like watching the AI write code and like steering it. Well, and then that's going to get really weird.

SPEAKER_70: If you want your business to be more productive, more cost efficient, and you want to have bigger impact, you need to use notion. We run our entire business on notion here at launch. And, you know, we were looking for a deal flow solution. And these CRM systems for venture capitals, my God, quarter million a year, I was going to build my own software $100,000 a year, plus $50,000 a year just to maintain it all the existing solutions, expensive, slow, janky, not customized. And then we built our own system right inside of notion at a fraction of the cost. With no developers, we just used our team.

SPEAKER_91: And we built all these different views of all the notes that we were already keeping inside of notion. So whether you're starting a new gym routine, organizing a trip with your friends, or even planning your company goals, notion is a flexible collaborative workspace that helps you make meaningful progress in every part of your life. Get started in seconds by choosing from thousands of templates for every task and make it your own. From to do lists to OKR trackers and so much more. Notion lets you build the exact system you want so you can work the way you work best. Here's your call to action. Get started with a free notion account at notion.com slash Jason. That's right. Make sure slash Jason is all lowercase notion.com slash J A S O N to get started for free right now.

SPEAKER_73: It's so funny that you mentioned that because also over the weekend, it was reported that open AI

SPEAKER_00: has hired about 1000 remote contractors over the past six months. Sure. And that at least 40% of those contractors are software engineers who are creating data for open

SPEAKER_73: AI models to learn software engineering tasks.

SPEAKER_60: They have it. And that's a, and that's a model when you think about it. Microsoft owns GitHub, uh, which has co-pilot. Yep. That's also getting sued by some open source folks for stealing their code. And it's debatable how much code is protectable. Uh, you know, it's kind of like a recipe. So how much can you protect a recipe? You can't, um, you can protect the name cronut, like a croissant doughnut, but you can't protect the recipe. So there's been an ongoing debate about that for many years. Uh, but this could be very powerful. So if they're hiring all these contractors, that's IBM did that for Watson back in the day. They would just say here, show us. You would show the same image to three people on three different continents with three different education levels, right? What is this? And then, oh, that's an orange. Okay, great. Oh, that's two oranges. Okay, great. That's two oranges on the beach. You know, whatever keywords they put in and they label, if two out of three of them or three of them or one out of three, then that you would put on, um, how sure you are of what that is. So you can imagine with code, all you have to do, Molly, is

SPEAKER_70: say to like 10 developers, write the code to, you know, uh, say hello world or do, you know,

SPEAKER_14: LinkedIn login authentication. It looks at the 10 pieces of code. These three worked and had no bugs.

SPEAKER_15: These three needed to be edited. And then you look for some commonality and then says, okay, this is the best way to do

SPEAKER_39: LinkedIn authentication for your website. Right. Now nobody ever has to write that again.

SPEAKER_73: Right. You just, instead you go to a version, you know, and, and listen, like, I know somebody who's doing this now goes to chat GPT and is like, generate me some code for this. And then has to, you know, human check it. Massage it. Like, cause it might be wrong or whatever, massage it into what they need.

SPEAKER_09: But it sounds like what they're doing. And, and I should note that 60% of the contractors were hired for what I call janitorial data janitor, right? They're cleaning up their labeling. They're creating these sets of images and clips and other information. And then these, these engineers who are specifically creating data, um, around software engineering, one of the things that they're doing is almost like commenting. So taking pieces of code and annotating them with the natural language, the human explanations behind what the code does to help associate the code base itself with the action that somebody

SPEAKER_141: might type in. For example, I want you to create me, you know, an auto LinkedIn login.

SPEAKER_73: So open AI, remember has that existing product called codex that is designed to translate text into code. They power.

SPEAKER_144: That's what's co-pilot is built onto it.

SPEAKER_09: Yeah. Right. Which also is text to code. So I'm just going to go back to what I said a couple of days ago, which is like, don't learn to code, learn to wire. Like the computers, the machines have been coming for this job for a long time. And in fact, Tesla's former head of AI, Andres Karpathy tweeted the following last week. The hottest new programming language is English.

SPEAKER_39: Yeah. I mean, people are talking about that, that if you can actually do the prompt engineering,

SPEAKER_15: like we talked about it all in the other day, I think here as well, um, you can have an interesting job if you tell the code how to program itself, right? Like, so I want you, when you say, I want you to write me a play about two podcasters talking about Sam Bankman freed. And then when you say in the voice of Shakespeare with a little Quentin Tarantino, uh, and that,

SPEAKER_14: um, rhymes like Eminem, it's like, okay, wow. Those are a lot of conditions that you added to it, but it understands them. And so that that's super fascinating. And I was watching Karpathy, um, do a YouTube video this weekend where he made his own chat GP two instance with like version two of it, the open source one.

SPEAKER_15: And it was like a really interesting video. I just watched it in the background while I was doing some work. It was like an hour and a half long video where he was explaining how all these predictive models work.

SPEAKER_60: Um, and he actually, um, showed like this Shakespeare AI, uh, which is on GitHub. And so people are taking like, you know, verticalized content, the complete works of William Shakespeare, and you can just go on a onto, you know, uh, GitHub right now, grab that Shakespeare AI and then put it into your software.

SPEAKER_36: So you're like, I want my, you know, um, podcast to be translated into Shakespeare and make us

SPEAKER_14: translate our, you'll be able to translate this podcast and have us talk like we're Shakespeare characters, but keep the complete intent and then republish it in our voices with us as Star Wars characters. Like it's going to be pretty crazy when you think about like what this AI is going to be able to do.

SPEAKER_73: Right. And then what are the aspects of that that are related to this kind of safety marketing pitch?

SPEAKER_09: I think what we're likely to see just as a, a little bit of a kind of bubble watch prediction is like, for example, this race, um, of this, uh, whatever that I've already forgotten the name. Anthropic is like cash in now with the slightest of differentiators because you're one of the eight people, right? Remember that moment in time where all of the autonomous driving startups got acquired Argo

SPEAKER_00: and you had the Waymo intellectual and they raised tons of money because there are only like 11 people in the world who know how to do this. Yep. It is quickly going to become a commodity. Yep. Or much like self-driving it's way, way, way, way harder than anybody thinks. And so it's going

SPEAKER_09: to get like speech to text. Remember when it was, um, like dragon naturally speaking or something would be, it was, they were like, yeah, it's 85% accurate, but that 15% is the difference between something that is functional and makes you sound like normal and something that doesn't really work. And so if AI gets 85% of the way there and then gets stuck much like autonomous driving, then now is the moment to get your $5 billion valuation and your aqua hire and slap whatever differentiator you

SPEAKER_154: want as many shares as you can. Yeah. And a lot of investors are going to lose a lot of money right

SPEAKER_14: now. Just saying in order to make money on this, you would need to have, if this was a publicly traded

SPEAKER_36: company with a hundred million dollars in profits on, let's call it 250 million in revenue, 250 million in revenue. Um, you know, if they were selling to enterprises at $25,000 a year or something,

SPEAKER_15: there's a lot of customers, um, you would have to have a hundred million to have a 50 times earnings to make a $5 billion valuation. Um, there's a long way to go. So as far as a bet goes, it's not a terrible bet. Um, because if the company sells for anything above the capital it's raised, you get your money back. So one important thing to understand here is if it's 5 billion and they've had a, let's say a billion dollars invested in the company, if the company sells, but for a billion, all those investors get their money first, obviously. So that's why sometimes people will take a

SPEAKER_60: optionality bet here, which is, you know, if it, if it happens to go to 50 billion, they have a 10 X,

SPEAKER_39: but that would be very hard to do. It's very hard to get to a billion dollars in profits,

SPEAKER_78: which is what it would imply if they got to 50 billion. Yeah. At least a billion in profits,

SPEAKER_73: maybe 2 billion in profits. Hard to. Yeah. Tread carefully. People tread carefully. Uh, speaking of treading carefully, this is just straight up Jason Bate. Oh, okay. Last, uh, last in our trio, our triptych of generative AI stories. Baidu, China's equivalent to

SPEAKER_00: Google is building its own version of chat GPT Bloomberg reported that exactly a CCP approved dystopian version of chat GPT. Yep. Uh, you know, to, to sort of tell you, like, you ask it a question. It's like, yeah, I got, I got my version of the truth for you. Sure.

SPEAKER_15: Uh, makes sense. I mean, yes. I mean, everyone's going to open source and you can fork it and build on top of it. And it's really going to be about the data at this point, you know, like you're going to need to have a hundred engineers, 50 engineers to program it, but it really is going to be about the data set. And so I would think Baidu has some very interesting data sets, very, very interesting data sets.

SPEAKER_73: I mean, ironically, right. All, all the data sets. I mean, China is effectively a surveillance state.

SPEAKER_09: Yes. So the data advantage that China potentially has, I find to be a fascinating question. Like, does China know more about people than Google? Honestly, I don't know if I honestly know the

SPEAKER_15: answer to that question, but I would say it's probably close to even what you could think about is, you know, when you're on YouTube and that whole like intellectual dark web kind of exercise. And then there was that lawsuit where, um, in France, uh, somebody who was murdered by terrorists said, Hey, we, this was kind of, uh, this person was being steered on YouTube towards more and more

SPEAKER_60: extreme content. Yeah. And that led to, so there was a whole, um, I did a bunch, by the way, a bunch

SPEAKER_178: of reporting about that. The, the ladder of radicalization, there's like a ton of actual

SPEAKER_14: research about how it happened. It's a thing. So if the algorithm can steer people towards more and more extreme content or more, let's just call it adjacent content. So adjacent content's either going to be, you know, it could be steering negatively. It could be stirring positively. You know, you get into

SPEAKER_15: some beautiful, uh, guitar solo. It could take you in a positive direction, right?

SPEAKER_14: Um, but it could also take it in a negative direction. As I say, and yet somehow it never does. No, it does. I mean, I see a lot of cool, like Pink Floyd people doing covers and Mark Knopfler covers going on in my, uh, YouTube feed. So that's positive for me. Good. Um, but here's what happens. You know, if it can steer people in certain directions to adjacencies, well, what the government could do with that is if they weren't just interested in making more money, they could actually know, oh, we're steering somebody towards being radicalized, or this person is doing searches that are the precursors to somebody becoming a dissident, right? Somebody becoming, uh, a protester. So they might know from some early signs of what you're searching for that you're eventually going

SPEAKER_60: to wind up as a protester there. And then it becomes minority report. We have the precogs, precognition, knowing this person potentially could become not a terrorist, but a protester, you know, a freedom fighter. Um, they've got too many thoughts about freedom. They're searching for the constitution. They're searching for freedom of speech. They're searching for certain religions, whatever it is. Um, and that's where I think this becomes super nefarious. And then you combine

SPEAKER_04: that with people's credit card data, location data. Maybe this person was looking in a bookstore window.

SPEAKER_70: And you know, like, hey, they have certain books in the window there. Yeah. And just the fact that the person was hanging around a bookstore, plus their YouTube data. Now you know something about

SPEAKER_82: them, and you could go pick them up and torture them and reeducate them. Yep. I mean, AI makes all of

SPEAKER_73: the what ifs into potential realities, you know, like when you imagine the worst case scenario of like,

SPEAKER_09: well, what does it matter if you have all these disparate pieces of data about me? Well,

SPEAKER_157: it weren't actionable. If all that data was sitting there in a database, you'd be like,

SPEAKER_191: who cares? Who cares about my location data? Who cares about my YouTube data? Right? Exactly.

SPEAKER_157: Yeah. It's like, oh, well, if the AI stitches it together. And now they're saying you're a

SPEAKER_15: potential terrorist, you're a potential freedom fighter, whatever, you know, framing you want to

SPEAKER_194: use. And you don't even know it yourself. They're like, no, I'm not. It's like, you will be. They're like, you will be just trust me, you're going to be part of the rebel alliance.

SPEAKER_178: Yeah, it's pretty obvious. Like, you just talked me back into the safety marketing.

SPEAKER_14: Just tell me what you mean. I mean, there are things that will be obvious, like deep fakes, right? It's obvious, you can just imagine in your mind. But it's the cases where we don't know what the AI is going to come up with the AI is going to find things that are going to be very challenging for humans to deal with.

SPEAKER_201: Mm hmm.

SPEAKER_04: You know, and when we find those uncomfortable truths, that's going to be really challenging for human brains to kind of understand it.

SPEAKER_202: Yeah, I mean, Asimov's first law is a really, really like if you ever sit there in the dark

SPEAKER_09: and truly contemplate what that means, and then apply it to what we think AI is about to reveal

SPEAKER_41: about us, like, it's, it gets, you get a little, you're talking about like a human, some despair can occur humans. The first law says that, yeah, the, that, that the prime, the only job the robot bot has is to

SPEAKER_00: protect humans, protect humans. So then the robots inevitably determine that the only way to protect humans is to kill all the humans because humans are so bad for themselves.

SPEAKER_18: Right.

SPEAKER_206: So, and like that's where I'm, that's where I'm going with this idea that what it

SPEAKER_207: uncovers about us is that we're the problem.

SPEAKER_14: Well, and then you have the seminal work of Megan, uh, you know, and Terminator. I have not seen that. You know, I've got to protect John Connor and then Megan's got to protect, you know, her, uh, little girl and it's like, oh, the doctor is going to put a shot in their arm and it's going to hurt. Okay. We have to kill the doctor now, you know, like in some horrific way or, yeah,

SPEAKER_06: somebody is administering some basic discipline on a child and it's like, oh, okay. They're harming the child after they protect them. Yeah. And kill this person. Uh, good times, good times.

SPEAKER_189: But what are the other two? There's three laws.

SPEAKER_00: There are three laws. Yeah. It's like a robot, uh, like what has to protect them, may not injure a human being or through inaction, allow a human being to come to harm is the first law. The second law, a robot must obey the orders given it by the human beings, except where such orders would conflict with the first law.

SPEAKER_217: Got it.

SPEAKER_09: Which is how they always end up killing us. And then the third law is a robot must protect its own existence. As long as such protection does not conflict with the first or second law. Clever, super clever, clever. I mean, it was to be clear, a narrative device, not a real law. I just want to. Yeah, we could, we could do better now that we know that the second law is what allows them to use the first law to wipe us out. Yes.

SPEAKER_11: It's a loophole if you will, but, um, but yeah, uncomfortable truths.

Chamath Palihapitiya: If you run software on AWS, Google Cloud or Azure, you know how out of hand the bills can get, right?

SPEAKER_70: The pricing uncertainty can really hurt your business. You get these surprise bills, but you can solve that right now with cast.

SPEAKER_91: Hey, I cast AI audits and optimizes your cloud cost and performance for you, which the cloud companies don't do automatically, right? You have to do it. You have to take control of this and cast AI is going to help you do that. They eliminate the stuff you're paying for, but that you don't use. And they search for less expensive hosting options within your cloud provider. And then you begin saving immediately. Listen to this on average cast AI customers save over 60% on their cloud spend. Imagine what you can do putting that capital back into your business, right? Everybody's in austerity mode. Everybody's in optimization mode and cast AI is going to do that. It's priced for a tiny portion of what you save. So before you go and sign any multi-year cloud contracts or make any drastic personnel decisions,

SPEAKER_70: check out what cast AI can do for you. So here's your call to action. Cast AI will give you a free cloud cost audit, which you can save and a personal consultation. Visit cast.ai slash twist to get started. That's cast.ai slash TWIST for a free cloud cost audit today. And we met with the founders last week and they are really excited to help founders and listeners to This Week in Startup save on their cloud bills.

SPEAKER_224: So give them a call. Go visit cast.ai slash TWIST. Speaking of uncomfortable truths, let's talk about Mr. Beast.

SPEAKER_226: Oh, he's awesome. Um, he's having a cultural moment of maybe of medium awesome.

SPEAKER_00: Okay. So we all know Mr. Beast is a huge deal, obviously, right? Like this is very popular.

SPEAKER_09: He's sort of changed, I think what it means to be a creator celebrity. Like he's got this potentially billion dollar business. All these people are investing him in it. He's got the burger empire. Like Mr. Beast is an empire.

SPEAKER_00: And this video posted this weekend and just purely anecdotally, I will say my son and I were

SPEAKER_09: watching YouTube last night. Like we do because nobody watches TV with their parents anymore. You just watched various YouTube videos and we saw this thumbnail hop up for this new video. Yesterday, Mr. Beast posted and here's the thumbnail. Yeah. And in case you're not watching, it's Mr. Beast with his, I've always found slightly empty and scary grin.

SPEAKER_00: Um, and then this kid with these like big, weird, fake looking, creepy crocodile tears running down his face and a bandage. And the video is titled 1000 blind people see for the first time.

SPEAKER_09: Amazing. So in the video, he took 1000 people who suffered from blindness caused by cataracts

SPEAKER_229: and he paid for the surgery to repair their eyesight. I watched it.

SPEAKER_14: Yeah, it was quite, this is a classic video on YouTube.

SPEAKER_108: Like there's that whole, uh, do they call them cochlear implants where you can hear again?

SPEAKER_13: Yep. There are tons of those, right?

SPEAKER_108: Yeah.

SPEAKER_208: They're usually one of those. I get so emotional when somebody can hear their child for the first time. And then there are these ones with blindness where people can see again for the first time. And so I guess the device he came up with here was if seeing one of those is amazing,

SPEAKER_108: why don't we do 10 of them, a hundred of them, a thousand of them?

SPEAKER_00: Totally. If one of these is totally engaging and pulls on your heartstrings,

SPEAKER_09: let's just blow that up. And so it is, it's just provoked this really interesting reaction. So I will tell you that last night we saw this thumbnail pop up and literally my 15 year old was like, that's creepy. Like his immediate response was that feels gross and exploitative. And so we skipped it. We were like, no, let's not watch it. That's again, I was like, yeah, you're right. That's there's something weird about that, that I don't like. And so then I wake up this morning and Mr. Beast is trending and I see that like the internet is going crazy because tons of people are having that exact same response and tons of people. And I will say much like in a predictive AI, I was like, I know exactly what's going to happen on the show. Jason's going to be like, that's awesome. And I'm going to be like, that's exploitative and weird.

SPEAKER_00: And we, and we are going to have the exact. And so I want us to be able to mirror the conversation that's happened on the internet today because people are having really strong feelings about this.

SPEAKER_13: Let me guess, because I could tell you the, dare I say, woke position on this. Don't. Well, I could tell you what the, what is woke about it? I'll tell you the woke position.

SPEAKER_226: Oh God. I'll tell you the. Now I'm regretting pushing so hard for this. No, no. I knew it was going to happen and yet I'm irritated. All right, let's go.

SPEAKER_14: Let's go. You heard what I'm going to say. This falls into the, this is going to trigger for people.

SPEAKER_240: Dare I say, um, what is the state of the world when a YouTuber has to pay for people's healthcare? Why don't we have healthcare?

SPEAKER_241: Yeah. Which is completely valid. One. Uh huh.

SPEAKER_14: That is the default. Uh, yes, that is like trigger number one. We don't have healthcare in this country, which is totally valid. But counter to that would be, well, he's a rich guy with a huge platform. And so at least they helped a thousand people and he's raising awareness for this. So what's wrong with that?

SPEAKER_70: The second one is going to be, what kind of world do we live in when people are used as props for him to make money and get more famous? Yeah. It's kind of that second one.

SPEAKER_208: He is using them for exploitive reasons and there's no dignity. And if he was a good person, he would just quietly pay for all of their surgeries.

SPEAKER_04: Right.

SPEAKER_248: And why doesn't Jeff Bezos also pay for a billion people? And if Jeff Bezos gave everybody a billion dollars and everybody be a billionaire.

SPEAKER_132: And we would have no poverty. Yeah. So those are the two reactions. I am certain happened.

SPEAKER_09: Oh yeah. A hundred percent. Yeah. And then there were the ones that are just like, you're also exploiting, exploiting disabled people, right? Like you're just moderate, monetize, monetizing the suffering of the differently abled. And, you know, so there was that version too. Um, someone, someone's tweet said, I'm tired of having to perform gratitude for wealthy people just to stay alive.

SPEAKER_11: It's like, yeah, that's the why I think it really.

SPEAKER_09: And then there's the, and then there was the really messed up joke that I don't want to, I don't want to dox my son, but he was like, yeah,

SPEAKER_00: he got a thousand more views. Hey, wow.

SPEAKER_262: Yeah. That's a dad joke. Um, sort of. Yeah. Wow.

SPEAKER_00: I mean, it's like, you know, and then I was like, yeah, what's he going to do next? Kidnap a thousand orphans and then make them work for him for free. And then say, I'm a job creator. Anyway, it's, I think it's like, it's, it's both of those things at once. Right? Like, that's amazing for those people to whom this happened.

SPEAKER_226: It's a raw shock test and read into it, whatever you want. They were a hundred percent exploited for him to get clicks.

Jason Calacanis: Like, and it just is like, you start to see the, I mean, this is sort of,

SPEAKER_00: this gets to what you were just saying about the YouTube algorithm, right? Every, if you want to win at the game of entertainment, you have to get, like,

SPEAKER_09: it wasn't enough for Mr. Beast to give some people some money after making them stand in a parking lot for a few hours or what I was 10 hours. It wasn't like you could do that. And then you can give someone a house and then you could give someone a car and then you can give them 50 Lamborghinis. And then you've got to create the Wonka house. Like the wheel of entertainment when it's click related is like to get more and more and more and more extreme. And then eventually you land on something like this, where both of these things are

SPEAKER_00: true at once, like yay. And also, uh, like there's something a little creepy about it.

SPEAKER_04: Well, you know, this is not like some brand new, um, phenomenon.

SPEAKER_274: Obviously there was a show called queen for a day. Yeah.

SPEAKER_14: And this TV show basically would bring people in to give their emotional sob stories and then

SPEAKER_70: give them, you know, a refrigerator or a washing machine or a hearing aid. This show, uh, was like in the forties.

SPEAKER_278: Fifties. Oh yeah. Yeah. I mean, this is American and popular.

SPEAKER_00: This is American entertainment at its finest. Like reality TV exists. It takes some schlub, give them a, hear a sob story. Like the American idol with the weepy thing. And they go and they visit the home and they're like, oh my God. And everybody in the family only has one leg. And I mean, it's just like, it could not be more. Like American entertainment is exploitation at its core. And Mr. Beast was the guy who seemed to be doing it a little bit better, right? You felt better about it. And then all of a sudden you see this kid with the big fake crocodile tears. And you're like, ooh, oh, I get it. Okay. It's just the same stuff.

SPEAKER_258: Ah, yeah. I mean, I think everybody.

SPEAKER_132: Just the NFL draft like Nick said all over again. Everybody who's in this video opted in to do it. Some of them not only got free surgery, he gave them 10 grand. Mm-hmm. Um, if the video pays for all that surgery, I think he probably breaks even on most of these videos is my understanding of it.

SPEAKER_14: Some of them he loses money. Some of it makes a little money.

SPEAKER_60: But like his bigger business is going to be like his, his, uh, edibles, his candy. And I guess he could do edibles too.

SPEAKER_283: Um, but I don't, I don't, it's like the brand, the brand.

SPEAKER_135: I don't have a super problem with it. I think people are looking for something to be outraged about. I don't think.

SPEAKER_264: No, I think they're both true. I think that I think both things are true. I think a little bit of it is both true at the same time.

SPEAKER_00: Like this is a little bit Jesus-y and the thumbnail is creepy as hell and it's exploitative and America's messed up and capitalism sucks. And Mr. Beast did this nice thing for these people, right?

SPEAKER_09: It's like, I think all of those things are a hundred percent true. And I'm delighted that Jason and I were basically able to role play Twitter for you today.

SPEAKER_285: In case you're wondering, you no longer need to click on the trending topic. You don't need to.

SPEAKER_15: You just, you're welcome. Pick your opinion that reinforces your tribe and you can pick a capitalism or one screenshot

SPEAKER_286: from the NFL draft in 2020. That's absolutely sad, but it's hysterical. Okay. Go ahead.

SPEAKER_287: Yeah. Totally. Oh yeah.

SPEAKER_286: So ESPN got roasted for this because every like, just read like this was a, so this is

SPEAKER_289: like T Higgins, Clemson wide receiver, by the way, this guy's awesome.

SPEAKER_290: He played in the AFC championship yesterday.

SPEAKER_292: So he's still in Oak Ridge finalist for 2016, Tennessee,

SPEAKER_04: Mr. Basketball sister, Kiki played basketball for Tennessee state mom, Camilla for 16 years.

SPEAKER_294: And this was like every single first round pick.

SPEAKER_295: They were like, here's John. His mom died of cancer. It's like, oh my God. It was. Yeah. It's a little craziness.

Jason Calacanis: America's gross. Y'all like, that's why, that's what I'm saying. We had higher hopes for Mr. Beast. I wanted him to be better than Oprah.

SPEAKER_04: Uh, listen, hands on a truck says something about America. Like, hands on a hard body. Hands on a hard body. Is that the whole, it was a documentary. I remember seeing it back in the day and it's like, okay, sure. Like it's a contest. It's fun. Nobody gets hurt except it's extremely painful and you pee your pants.

SPEAKER_14: I mean, I have a friend who, what are we saying about America that like a mom with three kids who's a single mom can't afford a car and she's got to torture herself for 47 hours.

SPEAKER_00: Like they're all true. I have a friend who feels this way about hot ones. Like we had this big art. She thinks that that hot one show where celebrities go on and eat hot wings until they cry is exploitative. And I was like, what?

SPEAKER_305: Like, come on. Like they all know what their poor celebrities are trying to get people to see their Marvel movie.

SPEAKER_308: After we have to show one, we have to show another one from ESPN. I'm sorry. Oh my God. We have another ESPN one. This is amazing. Stop it.

SPEAKER_312: This is not real. You made this up.

SPEAKER_313: I can't believe this is real. Listen to Chanel Junior.

SPEAKER_00: Wow. Okay. Hometown DeSoto, Texas. Hobbies include video games and Sudoku. Spent time volunteering in high school to honor local veterans.

SPEAKER_315: His mother, Annie, contracted and survived West Nile virus in 2012.

SPEAKER_318: It's like, what are we talking about? What is going on here? Oh my Lord.

SPEAKER_315: You know what?

SPEAKER_318: I think I changed my mind.

SPEAKER_322: Mr. Beast is just doing what we do. You know how many people work on productions like the NFL draft? Like hundreds of people. And they all just were like, that's fine. And they're like, yeah, air it. Post that. It's fine. Love it.

SPEAKER_108: Let's just put some more sadness there to re and just, can the guy just play back? Can the guy just be like a great prospect?

SPEAKER_324: Can he just? We need some more sad music.

Chamath Palihapitiya: It literally is like everything has to be sadness and trauma porn.

SPEAKER_328: And shout out to his mom for surviving West Nile in 2012 though.

SPEAKER_12: Yeah. I think he just moved up to in the draft based on that. That's right.

SPEAKER_332: Good genes, strong immune system. God bless America.

SPEAKER_278: All right. Mr. Beast is just doing what we do. He's just out here being America.

SPEAKER_14: I was going to write some fiction when I was in my youth and I wrote the scene. I got to pull it up. It's on one of my old laptops. And I wrote like this scene where I was like, you know, it'd be like really interesting. This is getting super dystopian watching like Shark Week. I just watched Shark Week. And I was like, I wrote the scene where people in this dystopian future would swim

SPEAKER_70: across chummed waters where like Air Jaws was and they would do it to get clemency for their family member who was in jail and it would pay back the state.

SPEAKER_132: And the state was so corrupt that they were like, yeah, like if one of your family members or you as like the prisoner want to swim across for entertainment on this pay-per-view, we'll take the money from the pay-per-view and pay off your debts. And like, you know, like five people jump in the water and swim from one platoon to another. You actually invented squid game.

SPEAKER_337: And like two of them get taken out by sharks and they make a 10 million dollars and they're heroes.

SPEAKER_338: Yeah, you invented squid game.

SPEAKER_132: And I literally invented squid game before that happened. And I saw squid game. I was like, huh. But then even more than that, there was just a Shark Week where a guy was in a plexiglass tube. And I guess, I don't know if we can show it here without it getting censored. We'd probably get a strike against us. But anyway, if you guys can pull it up. Guys in a glass, you can get a still photo of it. Guys in a plexiglass thing. And I'm like, that plexiglass is not going to survive Air Jaws. No way. Like if one of those big 20-foot sharks.

SPEAKER_340: Yeah, no chance.

SPEAKER_132: Goes full speed into that, it's going to shatter. And literally, I'm watching this video. It shatters exactly what I thought would happen.

SPEAKER_14: And the guy's left, like instead of being in a plastic cube, he's like left like floating on a sheet of plastic, while a great white shark has just demolished the box.

SPEAKER_344: Mm-hmm.

SPEAKER_14: And they're like, I wonder what happens. Yeah, there it is. So this guy is in a plexiglass, like in a coffin basically. Yeah. And you're like, um, okay. Yeah, that's a plexiglass coffin. And you're in a seal outfit.

SPEAKER_347: And yeah, that's a 20-foot shark. And yeah, the shark just comes out. And what do you know?

Chamath Palihapitiya: He's like, I'll take a bite.

SPEAKER_108: And literally this piece of plastic, which some, you know, person's like, yeah, that piece of plastic will be fine. The shark can't bite through it.

SPEAKER_352: It's like, oh. Okay, we don't have to keep seeing it.

SPEAKER_169: That is, I'm gonna have nightmares tonight.

SPEAKER_73: Yeah, the shark rocket eating the seal. I mean, it's also jackass, right? Like this is literally the grand tradition of American entertainment. I think it's just.

SPEAKER_91: Should this stuff be allowed online is the question. Does it inspire people to do more dangerous stuff?

Jason Calacanis: Oh, it 100% does like that. That was the whole right. That was the I mean, I remember again, doing those stories in 2016 about stupid Jake Paul and Logan Paul and like the stunts get more extreme and the tick tock challenge that

SPEAKER_00: like literally we had to stay home from school one day because there was that tick tock challenge that was like somebody you show up and kill somebody. Yes. I don't even know, right? Like this is.

SPEAKER_09: This is a 100% part. Stunt culture is dangerous. Stunt culture is dangerous. And now I think the argument is that now Mr. Beast is part of stunt culture.

SPEAKER_362: Should we touch the third rail? A wah wah.

SPEAKER_252: Should we go right to the third rail? Sure. Speaking of stunt culture. You know what? I'm kind of on one this morning. I made that orphan joke. Let's do it. Let's do it. You saw this Project Veritas thing with the Pfizer executive?

SPEAKER_11: No. Oh, you didn't see it. You didn't see it.

SPEAKER_00: I literally ignore anything that has Project Veritas in it because I'm just like, oh, no, thank you. Please tell me.

SPEAKER_108: So literally Project Veritas trended over the weekend. Um, you know, their stunt culture, they do kind of what Dateline would do, which is like

SPEAKER_70: frame people, you know, like people doing bad things, of course. Yeah. So they use dating as a pretext to get people to go on hidden camera. They get a Pfizer executive.

SPEAKER_14: Who predicts that he could be on Project Veritas. He's like, yeah, this, you know, there's always hidden camera stuff, but he starts lying about how he claims he was lying. He starts talking about how he's going to do gain. They're doing gain of function research at Pfizer and how it's great for Pfizer because they're going to sell so many vaccines and all this stuff. And literally then the guy James O'Keefe comes out and he's like, does the Dateline thing. It's like, hey, can I talk to you for a minute? And the guy's like, I'm a liar. I'm a liar. I was lying about everything just to impress my date. And now here we are.

SPEAKER_132: And the guy goes crazy and breaks the laptop. And there's a big fight scene. You have to see it because.

SPEAKER_369: Is this like recent or old? This is over the weekend. No, no, no.

SPEAKER_73: I mean like the actual, this footage, like did this, this footage occurred recently to suggest

SPEAKER_00: that Pfizer is currently working on engineering a virus that they can then magically cure and make a ton of money. Honestly, Pfizer responded as well.

SPEAKER_18: Did Pfizer respond today? Because they haven't responded over the weekend. Oh, really? What would even surprise me about any American corporation at this point? Like really in related news. Did you see the story about Exxon?

SPEAKER_00: Which was that like in the seventies, as far back as the seventies, Exxon had scientists who

SPEAKER_378: specifically and unbelievably accurately predicted everything that is happening now with respect to global warming.

Jason Calacanis: Like literally they were like, oh yeah. They were like, they're going to be tons of super storms. And like a lot of people are going to die in Bangladesh is pretty much going to be gone.

SPEAKER_00: And like, I mean like literally pretty shockingly good science.

SPEAKER_09: And then we're like, we should definitely spend billions of dollars on a decades long disinformation campaign to make sure that no one finds out about that.

SPEAKER_00: Imagine literally finding out that you are at the crux of possible human extinction and being like, we should probably cover that up.

SPEAKER_226: We should cover that up. Yeah. Nothing would surprise me at this point about any American corporation.

SPEAKER_208: But this is going to be open up a can of worms because the guy basically says like, yeah, we're basically doing and the, but you know,

SPEAKER_14: he's being led on and they don't release the whole tape. So you don't know how much he's being egged on by his.

SPEAKER_09: That's the problem is that Veritas stuff is not a trustworthy source.

SPEAKER_14: It's gnarly to use dating as a pretext to get people to make confessions like this. Like, I don't know. We like this does not say anything.

SPEAKER_388: It doesn't exist in the world of journalism, right?

SPEAKER_162: And so it's a, a YouTube banned the video. Other places didn't ban the video. It's like, or I, I don't know if YouTube banned it or not.

SPEAKER_00: And then what happens, honestly, what happens is that even if it's true and it probably fricking is, I'm sorry, it's probably fricking true. It's possible. Yeah, sure. It's going to be hard to believe it. Like you're going to, you will have, you will have, when you're an unreliable narrator, you rule out a huge chunk of your potential audience.

SPEAKER_11: Correct. Who won't necessarily believe that it's the truth. Mainstream media will cover it.

SPEAKER_14: Right. Mainstream media does not do these tactics. Like mainstream media is not going to break into an office building to get documents, right?

SPEAKER_04: They rely on leakers. I mean, I only did that one time. Compass here. So they would never send a New York Times reporter onto Grindr or match.com or whatever app they used to set up a date to then kind of lead the witness into confessing about things, whether the confession is true or not. Mm-hmm. They would never use those tactics.

SPEAKER_39: The guy from Project Veritas says the only way to get this information is to use these tactics.

SPEAKER_00: To be honest, even if they use those tactics, and I'm not saying that's never happened in the history of journalism, it's a pretty cutthroat sport, right?

SPEAKER_09: Like I don't, I'm not saying, I don't think that is common now. I don't think that happens a lot, but I'm just saying, even if that did happen and you got this like bombshell thing, your initial response would not be to be like, aha, we got you and then put it out. It would be like, okay, now I will embark upon a painstaking, years-long process of trying to confirm this before I report it.

SPEAKER_04: These are tactics that like Musaad, KGB, CIA use, right? Yeah. Trapping people with love and sex and dates.

SPEAKER_84: That's like, you know. Remember when Meg Whitman said that about reporters?

SPEAKER_09: What? She was like, oh yeah, yeah. This was not very long ago. And then she called me personally to apologize because she got, so she was in like in all hands and she was like, don't talk to reporters. Um, they will groom you like pedophiles do effectively. Like we could look it up and get her direct quote, but she was just like, you know, they'll, they'll try to make friends with you and they'll do this thing.

SPEAKER_00: And she kind of like compared journalists to pedophiles and then made this series of like personal phone calls. It was at Quibi. It was when she was at Quibi.

SPEAKER_405: Oh my God. Yes. Our use analogy was inappropriate.

SPEAKER_130: And I had just talked to her. Well, Whitman said an interview at the same time.

SPEAKER_135: None of us are ever perfect. I didn't intend it. I don't know. That's not how I feel.

SPEAKER_91: Whitman reportedly compared the way journalists cultivate sources to tactics used by called predators to groom victims at an all hands meeting.

SPEAKER_179: On Friday nights were mostly accurately reported. Okay. So she said it was okay. Yeah. Um,

SPEAKER_00: People said they were strange and off-putting. Which was such an, it was very interesting. I thought she handled it. I mean, frankly, she handled it beautifully. She literally personally called reporters that she had talked to, called them directly and apologized and whatever. But I, and then I'm again, as long as we're on third rail day, third rail Monday. I was like, I literally was like, well, they kind of do do that though.

SPEAKER_11: Well, yes.

SPEAKER_132: Um, we're out of here making friends, not grooming in predator kind of way, but you might build rapport.

SPEAKER_412: You build rapport, which is what predators do as well. Like a spy. Yeah. Like a spy would do you build trust, but then you pump people for information.

SPEAKER_252: The important thing is learn the source. You then try to confirm that information.

SPEAKER_00: You don't just go like, aha, we got it. We got the story. You confirm. You do the reporting. So here's the Pfizer response, by the way. In the ongoing development of the Pfizer-BioNTech COVID-19 vaccine,

SPEAKER_09: Pfizer has not conducted gain of function or directed evolution research.

SPEAKER_84: Okay. Directed evolution. Yeah.

SPEAKER_11: And this guy says they do.

SPEAKER_09: And this guy says they do. Now they say that. In the ongoing development of Pfizer-BioNTech, if you were really going to parse this, you'd be like, well, they didn't say they did it any other time, didn't do it any other time. But working with collaborators, we have conducted research where the original SARS-CoV-2 virus has been used to express the spike protein from new variants of concern. This work is undertaken once a new variant of concern has been identified by public health authorities. This research provides a way for us to rapidly assess the ability of an existing vaccine to induce antibodies that neutralize of newly identified variant of concern. We then make this data available through peer reviewed scientific journals and use it as one of the steps to determine whether a vaccine update is required.

SPEAKER_00: Express the spike protein. Is that that's not the same thing, right? As gain of function. I mean, I don't really know, right? Like, who knows? Medical research is crazy. And I would not put it past any pharmaceutical company anywhere in the world to engineer

SPEAKER_378: viruses that they can then magically introduce a vaccine against that, by the way, will not cure it because nobody here is in the business of cures. Yeah. Because cures don't make any money.

SPEAKER_14: You would, you would, uh, you want to think that the system does not have

SPEAKER_419: perverse incentives, but like hell today.

SPEAKER_04: There are perverse incentives in every system.

SPEAKER_162: And, um, yeah, they made a lot of money off of the vaccine and it did some good in lowering debts and it didn't actually keep up and I'm not taking any more boosters, but I don't know what my position is with my parents taking boosters.

SPEAKER_71: Maybe I want my parents to take boosters because they're at more at risk than I am.

SPEAKER_14: I definitely do. Yeah. You know, Joan Didion had a great quote at one point about journalism and like you specifically

SPEAKER_04: around sources, uh, the great Joan Didion, uh, and she, this is the quote. My only advantage as a reporter is that I am so physically small, so temperamentally

SPEAKER_14: obtrusive and so neurotically inarticulate that people tend to forget that my presence runs counter to their best interest.

SPEAKER_388: And it always does. That is one last thing to remember. Writers are always selling somebody out.

SPEAKER_50: It's a tough quote. Um, but it's not wrong. It's not wrong. No, no, it's a dirty job. It can be investigative journalism.

SPEAKER_09: Investigative journalism is a very, very cutthroat job.

SPEAKER_428: Yes. Full stop. And they, you know, we'll use a source.

SPEAKER_60: And so like, you know, the, the, I, I don't agree with project full stop. Don't agree with project. Veritas's techniques. And that's what I mean when I, yeah, it could be valid.

SPEAKER_09: Maybe, but I want somebody, I want somebody else to do that reporting. Like I'm never gonna, I mean, if I see project Veritas, blah, blah, blah. I'm always like, no.

SPEAKER_433: That's the problem with it.

SPEAKER_09: Like I automatically don't believe it. That's the problem with being shady is you don't get a benefit of the doubt.

SPEAKER_435: You just don't. Yeah. And yeah, I wonder what's gonna happen with this whole situation.

SPEAKER_303: Speaking of being shady and getting the benefit of the doubt. I love Mr. Beast by just so just, I know Jimmy.

SPEAKER_13: I wasn't talking about Mr. Beast. No, no. I just want to put a pin in that since that was how we got here. Oh, wow.

SPEAKER_14: I just want to say like, I think what he's doing is great. And if he wants to, you know, do philanthropy through his YouTube channel, I think it's awesome.

SPEAKER_373: Yeah. Generally speaking. All right. Yeah.

SPEAKER_53: I mean, you know, I think Mr. Beast is Mr. Beast is America and America is Mr. Beast.

SPEAKER_00: Yeah. There you go. Perfect. There you go. Um, also classic American tale of redemption in progress right here. No. Billy McFarland of Fyre Fest fame is back on Twitter and selling his services. We kind of, we have a little bit of.

SPEAKER_11: Something his services. Yeah. Oh yeah. Oh, he's an event planner. Uh, no, he. All in summit 2023. He'll help. Well, he will help make all in summit go viral. For sure. Yeah. That's his services.

SPEAKER_285: Who's got to start up. All the wrong reasons. Who's got a startup. He tweeted, uh, on Saturday and needs me to blow up the marketing. Blow up is the key word in that sentence. And then on Sunday, he tweeted a whole thread about how he got his first company to go viral.

SPEAKER_00: And then ended the thread with this tweet. Want me to make your viral company go viral DM me. And then there's a link to a page that includes his consulting schedule.

SPEAKER_09: Oh, nice. Where he's charging $600 for a 15 minute interview. Okay. $1,000 for 30 minutes and $1,800 for an hour long consulting call about how to make your startup go

SPEAKER_36: viral so we could have him on the show for 1800 bucks recorded here. Interview him for 1800 bucks.

SPEAKER_455: Talk to him about his life and times and see if it goes viral on YouTube when we make back the 1800 like Mr. Beast.

SPEAKER_162: Maybe, maybe I'm just going to tell every single one of my founders, please don't. Because even if he does give you great advice, I mean, imagine somebody takes this advice and lands them in jail, I mean, be careful folks.

SPEAKER_00: We have one additional side letter for 2023 for all of our deals. And that is, if you hire Billy, we're out. We sell our shares.

SPEAKER_13: Let me tell you a story. Story time. Okay. Story time. Story time. Since we're third rail Mondays, I had a founder who hired somebody who had a colorful background

SPEAKER_14: like McFarland, like high profile, had done like horrible stuff, said terrible stuff that just was super toxic.

SPEAKER_04: And he had met the person at a party. I'm going to make this an amalgamation. So it's not about a specific company, but imagine if you will, you know, you hire somebody like this

SPEAKER_70: and I'm on the board and I'm like, under no circumstances can we be at a startup responsible for this person's redemption because you have shareholders and there is no upside to hiring

SPEAKER_132: this person that could not be achieved through the thousand other people who have this skill who will apply for the job. So if you, as the founder CEO, meet a person, they have this colorful, checkered, disturbing past, and then you bring them into the startup and it blows up in our faces. Anybody who sees this person on our masthead, sees he has an email address from us, makes this connection, it's going to put us in the press for the wrong reasons. You have shareholders and employees, you have an obligation not to do stuff like this. It's not dissimilar to the Palmer lucky stuff, right? Um, alleged Palmer lucky stuff, he denies, whatever.

SPEAKER_248: You know, if you're doing something that could be embarrassing to the company, you're just taking, you're making it, putting everybody at risk. And why would you do that if you don't have to? Anyway, um, the person, you as a board member and investor, you have very little say in this kind of matters on data

SPEAKER_162: to average person, hire the person anyway, and then just engage with them. Um, because they just actually didn't get the work done. Um, luckily it didn't blow up, but yeah, you know, it's like, you don't really want your startup to be the platform in which somebody solves their problem, right? Right. It's not your obligation. And so if you're Billy McFarland's problem is it obviously needs cash. Uh, like everybody needs a job, I guess most people do.

SPEAKER_39: And, uh, if your startup uses his advice, even if it's brilliant advice and then people find out you used his advice, it's just like staying on your company, not worth the risk.

SPEAKER_73: Yeah. Well, our producers, uh, went ahead and came up with some other ways to spend that kind of money.

SPEAKER_09: So for the $600 that you could spend on 15 minutes with Billy, you could eat at the French Laundry with all the food and wine upgrades, about 500 bucks a person. That's good.

SPEAKER_00: Instead of the 1800 for an hour with Billy, uh, you could spend $150 more and sit two rows

SPEAKER_285: back from the floor to watch LeBron versus Katie at the Barclays center tonight. Oh, very nice.

SPEAKER_311: Side note. I had to include this because I'm about to fly to New York. You can get that kind of tickets. You can fly first class to New York, stay at a four star hotel for the same amount of

SPEAKER_482: game in that seat and the warriors and the warriors. Seriously.

SPEAKER_00: Literally, literally, quite literally. Yeah. Like I was like, my son's 16th birthday is coming up. I was like, Ooh, I'm gonna ask Jake out for like some warriors hookups, you know, like to get some good tickets.

SPEAKER_378: And I was like, uh, no, well I could, instead I could get him a car.

SPEAKER_208: Yes. It is a little crazy. Um, one of the tricks in the, if you're a big sports fan is to go to a smaller market,

SPEAKER_04: you know, make a trip out of it, you go to Utah jazz versus warriors sick courtside for,

SPEAKER_15: you know, two grand or three grand instead of 20 grand at Madison square garden or whatever, or at the warriors arena or Lakers. You get to have that course side experience. Yeah. People do that. Uh, they make a trip out of it.

SPEAKER_208: Same thing for concerts too, by the way, you, you know, you're, you're in a major city and you go see some major artists. It's kind of hard to get tickets, scalpers, whatever,

SPEAKER_14: but you can make a trip out of and go to red rocks or go to, you know, I don't know, Canada and be able to scalp tickets at a more reasonable price.

SPEAKER_490: So I'm like, I'm talking about the blue box, Canada, Canada, you know,

SPEAKER_18: Yuba city is the, Yuba city is the weird spot to go see big bands in, in the Bay area. The general it's like near Marysville, which city Yuba, Yuba city, Yuba city.

SPEAKER_496: I don't even know what that is.

SPEAKER_373: Like a lot of bands make a stop in Yuba city.

SPEAKER_497: Oh, why you be at Yuba city.

SPEAKER_18: I went to see a show there once and then the car broke down on the way back.

SPEAKER_09: And then we, I spent the scariest night of my life in a hotel with like, I swear a dead body in a pool. Like the room had been condemned. There was like a guy in a Camaro who was like, oh yeah, go to this Sacramento. Yeah.

SPEAKER_00: It's a hall. You want to take a reliable vehicle, but if you break down, Earl will come in a Camaro

SPEAKER_09: and he will, uh, direct you to the auto body shop and he'll tell them, he'll tell you that you could say that Earl sent you and then you will tell that story

SPEAKER_170: for the rest of your life. Hostile part seven. I can't believe I didn't die. It was amazing. Yeah. Don't die, please. Okay, everybody.

SPEAKER_04: Well, there's a show for today. Third real Mondays. It'll be a tradition here. Um, maybe we'll see you tomorrow.

SPEAKER_14: Oh no. So I mean, it shows, I mean, we might get three strikes and the YouTube channels over now

SPEAKER_06: because we talked about project Veritas and other stuff, but, um, yeah, good luck to us.

SPEAKER_506: Let's see how it goes.

SPEAKER_507: Let's publish. Publish. Boop. See you tomorrow, everybody. Bye.

SPEAKER_506: Bye. Bye. Bye.