SPEAKER_00: All right, everybody. It's Friday. It's a big show today. A lot of developments in the degenerate high movement, generative AI movement. It's moving out of the crazy pace.

SPEAKER_04: Uh, it is not degenerating. Yeah. It is only, uh, it's actually like a fractal right now. It's just, it's exponential.

SPEAKER_00: It's compounding a lot of people are working on it. And I think a lot of people have been working on it quietly are now starting to show their work, show your work moment in time, and a little bit of competition now has happened. We got reports that deep mind is gearing up to release their own beta chat bot to compete with chat GPT. And they're calling it Sparrow and they're addressing a lot of the issues we've been bringing up here around citations and fair use.

SPEAKER_09: Yeah.

SPEAKER_11: And I have some insights from seeing Sam Altman speak at an event last night about how quickly this industry might move. And we'll talk a little bit about the incentives for it to move so quickly, AKA the tens of billions of dollars.

David Friedberg: And then Molly is going to interview one of our launch accelerator classes, United by zero,

SPEAKER_14: which is helping people find more sustainable clothes through a Chrome extension.

David Friedberg: It's very clever tool. You're going to really like it.

SPEAKER_16: And it's great, simple, simple, but genius. And then finally, it's Friday. So, you know, it's time for a really impactful.

SPEAKER_10: Okay. Boomer segment about mental health. This show has everything. A little bit of everything for everybody.

SPEAKER_18: It's going to be a great show. Stick with us.

SPEAKER_20: This Week in Startups is brought to you by Linode. Linode is the leading independent public crowd provider. Apply to their RISE program for founder-led early stage startups and get your first year free. Plus, significant discounts for years two and three at linode.com slash twist. Crowdbotics. Great ideas can change the world. And Crowdbotics is the fastest way to turn those ideas into code. Get a free scoping session for your next big app idea at crowdbotics.com slash twist. And acquire.com, whether you want to sell a solo project or a booming startup with hundreds of employees, acquire.com has the tools, experience, and most importantly, engaged buyers to help you achieve your acquisition goals.

SPEAKER_22: Sign up for free at try.acquire.com slash twist.

SPEAKER_23: All right, everybody. It's Friday. I am exhausted. I think I might be having my second COVID because I woke up today and whoa.

SPEAKER_29: I just, you know that COVID fog? You remember that from two years ago when you get hit with the COVID or maybe a year ago, I got my COVID.

SPEAKER_16: I think it was just a year ago, right? Well, I had it a year ago because it was right after I started.

SPEAKER_31: I got mine right before because I got it at David Sachs' GOP COVID-19 super spreader event.

SPEAKER_33: Right. A holiday special. The David Sachs COVID super spreader.

SPEAKER_11: It's my understanding that that high horse is no longer allowed due to the summit. It's my understanding you have been forced to climb down off of that one because of our gathering in Miami. Anyway, I hope you don't have COVID a second time.

SPEAKER_38: I mean, everybody gets COVID.

SPEAKER_39: Everybody gets to host a COVID event and you get to take that with you forever. Your memories of COVID.

SPEAKER_31: I did a tweet storm about like, are we, how many people believe that COVID is a lab leak today, right? And do you realize it's like 50, high 50s in other studies that 70% of Americans, independent of political party, are now in the camp that, yeah, it's probably was a leak.

SPEAKER_43: Feels like a leak.

SPEAKER_44: I like had a conversation, Nick's like, so we're going right there. Are we a lab leak right off the jump? I think it's okay.

Jason Calacanis: I mean, well, it is, it is interesting how the level of controversy is like draining right out of that. I had a conversation the other day with a friend who was like, uh, something, something, we're not talking about that. And I was like, actually, I sort of feel like there was this big Atlantic piece. Like, it seems to almost be coming around to maybe like consensus that it sort of seems

SPEAKER_45: to, did a story, have almost certainly pro publica did that big investigation.

SPEAKER_49: Yeah. Family fair.

SPEAKER_31: Like it's kind of, and then there was this whole thing where like, I was like, I just thought, well, if I had shared this on social or the podcast, this episode would have been

SPEAKER_54: flagged and or removed. Yeah. If we had said that. So it's like, very interesting that the lab leak, yeah, it still will be flagged.

SPEAKER_55: Oh, okay.

SPEAKER_58: Oh, with the information, which I think is smart.

Jason Calacanis: I mean, look, this is like, nobody ever has perfect information. Right. Right. And when you don't have perfect information and it's the apocalypse, scary, mistakes are going to be made and it's scary.

SPEAKER_59: And, and you have a, you have a population that's in like fear and panic and lashing out at anybody, you know, I mean, they were blocking those. Stories and speculation and whatever, because they were trying to tamp down on anti-Asian hate, which was causing deaths.

Jason Calacanis: Yeah. Right. So it's like, I don't know. It's yeah. Did people screw stuff up?

SPEAKER_61: Mm-hmm.

Jason Calacanis: For sure. Are they going to again?

SPEAKER_61: Yep.

SPEAKER_62: I just think it's like an interesting thing to take with you going forward, right?

SPEAKER_31: Like, what do you take forward from it? And I'm like, in a panic.

SPEAKER_39: Like, yeah, things can get weird on the information front. Right.

SPEAKER_43: And this, I try to think of the first time in our lifetimes as Gen Xers, you know, being

David Friedberg: on the planet for 50 years or so, like, were we ever told you can't talk about something or that was like off limits. And it was the first time I can remember people saying like, yeah, you can't bring

SPEAKER_31: that up. So why not? Because there was an episode I had Balaji.

SPEAKER_71: I think there were not.

SPEAKER_35: There was no social media before, but certainly there, everybody has grown up with like the

Jason Calacanis: things you don't say, the topics you don't talk about. It was a little more unspoken because you didn't have things taken off of the internet because you didn't have the internet, but there were stories you didn't publish. I mean, there's always been stuff you don't say. Like, come on.

SPEAKER_31: Well, I mean, I just, I, yeah, social media is new, but the act of being censored. The act of having stuff taken down, right? Like I'd never had a blog post where they were like, take this blog post down because you discussed something. And then we had an episode. I want to pull the clip because I remember bringing it up and I was thinking about myself

SPEAKER_26: as in, you have the Balaji clip?

SPEAKER_77: I have. It's, it's a little long. It's like four minutes long, but he's. I don't want to play right now. I'm going to link to it in the show. You're asking him. He's like, I just want to say that I, this is not my opinion. I am just talking in factual evidence. This is not my opinion. Like, he's like really nervous. It's pretty crazy.

SPEAKER_31: Right. That's what I remember, I remember asking, this is an interviewer trying to be like a fearless interviewer. And I was like, so what do you think of the lab leak theory? And he, and this is like the Balaji who is like, he's out there. He's like, yeah, we should all vote on the blockchain. And like, there should be no government. And the, you know, here's my new plan for government. Like, kind of like a fearless, you know, you know, edge of, you know, uh, Overton window thinker, right? Outside the wide Overton thinker and he was, I think, visibly scared if my memory serves me correct, of actually addressing this on a tape podcast at that moment in time. Cause what, would you remember what month that was, Nick? It was, was that like month two or three? Yeah. It was month three of it. Hmm.

SPEAKER_52: Um, yeah, really interesting. Yeah. Um, I really would like to get to the bottom of that.

SPEAKER_84: It's kind of our, like, I mean, it is interesting.

Jason Calacanis: Like you're right that in our, the closest corollary I can come up with. And again, social media didn't exist. So it's like hard to, but when I was in high school, I did my, my mom got heavily

SPEAKER_59: involved in, then I got heavily involved in doing HIV AIDS education. So I would travel around. I went to native American reservations and schools and did this kind of like. Safe sex, you know, here's how HIV and AIDS actually. And that was sort of, that was like, right when it was still massively stigmatized and also had elements of birth control to it.

Jason Calacanis: Right. I mean, we were talking about condom use and saying like, no, it is not a thing that only gay people get. Right. And there were, it's, it's nowhere near the kind of the fear that people felt for being like canceled or disappeared from Twitter or whatever. Right. Right.

SPEAKER_11: But however, that was a very, there were very clearly like lines that certain places

SPEAKER_90: didn't want to cross or that they didn't want to talk specifically, I got gay people or. They didn't want to talk about AIDS at all. Yeah.

SPEAKER_58: Because it was like, well, if you did that, you sinned or, you know, and there were like, I mean, I grew up in a.

Jason Calacanis: In a town where they would like throw you out of school for bringing condoms. Like there's an HIV AIDS, I think is probably one of the times that that happened.

SPEAKER_59: But then you almost have to go all the way back to like McCarthyism to get that same kind of fear of like saying the wrong thing. Yes. And having your career be over as a result of it. But it really was not until magic Johnson came out and said, he said like before that, I mean, Freddie Mercury would never admit that he had AIDS. Like, or he was gay or that he was gay, exactly. And I think never did. But like the AIDS crisis is the last time in American culture I can think of where there was that level of like secrecy, not a conspiracy theory.

SPEAKER_31: We're not talking about conspiracy theories. Like that's a different thing, but just generally the public did not want to address a topic.

SPEAKER_97: And if you brought it up, you know, beyond uncomfortable, you would censor yourself over

SPEAKER_99: it, censor yourself. Yeah. To, to talk about the issue.

SPEAKER_102: Yeah. Um, yeah, I just like, I really feel like we should get to the bottom of it. In a non-political way. I hope we can. Yes. Like we do need to know this.

SPEAKER_31: We need to know what happened, right? Like, is it such a big deal to just want to know what happened? Like just for the future and for like, why did we go through all that? Like, and if we're responsible, like if we funded it, I, I don't necessarily have a problem with that. I'm, I, I don't have a position on like, was gain of function research, a smart thing to do or not. Like, I, I, it's kind of above my pay grade. It sounds to me like it's a dangerous thing to do.

SPEAKER_39: And to do it in a city. So, I mean, you can be Monday morning quarterback all day long. It seems like maybe you should do it on an island somewhere.

SPEAKER_05: I don't know. Like maybe not in a major city. Like with this guy named Dr. Moreau, like you would do it over there.

SPEAKER_31: I mean, literally like doing it on an island is, it seems like.

SPEAKER_105: Pretty obvious idea or, you know, uh, yeah, really.

SPEAKER_11: Kind of come out someday, but we will be probably pretty old by then. Right? Like it's going to be a long time before you can take the emotion out of that conversation.

SPEAKER_109: Sometimes we get a partner here that has an offer. That's so amazing. I barely need to read you the ad copy. I just read you the benefits. Linode has a startup program. It's called rise R I S E, and it offers more than just free credits. Startups get up to $10,000 per month in year one credits, followed by a 50% and 25% discount in each of the next two years. And there are no caps and you get a lifetime discount based on your usage and you get free 24, seven, three 65 award-winning customer support by the phone, email, or social media. There are no tears. There are no handoffs and you get cloud consulting experts to ensure your tech stack scales seamlessly. Plus community connect with other program members, alumni, advisors, and more. And shout out to our friends at Linode, which was acquired by Akamai last year. Congratulations to the team with Akamai plus Linode. You also get access to leading security and CD and solution. So you're going to be snappy, snappy all around the globe. If you're a class sensitive, but you want amazing hosting, the solution for you is Linode. That's the answer, right? You want to watch that bill. You want to watch that bottom line, but you also want top tier service and speed. So visit Linode.com slash twist and you will get $500 in free credits and you can apply for their startup program. Rise rise members receive up to $120,000 in free infrastructure credits during their first year and up to 50% off in the years after that.

SPEAKER_31: And speaking of taking the emotion out of it. You know, they're the national archives have been releasing the JFK assassination papers three, four or five times now. And the CIA is like supposed to be under mandate over like multiple presidents, multiple, you know, Congresses or whatever to release this information. And they keep saying they're releasing it and 97% or 98% is released. And it's like, yeah, we know, can we get the final 2% is one, what we want, that's kind of the stuff. Yeah, it's kind of the stuff. And they keep redacting it. So, uh, welcome to this week in conspiracy theories, but I do think it's relevant, um, that we talk about this. Like we definitely need to figure out what's good. What are they trying to hide in the JFK assassination stuff? I w I was like literally looking at both sets of media last night, uh, two nights ago.

SPEAKER_113: Yeah.

SPEAKER_54: Democratic media, you know, MSNBC, Fox, everybody in between is like, yeah, like meet the press. Everybody's been talking about this very quietly though. Right. Um, why can't we just see this? Like we're, we're adults here and it's, I guess maybe there's people who are still alive would be impacted by it. Or maybe they did it. Yeah. I don't think they killed the president, but it sounds like maybe a nine 11 situation where like, you know, they were taking flying lessons in Florida and we were watching them and, you know, we lost track. Oh, like maybe they knew. Right. I mean, they screwed. I think that's more, it's like a coverup of like embarrassment. We screwed up. We could have prevented it. Like, what if there's a document that says like, we're gonna kill the president at this location. Yeah. And the person's like, yeah, I don't think this is credible. And then he dies, you know, or more.

SPEAKER_59: Right. Or that there are, there's evidence. There's probably, don't you think there must be evidence of co-conspirators? Like it was not just alone. Yeah.

SPEAKER_00: Like no way, which is what I think the issue is with the.

SPEAKER_124: Which I think is the potential issue with the COVID stuff is like, we're co-conspirators in such, in a sense that we probably funded it.

Jason Calacanis: I just heard a story the other day that said that they had been talking about doing a COVID commission that was kind of like the 9-11 commission that has just like quietly died. And I certainly hope that that's not the case.

SPEAKER_59: I mean, we really like America's owed explanations for the fact that we have the highest death rate per capita of any developed country in the world. We did a bad job. We are owed the fact that Jared Kushner refused to talk about pandemic planning and throughout the book. And, you know what I mean?

SPEAKER_10: Like that stuff really happened and we're owed some conversations about PPE, like. Yeah, let's have a commission. And why would we not have a commission? Where they may.

SPEAKER_43: I'm so sick of the partisanship. It's like the partisanship is so weird. It's like, okay, the COVID happened.

SPEAKER_31: Kids got impacted by it. People died. We had a suboptimal response. What's actually going on here?

Jason Calacanis: You know, like we'll never be the same. Like, we can probably take a look back at this, but you know, I mean, it's, I think it's still too. It's like still too hot.

SPEAKER_31: Yeah. I feel like this is the year though, where like, maybe we'll kind of have enough distance from it that we'll say, you know, this really did screw up the economy. This really did screw up people's psychology, you know, psyches this, you know, people suffered education suffered, you know, which we'll see all the second and third order effects manifest themselves. And then it'll be like, you know what? Let's make sure this doesn't happen again. Let's figure out why, you know, this all happened. So I hope so. I would hope so. That is my hope. Get it together.

SPEAKER_134: Or we could just ask chat GPT. And then tell us what happened. You're right. You know what? Yeah. Yeah. Yeah.

SPEAKER_137: Yeah. Uh, you.

SPEAKER_31: Dear chat GPT. But there's all COVID stuff comes up, uh, because just in the pregame for the show, Molly left her house. Yeah. And you went to the city. It's over now.

SPEAKER_10: I definitely, I'm doomed. And I feel like I might have COVID for a second time. Two times this week, almost three. I was supposed to go out on Tuesday, but I had like too much storm damage situation.

Jason Calacanis: But yeah, I have left the house two whole times.

SPEAKER_11: And last night I went to an event, um, called Strictly BC, which is the name of a newsletter and podcast hosted by the TechCrunch senior editor. Yeah. Um, it was interesting.

SPEAKER_36: Something. Uh, Loisos. Yeah.

SPEAKER_147: I think I met her.

SPEAKER_10: Yeah. She's good. Good journalist. Great shoes.

Jason Calacanis: Um, she interviewed Alfred Lynn from Sequoia. Oh, great. Some kind of testy questions about SPF and the whole, as you might imagine a little bit awkward and, uh, also Sam Altman talking about open AI.

SPEAKER_10: Um, both of them at the same time or one, then the other, no one, then the one. And then like a panel of these poor fintech founders who, you know, who was, everybody was like, is Sam coming on now?

SPEAKER_93: When is Sam going to come?

SPEAKER_154: Oh, that's yeah. It's really rough. Yeah. It's like opening up for the Beatles. Yeah. Kind of get off the stage.

Jason Calacanis: It was brutal. I felt really bad for them. Um, yes. And then Sam Altman came out and was interviewed and talked a little bit about chat GPT.

SPEAKER_11: Oh, great. And.

SPEAKER_136: Any interesting takeaways?

Jason Calacanis: So a couple that, well, you know, it was a very, it was like kind of a surreal as some, as San Francisco can sometimes feel, it feels like a sci-fi in action. So it was like a weird, surreal collision between kind of the human world and this description of this like future AI universe, including like, I had to actually step out while he was speaking.

SPEAKER_59: Because my mom called because she was getting out of the hospital, like it was, you know, but I, my sense is that while I was gone, there was some pushing on the question of the financial model of open AI that, and that those questions weren't really answered. So then the takeaways mostly happened during the QA, which I was, where I was there, he said a couple of things of note. One that they're going to move fairly slowly in terms of launching GPT four. Cause remember chat GPT is based on GPT, which is the, the framework, the neural framework for deep learning and chat GPT is based on GPT three, which is kind of the current model.

SPEAKER_158: And four is expected to be, you know, orders of magnitude smarter.

SPEAKER_39: The amount of connections or chunks of data. I'm not sure there's some word they use for, you know, the corpus and the corpus is a magnitude bigger.

SPEAKER_31: And people have been making visualizations of that magnitude where it's like, here's a grain of sand and here's the planet earth, you know, like kind of situation. Right. I don't know that that makes the output that much more dramatic, but should make it more comprehensive, I guess.

David Friedberg: So then maybe the percentage of times you get a hit parameters. Yeah.

SPEAKER_163: Maybe, uh, um, he, someone asked, he, he said in general, we're going to release technology much more slowly.

SPEAKER_59: And people would like, we're going to sit on it for much longer. He was cautious, you know, talking about being cautious in the implementation. He said, I, I heard him say, they taught, you know, he was of course asked about its use in schools and for learning and whether there could be a site, your sources model built in.

SPEAKER_43: Which we'll talk about that was what I've been harping on is like, what are the sources here?

SPEAKER_59: Yeah. It was interesting because he basically said, you know, like, of course, we're going to talk about things like watermarking and source citing and all of that. But he said, you know, I would still be cautious about any policy that says you can or can't use this because nothing is going to be perfect. Like people are going to be, figure out how to cheat with it. Like it exists. And sure it's more important to adapt learning to it than adapt it to education.

SPEAKER_133: Oh, wait. Oh, that's interesting.

SPEAKER_168: Yeah. I'm paraphrasing his vibe. Yeah. Yeah.

SPEAKER_39: No, but as a concept, it's, it just reminds me of the calculator discussion. My understanding is totally you're allowed to bring calculators. Yeah. How?

SPEAKER_10: Yeah. Not in the SATs. I had to go buy like a Texas TI, not in the SATs.

SPEAKER_172: Do you have a great idea, but you don't have a technical co-founder? Well, boom.

SPEAKER_109: Crowdbotics is your CTO as a service. You know how it is. A lot of times these amazing ideas you have, well, they fail in the planning phase. Crowdbotics will make sure that doesn't happen to you. They offer pre-built app templates to help you build your MVP, your first version of your product faster than anybody else can. This means you can stop building from scratch and start using the same architecture that all the industry leaders do. There's no reason to reinvent the wheels. Not when you have a great partner like Crowdbotics. You can go from idea to spec and spec to code super easily because they do this for a living. And if you're not sure where to start, well, Crowdbotics offers professional scoping. This helps you flesh out your project at the MVP stage and beyond. And here's why Crowdbotics really gets this done in an awesome fashion. They have automated DevOps, maintenance, security updates, and app store publishing. You don't have to worry about all that blocking and tackling. They do it right, right from the start. And the best part is you own the source code. You own the IP. So, let the folks at Crowdbotics show you how all this works. You can schedule a free scoping session and get your detailed build plan done. Buy Crowdbotics at crowdbotics.com slash twist. That's crowdbotics.com slash twist. Again, it's free. You got nothing to lose and you got everything to gain. Crowdbotics.com slash twist.

SPEAKER_111: No. No.

David Friedberg: So, this is, I think, going to be like a groundswell. I have an idea. I don't think, I think there should be two types of writing tests from now on. One should be, you write something without the tools and you get graded on your ability to write without steroids and then you get graded with your ability using Grammarly, ChatGPT, et cetera.

SPEAKER_39: Because you'd still need, we still need to know that people can write. Right? Or not.

Jason Calacanis: I don't know. Right. I mean, now they're using it. There was that kind of pretty heartwarming story about how they're using it for like a guy had started a business. He's a recent immigrant.

SPEAKER_11: He has English as a second language and is dyslexic on top of that. Right. And so his communication skills with customers are terrible, but it's a one man business. And so he started using ChatGPT to write customer emails.

SPEAKER_36: Oh, that's heartwarming. It's unbelievable.

SPEAKER_59: Like I just was, whoa, because now it's like he can actually win business. Okay.

Jason Calacanis: So we may find over time that you have to learn different things. One thing I did think was really interesting is that Connie, when she interviewed Alfred Lynn, asked him about AI and whether Sequoia is bullish on AI.

SPEAKER_59: And he said, one thing I want you to ask Sam Altman is when do you think, you know, GPT or ChatGPT could win the like international mathematics Olympiad versus being able to repackage and regurgitate information based on what already exists. Like he seemed somewhat skeptical of the thinking capability that we hear so much about as opposed to the ability. Well, it is true that what we're seeing really is like a cool UI on a big, huge database, the big, huge database being the store of all human knowledge, but what ChatGPT really is doing is just like presenting it, repackaging and presenting, which is not the same as thinking.

SPEAKER_39: Thank God I have, yeah, so I'm thinking about my own biases here as a writer by trade and thinking I'm a much better writer than, you know, 98% of people on the planet or 99% of people on the planet.

SPEAKER_23: Uh, I wonder if my bias towards like, Hey, here's a blank piece of paper, write something compelling is me wanting to protect my franchise and saying like, Hey, you know what?

SPEAKER_31: Like, I don't want you to be able to use steroids and compete with me.

SPEAKER_43: But the fact is, you know, the ability to write a coherent blog post now is gotten commoditized by this kind of technology. Right.

SPEAKER_32: And the difference between creating and packaging is the next leap, right?

Jason Calacanis: Like explain what that means, creating and packaging. So you can take existing information.

SPEAKER_59: Like it emerged this week that CNET has been having AI write blog posts for the money for part of its site for what? I don't know that it was even presented as a scandal, but it's, but it is interesting, right? It's like the story is say by CNET staff.

SPEAKER_11: And when you click through, it's like an AI generated the bulk of this. And then a human editor came in and edited it and made sure that it was coherent and then sent it out.

SPEAKER_43: Oh, wow. CNET is quietly publishing entire articles generated by AI. Yeah. Oh, that is gross.

Jason Calacanis: That you can totally do. Well, is it gross or is it efficient? Like, do you really need a person to go and be like, I Googled all of the credit cards that are available and I found their interest rates. And I put them in a list and I feel like, why wouldn't you have an AI do that base level work? And then you go in and you, I mean, it's exactly what you've been saying about the producer, but then you have a human go in and add and make it sound like a human wrote it.

SPEAKER_31: I feel like I'm a rock dad now because like, for me, uh, when I hear, you know, some modern music and I'm like, that's not the person's actual voice.

SPEAKER_39: That person doesn't know how to play an instrument. Right.

SPEAKER_31: I do not get pleasure from the performance and then I'll watch a Mark Knopfler, Dire Straits, you know, you know, Alchemy concert or Pink Floyd, you know, concert. And, and I will go onto YouTube. I'm turning into a rock dad and I look for a live performance and I want to see the live performance because even the studio albums by Dire Straits or Pink Floyd, which are extraordinary.

SPEAKER_204: That is art. That is art.

Jason Calacanis: Yes. We're talking about aggregation. We, you've literally been saying on this show that it is like, yeah, the producers should just have Chad GPT find the average salaries of every CEO.

SPEAKER_208: They're not replacing us here while we talk.

Jason Calacanis: No. And we're still not replacing our producers because we're still going to ask them to package that. No, no, no.

SPEAKER_160: To give that. What? I don't know. Producers. I don't know. Somebody just have Brian take a day off and then have Chad GPT reply.

SPEAKER_168: What I'm saying. Let's see if we can tell the difference. This is the same conversation as, as academic. We should not.

SPEAKER_10: That sounds right. Shots fired. We should not pay people to Google stuff anymore. That's stupid.

SPEAKER_212: We have computers to Google stuff now. Brian GPT. Yeah. I, it's that. I just, I, I'm giving you my reaction. I see that you're offended.

SPEAKER_215: I know, but no. I'm offended by CNET doing that. I feel like I will not go to CNET. Oh, come on.

SPEAKER_31: I know if I, whenever I see a non byline story, I'm already like, eh, but this just feels to me like, uh, icky.

Jason Calacanis: I just wish somebody would, we're just gonna stratify. We're gonna, there's gonna be the stuff that, that computers should do because it doesn't make sense to pay a human to do them.

SPEAKER_35: And then there's stuff that, and when it comes to creating, when it comes to generating ideas, when it comes to the math Olympiad, right? Like all of that is still the terrain of humans and will remain so creative writing is still ours.

SPEAKER_219: But for God's sake, like putting together a list of credit cards and their rates. Yeah, no, no.

SPEAKER_31: I get that. Have a computer do that. Yeah. I mean, I don't mind like a little web research. Yeah. I mean, the, the, the other conclusion I've come to is I'm coming to a lot of conclusions about this technology. Mm hmm.

SPEAKER_222: Now, um, citations, hypotheses.

David Friedberg: I, for me, their conclusions, um, I think these things are stealing people's content. Um, and I think it's unfair, you know, I'm, I'm past the wow.

SPEAKER_31: And I'm now I'm to the, like, Hmm. And it feels incredibly unfair to rewrite with this technology, other people's source material. And I think the courts are going to side with me that this in the four part test for fair use, if you are using people's work, there is fair use for it.

David Friedberg: And one of the tests, if you're breaking fair use, and this is a qualitative test that a judge has to make a decision on in a jury, uh, depending on the jurisdiction.

SPEAKER_31: And I think that the, the test that's going to really fail on is that it's going to, uh, not allow the original authors. To generate revenue, to do commerce, um, from their work and, and, and future work. This is going to take away future work. So if somebody wrote a history of, I don't know, presidents and it's been digested by chat GPT and it's regurgitating that that person who spent their life's work writing about, you know, a thousand, uh, 10,000 words on every president.

SPEAKER_54: It's impeding their ability to do work in the world. And, uh, it's going to be pretty easy to figure out where they got this information from and the lawsuits are going to come fast and furious. And there needs to be, um, a really fair model here. And I think citations and payment is going to, and royalties are going to have to be put into this, just like the sampling of songs, you know, um, there became like a policy around this. I think that's going to be the thing that's going to put a lot of cold water on this real quick.

Jason Calacanis: So related to this, and then I do want to go back to one other impression from hearing Sam speak last night, but related to this, there was another piece of information yesterday from a time magazine article about Google's deep mind. Um, Demis Hassabis, deep mind, CEO and co-founder spoke to time and said that to keep pace with open AI deep minds, they are considering releasing their own chat bot called Sparrow for a private beta sometime in 2023.

SPEAKER_11: Super cute name, by the way. I love Sparrow. Sparrow's a great name. Sparrow's great name. But deep mind is delaying its launch to work on reinforcing learning based features that chat GPT lacks, such as citing its sources.

SPEAKER_59: And hasaba said it is right to be cautious on that front. So they might actually be slowing it down, not just because they're afraid it'll be too smart too soon, but because they are agree with you.

SPEAKER_36: Right. Or they need to address this, the sourcing. And we have some screenshots of some of the citations. If you want to see what they look like.

SPEAKER_232: Oh, really? This is screenshots from the Sparrow app?

SPEAKER_233: Yes. Yeah. It's a deep mind blog. Stay with us, Jason.

SPEAKER_31: How come I didn't know? Oh, right. I got COVID. I'm sick again.

SPEAKER_234: I think I might have COVID folks. It's in the notes.

SPEAKER_31: Anyway. I know it's in the notes, but I'm acting surprised here that this actually came out. Oh, yeah. This looks exactly like the chat GPT.

SPEAKER_235: IOS app that I shared with everybody when I was asking what are the best restaurants in Yonfelen who has the best.

SPEAKER_236: But it would.

SPEAKER_237: Hold up one more time. I'm sorry. Hold this up. A little slow here on the draw. Thank you producers.

SPEAKER_238: Uh, they work on. When did the current expedition to the ISS International Space Station launch is the question asked by a user inspired says on March 30th, 2022 with NASA astronaut Thomas Marshburn.

SPEAKER_240: Taking over as ISS commander and then.

SPEAKER_137: Oh, look at that. And then a full citation from Wikipedia.

SPEAKER_31: So it's the rewritten version of that. Mm hmm. So they made it as short and tight as possible. Brevity. Mm hmm. Fact based.

SPEAKER_14: Perfect grammar. Giving you what you want. And it just pulled it from the Wikipedia and rewrote it. And that's what's happening here.

SPEAKER_77: It includes a. If you look at the top right of the speech bubble, it includes a little link. It looks like it's a one click to touch link as well.

SPEAKER_245: Okay.

SPEAKER_31: So that to me feels like fair use.

David Friedberg: Uh, if it's non-commercial, if that puts an ad in there, um, and I don't have to go visit that website.

SPEAKER_10: Uh, so now like, but I gotta ask you, how do you feel about Google and Google snippets?

Jason Calacanis: I mean, seriously, like never, I've never liked it. Just a presentation layer for like the other side of this is isn't chat GPT just a presentation layer. Okay.

SPEAKER_249: For publicly available information. They're not stealing the information. Okay.

SPEAKER_251: No, no, actually I can answer that. Um, this is a Wikipedia page. Let's replace it with a Yelp page. Yeah.

SPEAKER_31: Uh, and let's say what are the three best restaurants? Uh, you know, in Yonville, uh, for done and it uses that corpus. And then it says the three best is this, and then it links to Yelp.

SPEAKER_254: Um, and, uh, I feel like it's worse of it.

Jason Calacanis: I think it's worse if it links to Yelp, because if it's giving you the three best restaurants in Yonville, it's pull, it's pulling from a corpus of information about restaurants in Yonville. That includes Yelp and Google reviews and everything everybody's ever said and Michelin and right. It's like, it is just condensing and presenting the consensus of the internet on this topic. Yeah.

SPEAKER_54: And the reason this would be illegal and a judge would find it illegal in my mind is my opinion. Okay. Uh, I think a judge will find this to be illegal and will acquire permission, uh, because it infringes on the ability of Yelp to do commerce in the future, uh, based on the work. Based on the work and their work product, they own all that content on their website. Now Google has taken their work product, created a better version of it, a derivative work, clearly a derivative work. And it is impeding Yelp's ability to do, um, commerce because you do no longer need to visit Yelp. Uh, now the link is a nice courtesy. But if that link does not get clicked, which I don't think it will, if you told me, you know, Bouchon has the best duck at French Laundry and I don't go. To Yelp, you have just impeded their ability to do commerce in the world. And that's when the unfairness in fair use gets triggered in a content owner, uh, in, in a creatives part, and then they get a lawyer. And so that's why I think this is slightly different. And if they do it from three sources, you know, like if, if I wrote a meta review, um, like Metacritic does, a lot of people had a problem with Metacritic. Um, but Metacritic is very careful. If we pull up, pull up a Metacritic page for tar, uh, and then, uh, queue up, uh, um, a rotten tomatoes. Cause that's the equivalent of this Molly is this like summation, right? And what you'll find is Metacritic and, um, rotten tomatoes.

SPEAKER_31: If you pull up the tar page, when you start scrolling down here and you see the, the capsules, they have been extremely delicate here. Uh, let's go down. There it is. Uh, they, and I think they work with these publications. I don't know if they have permission, but I think they're working with them. Look at how little they do. Look at how they put the logo of the company. Mm-hmm. They keep it to a sentence or two and like sin view. It's that third one down with a hundred. It says this film will draw you in and demand a second viewing. They are very, very careful here.

SPEAKER_262: The percentage of that sin view review is very short.

SPEAKER_14: Um, and the ones on the right are obviously the user reviews. They have full rights to those.

SPEAKER_31: This, you know, is very close to what chat GPT is going to be doing across large corpus of users, but you don't see the logo on chat GPT and it's rewriting it. It feels like if they're rewriting it. So the equivalent would be if I wrote my review and I took those five insights, rewrote those five insights and presented it as my review. Now this hasn't never happened at scale. It's happened with content farms on the margin, you know, cribbing people's opinions and hot takes. But it's, it hasn't happened at scale as a product that would annihilate ever needing to go to the source material. Yeah. So if we did, what are the five? Uh, how about this? Write a review of if we do a chat GPT right now, write a review of the movie tar in the voice of an elite movie critic. And we compare that to like those snippets. That would be the intellectual equivalent to me, but then it gets presented as chat GPT's work product, but it was based on all those links.

SPEAKER_199: Feels very unfair. Yeah.

SPEAKER_59: I understand what you're saying. I, I, you're probably right that there will be a bunch of lawsuits. I think like this is a, I think this is just a, is, but a part of this conversation.

SPEAKER_11: It's unprecedented. I think it's unprecedented.

Jason Calacanis: I think that they, I think given how many times Google has won lawsuits, how Google has sort of kept Yelp from being able to, you know, even make a dent. And it's, I think that the ability to say chat GPT is not a, what they're saying so far, and this will change when things become products. What they're saying so far is like, this is not a product chat GPT is not presenting you anything other than an answer that it got from the corpus of human knowledge.

SPEAKER_59: So like, if you asked me, uh, to answer a question and I just was like, oh, I know this cause it's all in my brain. And I gave you the answer. Hmm. I wouldn't be stealing that answer from like, if I was like, oh, well I read. Yeah. Alexander Hamilton biography. And I gave you the answer from memory. You wouldn't be like, you stole that from the, you're putting the, I mean, granted, I'm not doing it at scale, but okay.

SPEAKER_31: That's when it becomes, when it's productized is the issue. And you know, Google has gotten away with it on the margins here in the US. They've really gotten close to the line, but then in the EU and Australia and Germany, we've seen action get taken against them to not, you know, favor their own services.

SPEAKER_14: This is just wholesale replacing services as presented today.

SPEAKER_109: Micro acquire is a startup acquisition marketplace that helps you sell your business quickly and easily online. The acquisition process was never described as quick and easy before micro acquire, and they changed that. And here's what you need to know in 2023. Micro acquire is rebranding themselves to acquire.com. Oh, you know, I love a good domain name. Congratulations to the team over there. Great branding. They want to show the world that they can help any startup of any size get acquired. They want to help founders achieve life changing outcomes. And they want to build tools that make acquisitions easy. They also want to foster a new generation of entrepreneurs. The stats on acquire speak for themselves. An average 12 month revenue for startups listed is now almost $600,000. There are over 35,000 messages between buyers and sellers in any given month. This is a marketplace that is highly active. And now you can sell anything from a solo project to a booming company with hundreds of employees. Acquire.com has the tools experience and most importantly engaged buyers to help you achieve your acquisition goals. And if you're on the buy side, you can join over 120,000 buyers with skin in the game. Buyers can browse listings for free. And of course, it costs nothing for you to list or sell your business on acquire.com. If you're thinking about selling your startup or looking to acquire a business in 2023 and beyond, sign up for your free acquire.com account now. Get more info at try.acquire.com slash twist.

SPEAKER_10: Well, it's yeah. So, okay.

SPEAKER_59: That actually gets into the other thing that I think is interesting about this conversation is like, we've talked a little bit about how, I mean, let's go back to like what open AI was founded for and primarily funded, I think by Elon Musk, right?

SPEAKER_35: Like it, he gave a donation to them or he gave a donation to make it open source came into existence to be an open source project to keep AI from killing us. Right. It was supposed to be for the overall good of humanity. Now we've talked about how all of a sudden it's got a massive valuation.

Jason Calacanis: All of a sudden it's gonna power Bing and get productized and maybe suit out of existence.

SPEAKER_59: All of a sudden it's gonna force all these other products to come into existence. And it's being, uh, it's, it's now set up as this, uh, it's a capped. What's the word? I'm capped revenue company, but it's capped at $90 billion. It's very sure. Like the point at which they have to start, you know, I don't know, putting back money into the public coffers or open sourcing this or whatever is after it's like one of the most valuable companies. Well, I think you make cap in history, cap profits. Thank you. I could not think of that word at all. It's capped profits and the profits are capped at $90 billion, which is an insane amount of profit. So then Sam Altman is asked last night about what is the best case and what is the worst case for this technology? Hmm. And you know, he's like, look, the best cases are so good that you sound like a crazy person if you try to enumerate them, which is of course what anybody in tech says about anything.

Jason Calacanis: Yeah. And then he's like, in worst case, and I wrote this down, he goes, the bad case. And I think it's important to say is like lights out for all of us. Okay. And I was like, well, wait a minute. Wasn't this thing started to make sure it wouldn't be lights out for all of us. And now you're saying like, we're going to go ahead and make our $90 billion.

SPEAKER_59: And then, and by the way, he was asked like what it would look like for them to try to protect humanity after the $90 billion is made.

SPEAKER_36: And he's like, well, we're worried about that when we get to it basically.

SPEAKER_26: There's a very, this discussion came up on all in, uh, which just got published this morning.

SPEAKER_14: Uh, at the same time we're publishing this, uh, and, and Friedberg brought up this blog post, which from 2015, this is the opening blog post of open AI. And if you just look at that first paragraph, open AI is a nonprofit artificial intelligence research company.

SPEAKER_39: Our goal is to advance digital intelligence in the way it's most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. That's what I'm saying.

SPEAKER_170: Since our research is free from financial obligations, we can better focus on positive human impact. See. I haven't taken this blog post down and I, I guess.

SPEAKER_137: Once again, I am exactly as smart as David Friedberg.

SPEAKER_291: Wow. Yeah, of course. Um, Smarter.

SPEAKER_31: You were gonna say, yeah. IQ or EQ. I mean, these are both intelligence things. Like he might, I mean, he might have some IQ points on both of us when it comes to folding proteins.

SPEAKER_26: And then on EQ, he might be. For sure. Um, anyway, he might say that himself. But anyway, right?

SPEAKER_296: Like all of that seems to be gone now. And he's literally answering questions being like, yeah, it could probably kill us all.

SPEAKER_206: So we're gonna go slow on GPT-4, but we're still gonna do it. Cause 90 billion dollars.

David Friedberg: I have been researching the other products. Somebody came at me on an SEO person came at me on the Twitter.

SPEAKER_31: And was like, Jake house, take blah, blah, blah, blah, blah. Uh, you know, all these startups and venture capitalists are going to back companies that are dependent on chat GPT.

SPEAKER_14: And then they're all gonna get rug pulled by chat GPT. Just like Farmville got rug pulled and all the Facebook ones did. Right.

SPEAKER_54: And I thought about that for a second. I said, actually, I actually think the other open source projects. Cause Facebook put their, uh, pie torch. Is an open source project that I think came out of, or it was primarily driven by Facebook's work. There's so many open source projects. The data sources are going to be blocked by chat GPT. Now that they're a for profit company with a $29 billion valuation, anybody who let them use their data in that last one is now gonna block it or ask for money. And they're gonna ask for all kinds of citations and conditions. If they used Quora data, if they used Wikipedia data. Yeah. You know, Wikipedia is under creative commons, but you know, it doesn't mean Wikipedia can't change, uh, their position on chat GPT using their data. It just all kinds of folks, Craigslist, Twitter, Facebook, you know, I don't know what's in the chat GPT mix. I assume Quora, Facebook are too smart to allow chat GPT to crawl them. And that chat GPT is too smart to take their data without permission.

SPEAKER_308: But there are a series of companies. Uh, Craigslist, remember Craigslist and Craig Newmark. Shout out Craig. He wouldn't let anybody crawl Craigslist.

SPEAKER_311: You couldn't make an app based on Craigslist. I think you still can't. You still can't.

SPEAKER_54: No, it's like, this is our data. Mm-hmm. Quora. He got it. Smart cats, right? He's, you're not taking the Quora data. Yeah. Uh, Facebook, Instagram, LinkedIn, LinkedIn fought against the scraping. Uh, and that Israeli company, I think won the ability to scrape a certain amount of web data. Like these companies are savvy. Yep. They have AI people and there are open source projects. So the ability to answer questions is best represented by probably the data on Quora. Uh, in a structured fashion. What is the data on Facebook, Twitter, uh, and LinkedIn worth as a corpus of data?

SPEAKER_251: ChatGPT is going to lose the rights to use any of that data. Yeah.

SPEAKER_31: So ChatGPT's training data is gonna get demolished and then proprietary will happen or they're gonna have to license it. So I think that the public might point to the person who was kind of dunking on me.

David Friedberg: And I was like, well, I think this gets commoditized real quick.

SPEAKER_319: I think this is gonna be a, a quick race to commoditization. Yes. Just like cloud computing is commoditized. 100%.

Jason Calacanis: And he seemed to be saying that last night too, Samuel. Oh, did he? Okay. So he realizes that. For me, this is, it's so interesting. Cause it's like, it's the precipice of change moment, but also I think almost immediately the bubble pot moment. Like don't, you know, cause then he, then some, he of course said, it's a great time to start a startup and whatever.

SPEAKER_11: And some, you know, young guy was like, well, my friend wants to start a company. What kind of company should he start? And Sam Altman said an AI company. And I was like, no, no, no, no, no. I actually went up to him after.

Jason Calacanis: And I was like, start a climate tech company like this. You're not going to be able to access the data. You're not going to be able to outcompete Facebook. I think meta Google, like the corpus of data that GPT already contains. Like, this is not a startup.

SPEAKER_59: Like maybe you can become a layer on top of it, but I think a lot of people could lose a lot of money, like running into the space before we know.

Jason Calacanis: Whether the lawsuits have been settled, whether the data is going to be accessible.

David Friedberg: You know, here's what I, yeah, I think there could be, um, a very interesting verticalization that occurs here.

SPEAKER_31: If you look at the corpus of legal data and we talked about, you know, the do not pay legal case, but then we kind of quickly jumped off into.

SPEAKER_39: Hey, what if you fed it all that data and then made tools for lawyers or tools for, you know, people convicted of crimes and or who are interested in the law?

SPEAKER_31: Well, somebody could go and start taking every single, uh, court case in a jurisdiction in a vertical. Build, take an open source chat GPT, hire 10 lawyers and train a very nuanced version of chat GPT and a vertical. Right. I would back that company. Right. If there's like three or four entrepreneurs, two attorneys, two AI folks, and they want to make the legal version of chat GPT and they want to start ingesting their own data, training their own data. And that is something that chat GPT is going to have a very hard time beating. Right. Because this version would know jurisdiction, specific attorneys, different, um, uh, what do they call the different, um, court courts?

SPEAKER_09: Like there are different levels of courts, right? Right. Right. There's like appellate and this and that. Exactly.

SPEAKER_31: So, and then there are different themes and types of cases. So if you really start to hone that in, it would be the difference between like the chess, uh, AI and the poker one.

SPEAKER_14: Yeah. They might have things in common, but you know, you, you might be able to build a business just in vertical. So.

SPEAKER_334: Yep.

SPEAKER_14: I could see that. More to talk about this thing has really taken over.

SPEAKER_54: So much more. By the way, I would just add. Yeah. And so how was, what was the vibe leaving this event in, uh, uh, San Francisco?

SPEAKER_59: Like, you know, I think it's a, it was a really interesting split. Like there's the, there are the people who are like, this feels kind of BS, you know, like, I think a lot of people are wondering, like, does this have a Theranos vibe? And then a lot of people are just like, this is everything. And it's going to change the world. It was a real, I just in the elevator on the way down, there was also a weird moment where like a lady fainted and yeah, a woman fainted listening to it and had to be. And I happened to be right by her. And we were like, bringing her water. And everybody was like whispering, like, don't disturb Sam Altman or what?

Jason Calacanis: And he just sort of stared over at her. And then when she was in the chair, like seeming to kind of recover, he just resumed speaking, but didn't acknowledge it. Like it was such a, I mean, it literally like, the whole night had this kind of like humans versus robots vibe. And then I was walking back to my car and the only moving thing in that part of San Francisco was like a security robot up and down the street. And I was like, sometimes San Francisco literally is just sci-fi in action.

SPEAKER_339: You walked on the street alone in San Francisco. I'm a little worried. I maintain the defensive posture.

SPEAKER_199: I think we need to send it this week in security team. The robot was right there.

SPEAKER_340: I would have been fine.

SPEAKER_43: I don't think the robots got countermeasures. Uh, that's fascinating. It was a weird night.

SPEAKER_30: Did the woman faint like, because somebody was like, and this thing is going to get rid of, you know, every poetry author. And she was like, oh my Lord. She's like, I cannot, I cannot.

SPEAKER_09: And just fainted from like some specific claim that was made. I do not know the reason. Maybe blood sugar.

SPEAKER_342: I do declare Sam Bagman Freed is going to be let off and she just fainted.

SPEAKER_344: It was kind of a relentless parade of panels. Like nobody got to. Anyway.

Jason Calacanis: Um, it was a weird, it was a, it was a surreal. I'm not going to lie. Like I felt like I was either there at the birth of Skynet. Right. It was like, just this way, or I was there for just like the next big bubble.

SPEAKER_10: And I was like, just stay focused on the climate. This is real. This is not web three.

SPEAKER_31: I mean, listen, this, this is VR, AR VR hype cycle. Didn't work out. Yeah. Web three hype cycle. Didn't work out. Doesn't mean it can't in the future.

SPEAKER_14: Sometimes you get false starts. Uh, I, I suspect VR will be a false start to AR. This is not a false start. I mean, this is just clearly the future of everything. And we, we've been saying that machine learning and AI are gonna. Yeah, absolutely. I mean, you can see it in the tick tock algorithm. You can see the YouTube algorithm. Like those were just, this is just the start.

SPEAKER_308: This is just the start. But it does seem to be, the pace is what's unique here. That's the question, right?

SPEAKER_59: Is the pace fooling us or is it actually happening exponentially faster? Because if it's all happening at the same time and it becomes a race for $90 billion. For example, we probably, it probably is lights up.

SPEAKER_351: The pace here is, you know, it's, it's that compounding.

SPEAKER_54: It's like, how did you go bankrupt? And it was like slowly and then all at once. Like, how did the plane crash? It was like. Yeah. Slowly and then all at once. Like that's the moment we're in right now. You, the compounding nature of this because of cloud computing, because of the corpus of data sets, because of the number of people working on it. And it's self-reinfor, self-reinforced learning. And the money. And the money.

SPEAKER_352: Yeah. And the money.

SPEAKER_10: It was, it was a weird, like, I'm not gonna lie. Like it, this is real pretty intense just being like, is this real?

SPEAKER_354: And if it's real, like what the hell happens anyway? Yeah.

SPEAKER_54: Keith made a good point on the last podcast. He's like, you know, people are only sharing like the most incredible results from GPT three. Right. So it's like, it's the top 10% we're seeing. And so, yeah, one out of 10 is really, you know, you can take it to the bank. It's really impressive. Yeah. They're nine out of 10 or not. Yeah. I get the sense that like with the next check, GPT four check, GPT five, it's gonna go from one in 10 being absurdly impressive to one in five to one in three to, you know, nine out of 10.

Jason Calacanis: Yeah.

SPEAKER_59: I mean, I think that's the part where it's real. So then how fast is it gonna go? Like, hopefully he really means it when he says they're gonna take their time with GPT four. So they don't break everything, but I doubt it when $90 billion are on the line, that thing's coming out. We're all gonna die.

SPEAKER_359: I'll be slowing down. This thing's accelerating. Okay.

SPEAKER_59: Yeah. All right. All right. All right. All right. All right. All right. All right. Well, we've started our interviews for launch accelerator. Yes. Yes. That cohort just wrapped up. I am. I've actually just so happened that the first interview we did coming out of LA 26 is one of my climate companies. Oh, great. Tell us. Yeah. Lex key fober is the founder and CEO of United by zero, which has kind of two products. They've got a Chrome extension that provides sustainable alternatives to fast fashion. So if you go to H and M and you search for like a dress or like, Hey, you know what? You could actually get these things, which would be way better and gives you those options. And then like honey, basically for sustainable fashion, but also a JD power element where they do life cycle analysis for the fashion industry informed by consumer buying preferences and have a SAS product that they sell to companies.

SPEAKER_14: United by zero. I thought this was a very interesting company. We found, uh, you install the plugin, the Chrome extension rather. And while you're looking at options, you find a pair of jeans you like. Maybe you want the more sustainable version. Uh, and it gives you some options and that's really super slick.

SPEAKER_284: And then they can take all that data.

Jason Calacanis: They can do life cycle analysis for individual products for a company.

SPEAKER_59: Like they're in conversation with Vori, the super soft sweatpants so that they, then they could potentially have a score available for all their products. And that's the SAS. That's the SAS side too.

SPEAKER_32: Enjoy. Enjoy everybody. I'll let Lex explain it. Yeah. Yeah.

SPEAKER_16: Okay. Everybody. So the launch accelerator just completed its 26th cohort. And I'm going to do my favorite thing, which is sit down with some of the founders from that cohort over the next few weeks to dig into their businesses. First up is actually one of mine, one of mine that I championed United by zero CEO, Lex Keefauber. Welcome.

SPEAKER_366: Thank you, Molly.

SPEAKER_367: And I did not know that we were one of yours until right now, but that makes me feel like a thousand times better. Oh, heck.

SPEAKER_16: Yeah. All right.

Jason Calacanis: Well, tell me, I already know, but tell our listening audience, if you don't mind, what does United by zero do? Absolutely.

SPEAKER_367: So you can think of us as two things. We're like honey meets JD power for sustainable, sustainable shopping. And so what that means is that we can help you find more sustainable alternatives for the products that you're already shopping for that have been vetted so that you know you're actually buying things that are not going to screw over the future. Um, and so this is like really a visual product, which is very helpful to see. So if you head over United by zero.com, you can see a demo, or if you're watching this, I will show you briefly what that looks like. Amazing. It's both. So it's a browser extension that you can download and then use on all of the major fashion retail websites. So right now I'm looking at h and m.com and I see this beautiful black dress pulled up. And with this browser extension installed, you can see that there's a little button right by the title that is sitting right above it. And when you click on that button, a pop up window appears from the left side of the browser, and that shows you more sustainable alternatives that are in the same style. So we use image recognition and natural language processing to find similar type of clothes that you're already shopping for that are from our catalog of more sustainable alternatives. And we populate those right in the screen the same way that honey will show you if there's deals, we'll show you if there's more sustainable alternatives for the clothes you're already shopping for.

SPEAKER_59: Amazing. And do you partner with?

Jason Calacanis: I mean, because that isn't just like a little extension that I go click at the top where my password extension is. It's like embedded on the site. So are you partnered with these retailers in some way?

SPEAKER_367: We're, I mean, shockingly enough, no. H&M doesn't necessarily want us to be showing more sustainable. I was confused.

Jason Calacanis: I was like, how did you get that nice little button on the H&M website? It seems like they wouldn't go for that.

SPEAKER_367: Well, the technology is a fantastic thing. We're able to overlay on whatever it is that you're looking at. And so we can figure out where the title is and then put the button right there. But there's also two parts to this, right? One is how do you find more sustainable alternatives? And the second is how do you know that they're sustainable? So the second thing that we do is partner with sustainable brands to do a full level accounting of how much carbon their products use, how much water their products use, how much land resource, all of that in the distillate into something that's super easy for consumers to understand, but verified by science. So we're doing like the hard work of the homework and we're giving you those more sustainable alternatives right in the process of you, the consumer shopping for them. Amazing.

SPEAKER_58: And it's just that simple for consumers. And how often do you find that people choose to buy that more sustainable alternative?

SPEAKER_376: You know, the Pareto principle is just like in effect everywhere.

SPEAKER_367: So we find that about 80% of people want to explore more sustainable opportunities and then 20% of people are actually going to click through and then go on to being converting. So if you're a sustainable brand, you know, and you're marketing to everyone, 20% of your shoppers are probably going to drive 80% of your sales.

SPEAKER_372: And we can help figure out who those 20% of the shoppers are and then drive them to your front door.

Jason Calacanis: But if 20% of the people who click on your extension convert, that's a high conversion rate based on what I know.

SPEAKER_366: I mean, not everyone buys 20% leaves H&M and goes to that.

SPEAKER_367: And then your conversion rates when you're on their site are sort of optimized by whatever your brand conversion rates are. Right. Totally.

SPEAKER_371: And then how do you make money from this?

SPEAKER_367: So we make money in two ways. This is an affiliate play. So like we make money when we drive customers to brands place and they buy things. But then we also we have a SaaS product and that SaaS product is when we work directly with brands to help them understand what makes their product sustainable. And so we've built the software tools to create what's called a lifecycle analysis, which is sort of like a full scale. Let's understand what goes into your product. And we've done that in a way that's scalable. So we can only charge. We only need to charge 50 bucks for you to get the first level of that lifecycle analysis. And then we give that back to the brands. 50 bucks per product? No, 50 bucks for everything. Total. And then what we do is like, yeah. And then we build that into a widget that the brands can put on their website to show all their customers, hey, this is how much carbon, this is how much water. And then we charge per product for using that widget on their site.

SPEAKER_372: And that's kind of the J.D. Power part of it. Exactly. So Bain came out with a report recently.

SPEAKER_367: It said a ton of people want to shop sustainably, but not everyone's doing it. And it's this big problem in the industry, right? It's this say-do gap. We all say we want to do something, but the action isn't falling through. Mm-hmm. And they dug deep to figure out why. Why is it that people are not falling through with their words? And there's two main issues. One is they can't find more sustainable alternatives where they're already shopping. And that's what our extension helps solve. And the second is that it's really hard to tell what's sustainable and what isn't. And so that's where the J.D. Power starts to come in, where it's like, well, we can do that math for you. So you don't need to have a advanced degree in textile manufacturing to know that this thing is actually not only is it good, but like the brand behind it has put in the work to understand what the supply chain looks like, how the fibers are manufactured, everything that you need to know to build a company with integrity.

Jason Calacanis: And so when you partner with a company on that SaaS product, they provide you with that data and you, what, you have like an algorithm to kind of quickly turn it into a bit of a score? Yeah.

SPEAKER_386: Well, I mean, it's like everything. It's like, it's both like incredibly simple and like incredibly complicated at the same time.

SPEAKER_367: So the simple part is like, if you think about, let's take, you know, a cotton shirt. And so that cotton was grown somewhere. Let's say the cotton was grown in Egypt and we know about how much energy it takes to go grow cotton in Egypt. And then if you ship it to Vietnam for processing, and then you use X, Y, and Z dye on it, and then you put it in a package that's recyclable and then you ship it to someone, you can add up each one of those component parts to figure out how much carbon this product used. And so we can do all of those times of calculations using publicly available data. So we're not reinventing the wheel on that. We're just applying software to it to make it a lot easier. And then we're applying design and UX to it so that it actually means something to consumers.

SPEAKER_372: And you don't have to like read an inscrutable report to be able to have trust in the products that you're buying.

Jason Calacanis: And how does it show up to consumers? Is it like a badge? Is it a score? Is it a, you know, gold, silver, bronze kind of ranking?

SPEAKER_386: This is actually, this is the thing that I'm like most excited about everything that we're doing.

SPEAKER_367: People are not monoliths, right? We're not all the same. And sustainability can mean every, like so many different things to so many different people. So I have, I have kids, right? We were talking about this before a kid and then another one come in. And when I'm buying a clothes for them, I'm thinking like, how many, was there any toxic chemicals used in this? And toxicity is an important part of it. And when my wife is shopping for a dress to go out, she may want to know, well, can I recycle this? Or is there an easy way to resell it? Or can it be given back to the company and then not thrown into a landfill? Also important. When I'm buying a jacket outside, I just want to know what's the thing that has the least amount of carbon that's used, but special. All these parts are aspects of sustainability. Because our browser extension is something that people interact with, we're able to test so many different ways of taking that data and putting it in front of consumers to create the best outcome for the sustainable brands. So there's not one answer to that question, like which is the one monolithic way to represent this? The real answer is, we need to be able to adapt the information to the person in the instance that they are shopping, depending on what they're shopping for. That doesn't mean change it. The information is the same, whether it's like carbon, water, toxicity, land use, what have you. But it depends on what is what you need to show that person in order for them to feel like this is something that's going to compel them to change their behavior. And then we're using like different AI models to like create more dynamic ways of saying, how do we make sustainability a selling point and not just some random number that no one knows what to do with?

Jason Calacanis: So what might it say? Would it say like low toxins, like you said, or very low water use or it would highlight like you would see in that display of three different alternative black dresses?

SPEAKER_58: Maybe each one of them would have a different attribute and you could be like, I'm about the water.

SPEAKER_367: Yeah. So we have like an overall score that we've developed called the Z score. We also have things like how much carbon was used, how much water was used, and there's other elements of it as well that are adjacent and important. Like, is it a woman owned company or was it made in the United States or is it does it have a circular business model incorporated into it? All of which are important factors of people making decisions and to some degree have to do with this like general ecosystem around sustainability. But they each have different level of importance, depending on who you are, what you're shopping for, a million different factors that go into us being very hard to pin down in any one time.

SPEAKER_16: Well, I will try to pin you down on our two key accelerator questions.

Jason Calacanis: What is, as you look forward, your path to $10 million and your path to $100 million?

SPEAKER_367: So the $10 million path is pretty straightforward, which is we're starting in the fashion industry for a lot of reasons, but one of them is that there's 28,000 different independent fashion companies. We only need 2,000 to get to $10 million in MRR. And so there's low hanging fruit in terms of the companies that want this information and need new customers. But the real interesting question is the $100 million question, or even I would say the $10 billion question. And that goes to the basic understanding of what is a consumer conversion funnel. And for the last 15 years, it hasn't changed that much. You have awareness, intent, and then conversion. So an awareness, meta, Facebook, and whatnot dominates that, right? You want to get your brand in front of people. That's been the best place to do it for a long time. Intent, you go to Google, you Google something, they serve you an ad. That's, again, a trillion-dollar platform. And conversion, Amazon has optimized that in a way that all these three things are sort of the full stack of consumerism. But there's this fourth step in the funnel that we're seeing developed right now, and that's integrity. That's people asking the question, is this product right for what I believe in? And you're seeing that largely with like millennials and Gen Zs. And that's a really complicated question that requires this sort of kaleidoscopic view of what's in the product and what do you believe in? But as that becomes something that's built into our thought process of, am I going to buy this thing? We need a way to solve for that step of the funnel. And that's what we're building. United by Zero is building that integrity step into purchasing, no matter where you're purchasing, whether it's on the internet, on your phone, in the metaverse, in real life. It's the same question. Does this thing match my values? As we get to understand more about what matters to consumers and how that relates to the science, we're going to build tools that are going to really be able to develop and then own that step of the funnel.

SPEAKER_399: And then finally, I know you also have a podcast. This is your chance to plug it. Wow.

SPEAKER_367: I didn't expect that. So, yeah, this whole thing came. I have a podcast called Who Saving the Planet. Well, it's related.

Jason Calacanis: Right. Who Saves the Planet is the name of the podcast.

SPEAKER_367: Who's Saving the Planet. Who's Saving the Planet. We love, you know, that we're not done yet. There's no work in progress. Exactly.

SPEAKER_402: It's a gerund. Yes.

SPEAKER_367: Exactly. So, yeah, it's been on for a couple of years now. We've got over a hundred episodes and we've interviewed CEOs of companies that are trying to save the planet. From billion dollar companies like the guys behind Solugen to startups. And United by Zero started there and started listening to these founders say, we've got these two major problems. One is nobody knows what sustainability means. So we don't know how to actually get an ROI for making more sustainable products. And the second is like, how do we break through the noise of all the greenwashing and the confusing marketing to actually reach the customers who care? And so that's what we do. We help like reach the people that care and clarify the issue for everyone. Awesome.

Jason Calacanis: Lex Keefaber is the founder and CEO of United by Zero. Congratulations.

SPEAKER_408: Thank you very much, Molly. All your success.

Jason Calacanis: And I look forward to much more.

SPEAKER_408: Take care. All right.

SPEAKER_102: Cheers. All right.

SPEAKER_284: Thanks, Lex. Great job on the interview and good luck with your company. Congratulations on graduating the launch accelerator. If you want to come to the launch accelerator, launch.co slash apply. Um, yeah, we put a hundred K in your company. We work with you for, I think 14 or 16 weeks, uh, introduce you to hundreds of investors and try to help you increase product velocity, raise money. And yeah, generally help validate you and introduce you to folks in Silicon Valley, uh, to get your startup at a second base just a little bit faster.

SPEAKER_308: Yeah.

SPEAKER_11: We'll be talking to all seven of those founders from our cohort over the next seven, not seven weeks, the next few weeks. Uh, and then yes. Okay. Boomer.

Jason Calacanis: It's Friday variety show. You know, you love it. Yeah. This time. Let's see.

SPEAKER_11: Rachel has Juliet Meskers co-founder and CEO of Moth, which stands for minds open to heal, teaching people to prioritize mental health at the same level as physical health for education. Actually, they offer classes at companies and schools that cover mental health issues like CPR kind of for keeping your brain.

SPEAKER_10: Okay.

SPEAKER_135: Which I mean, seriously right now. Yeah.

SPEAKER_418: We need this more, more mental health services equals better. People are on edge.

SPEAKER_284: It's a tough out there. Yeah. And, uh, yeah, it's, it's great that the stigma is off of this and we're really addressing it post pandemic.

SPEAKER_419: It just seems like this became so acute for so many people. Yeah.

Jason Calacanis: People are really cracking and Gen Z in particular is so aware of it, but I'm glad that that's, it's interesting that that's probably the age bracket.

SPEAKER_11: That's like the most more founders are going to come more Gen Z founders will probably go into mental health and climate than I would imagine lots of other areas.

SPEAKER_284: Yeah, this is, uh, I mean, it's a big discussion. We'll have this discussion more on this program, but yeah, mental health is something we gotta it's in some ways.

SPEAKER_54: I think as a society, it's really great when you see, like, as we, um, start addressing issues and you're like, do we need to address that? Do startups need to address? It's like, well, why not? Right? Yeah. Why wouldn't we do that? Like, okay. Yeah. We got seat belts in cars. Uh, you know, that kills people. And, uh, we have people suffer from cancer. Okay. Yeah. We we're working on people surviving cancer and treatments and early, you know, w w w where's the focus on mental health?

SPEAKER_00: Where's the early detection for mental health, the mental health crisis we have in this country? Yeah.

SPEAKER_54: Where's the early detection for people who are, you know, going to become addicted and have addiction issues. Well, we have that for cancer, right? Yeah. People go get, uh, checked for different types of cancers and screenings. I think there's gotta be a lot more awareness of this and, and really getting ahead of it before people have acute problems and God forbid harm themselves, other people, you know, just feel despair. Let's get ahead of this people. And I think founders, you know, really start looking at, I'm very proud of the calm investment as but one example of just, you know, being able to reduce suffering in the world. It may not be meditation. It may not be. I don't know. It may not be curing cancer. Right.

SPEAKER_00: Uh, but if it is reducing suffering for hundreds of millions of people in the world, that's well worth it. I think it does ultimately save some lives.

SPEAKER_02: Yeah, for sure. Yeah, definitely.

SPEAKER_268: So enjoy this interview. Uh, and thanks Rachel.

SPEAKER_431: Thank you, Juliet Meskers for joining us today on the latest segment of Okay Boomer.

SPEAKER_432: For those of you who don't know, Juliet is the CEO and co-founder of Moth and Moth stands for Minds Open to Heal. Juliet again. Thank you. Thanks so much for having me. I'm really excited to be here. So first things first, I'm going to need you to introduce Moth. What do you guys do? Um, I already said what it stands for. So we have a little bit of an idea, but run me through that.

SPEAKER_441: Yeah, so Moth is a mental health education and training platform. And essentially what we do is we provide CPR, but for mental health. So in the same way that if someone was going through cardiac arrest, experiencing cardiac arrest, someone who is trained in CPR would provide CPR until the ambulance arrived. What we are teaching people to do is to be that initial bystander in a mental health crisis. Um, so that they can assist someone through a panic attack, through a manic episode, or just through experiencing day to day stress, anxiety, depression, so on and so forth. And then connect that individual to further treatment. So it is the essential training that everybody should have so that we're in so that when we're in difficult situations, we don't panic and say or do the wrong thing. So it helps prevent liability. It helps improve company morale, corporate morale. Um, we train corporate companies and we train in universities and schools as well. That's awesome.

SPEAKER_432: So I actually was a research assistant at Penn State where I went to college and we did a bunch of these trainings and none of them had to do with mental health. So it's really cool that you guys are filling in this gap. It's obviously needed. And if you don't mind sharing, I'd love to know more about the story of Moth and where this actually came from. Yeah.

SPEAKER_441: So the company used to be called mental health global network. And I chose that name because I was 22 when I started the company and I needed something that sounded legitimate.

SPEAKER_450: Um, so I just thought mental health global network, that sounds real.

SPEAKER_441: Um, and for a long time, for the past three years, we've been trying to pick a new name, something more brandable, something more memorable. Yeah. And we went through this process this year of hiring a professional namer. And it was a four month process of going through 500 names with a marketing team. And finally we came to a name that we all liked and we took it to the trademark lawyer and they said, you can't use this. Um, so we were super disappointed and we were back at square one and my brilliant graphic designer slash artistic director said, well, maybe if I come up with a logo first, that'll help spark a name. So he did some research and he came up with a moth for a logo. And I was a little skeptical at first, but he explained that the most notable characteristic of a moth is its attraction to light. And it's known for its psychic abilities, for its wisdom. And even though it often lives in the shadow of a butterfly, it also is just as essential to the ecosystem. It also goes through a metamorphosis and it's a really beautiful creature at the end of the day. And so came up with a logo for a moth and I thought, well, we're never going to use the name moth. So let me look up what moth means in Latin. And I found the name Tania and I thought, oh, that's really pretty. And I sent it to my board and someone wrote back and said, yeah, Tania works, but also moth.com is available. It's open. No one has it right now. So we were like, okay, that is a sign. So we decided to go through with the name moth and we teach people how to guide others towards the light through periods of darkness. So the whole attraction to light characteristic was really meaningful to us.

SPEAKER_460: That is so interesting. Like starting from the logo and moving forward.

SPEAKER_431: And Jason actually talks a lot about how hard it is to secure URLs like that. He has a history in, in the media side. And so that's super cool that you've got a, a four letter URL. Congratulations. I know that's, that's quite tough.

SPEAKER_432: Yes. Well, we're still saving up to buy the URL, but it is available.

SPEAKER_465: So, you know, it's there, you know, it's there.

SPEAKER_441: We just like need an additional $40,000, but we're going to get it. We're going to get it.

SPEAKER_432: I believe in you. That's really cool though. That's available. It's kind of surprising. And I love, I love the sentiment behind, behind the moth logo. And I've seen you guys on Twitter and it looks really cool. It kind of reminds me almost like something that somebody would have tattooed on them. Like, it looks very pretty. Like it's a pretty logo. Yeah.

SPEAKER_438: My graphic designer actually did get it tattooed on him. So no way. No way. Dude.

SPEAKER_431: That's incredible. Yeah. For you as a founder, I know we are around the same age. What was your founder journey? Cause it sounds like you've been doing this for quite a long time.

SPEAKER_441: Yeah. So my founder journey really started back in college. I had personally always struggled with my own mental health and my university went to University of Delaware and we had a really bad suicide problem. And the school really wasn't doing much to address it. And it was really getting swept under the rug over and over again. And after one student who was this really popular, exuberant, funny, athletic, charming, intelligent human being took his life, a boy named Connor Mullen. And a group of his friends were kind of like, fuck this. Honestly, like the school's not doing anything once again and we need to do something to address this. So they started a group called friends for friends, which I quickly joined. And the group's goal was to raise mental health awareness and suicide prevention to facilitate friendships on campus between, um, like-minded individuals from all across different types of social groups and to raise money for the counseling center. And I was getting really involved in the group. I joined the marketing team. We were doing all these suicide prevention and awareness campaigns when I ended up losing

SPEAKER_444: a good friend of mine from my childhood to suicide.

SPEAKER_441: And I really ruminated and felt guilty for a really long time and asked myself, you know, how could I consider myself to be such a mental health advocate? And yet I totally missed the warning signs in my own friend. And what I realized is that there's this huge gap in the mental health movement that sits between mental health awareness, celebrities, athletes, brands like Mad Happy coming out and raising awareness about the fact that mental illness is real and it's legitimate. And then on the other side of the movement, we see companies like Talkspace and BetterHelp and insurance companies and EAP programs making mental health resources more affordable, more accessible. But that gap that really exists in between is education. And 70% of people who receive a mental health diagnosis will not seek out treatment because of the stigma associated with doing so. And when you really look deep into stigma and where it stems from, it comes from a lack of knowledge. People tend to fear what they don't understand. And growing up, we have PE and wellness class and we have this general understanding of the importance of taking care of our body. And then we have this concept of learned empathy where we generate this understanding through societal norms that if a coworker receives a cancer diagnosis, you're going to offer to pick up some of their workload while they're going through treatment. You're going to drop off a castle at their house. If you're a student and your classmate tears their ACL, you're going to offer to carry their books to class while they're on crutches. But when it comes to mental health, we never received that formalized education. And so that fear of the unknown leads individuals to making negative assumptions. And those negative assumptions perpetuate stigma. And then, like I said before, that stigma then serves as a pervasive barrier that prevents individuals from getting the treatment that they need. And so at the time I was taking an entrepreneurship course and a nonprofit management course. And in both courses, I had to come up with the idea for a nonprofit and for a business. And the idea for Mental Health Global Network, which is now Moth, was really born through those two classes. And I was supposed to work in the fashion industry. I had a job lined up and really last second, the week after I graduated, my job offer was pulled. And I said, I'm going to give myself 24 hours to cry about this and be sad. And then I'm going to actually start the company that I thought of in school. And so I called up the smartest, most capable person I could think of joining me as a co-founder and pitch the idea to them. And they came on for the first three years. And that's, yeah, that's how I got to where I am now.

SPEAKER_432: That's awesome. And I'm incredibly sorry for your loss. I know that's a very hard thing to go through. And if anybody out there is struggling, know you were loved. I know the suicide and crisis lifeline is 988 in America. It's open 24 seven. So you're doing incredible work here. Sorry you had to find out that this was a gap that needed to be filled the way that you did. But I'm incredibly happy that you're out here solving this problem. And obviously there needs to be more people acting in this space. What advice do you have for others trying to get into the mental health awareness and a suicide prevention space?

SPEAKER_441: Yeah. So I think my number one advice, and it is kind of controversial advice, but I would say don't start a nonprofit. I think a lot of people who feel like they have a mission, they have a passion towards a purpose that they want to fulfill, think, oh, I'll start a nonprofit. And the thing about nonprofits is that they're actually very difficult to maintain. They are very difficult to sustain because of all the different rules and regulations you need to follow in order to receive those tax breaks. And the nonprofit market is super, super oversaturated, especially when it comes to mental health. And so one of the biggest ways that nonprofits make their money is through grants. And when there are all these different nonprofits going for the same grants, it's very difficult to secure that funding. It's also a lot of work to ask people to give you money, to give you donations. It's actually a lot easier to have a service that you provide and receive payment for. And a lot of nonprofits do have that as well. But what I learned is unless you come to the table with a lot of funding to begin with, whether it be group funding or your own personal net worth that you're able to put into a nonprofit, it's very difficult to manage one. So I would say, go ahead and take that leap, start a company, start a business, and then with the funding that you're able to raise through that business, through profit, then you can find a cause that you want to donate it towards. And so for us, for every paid training that we do, we provide a training to a community in need that can't afford it. And so that's our way of being able to give back without having that nonprofit status.

SPEAKER_460: Yeah, that's awesome that you're able to do that. And thank you for that interesting explanation.

SPEAKER_432: I did not know, and I still don't know, I feel like, a lot about how nonprofits work and the backbone behind them. And it's definitely something that I want to be looking for, like looking forward to educate myself a little bit more on, especially with how many companies now are nonprofits. This like near the holidays, I've been seeing so many advertisements to giving back, definitely top of people's mind. And this feels like a very timely startup to be talking about with the holiday season. I know this can be a really difficult time for some. So thank you so much for coming on and letting people know that you, Ma, you guys are making a difference. I'm super excited to see what you do in the long term. Thank you so much. I really appreciate it. I've loved speaking with you today. And the last question, where can people find you and where can people find Moth if they'd like to reach out?

SPEAKER_441: Yeah, so they can go to Moth Health dot com to check out more about our business. Right now we are in the process of also building a marketplace. So in combination with our mental health education services, we're also building a marketplace that will be launching in the next few months so that people can come to our site and put in how much they can afford for treatment. What kind of treatment they're looking for? Are they looking for a dietitian? Are they looking for a therapist, LGBTQI plus counselor, so on and so forth? And we'll give a recommended list of outcomes for them. So it's kind of like an Expedia for mental health. So that's something to keep an eye on in the next couple of months. And to contact me directly, you can go to you can email me at Juliet at Moth Health dot com. We are Moth Health on Instagram and on Twitter as well.

SPEAKER_435: Awesome. Thank you so much, Julia. Thanks so much, Rachel.

SPEAKER_507: All right, everybody. Great, great week. We got through it. We do have a show on Monday.

David Friedberg: Acquire dot com. Previously micro acquire founder Andrew guys. Daggy is joining for an interview with me about M&A and mergers and acquisitions. Top of mind. A lot of companies are shutting down, giving up on the dream, but they, you know, they built some asset and Andrew has this great marketplace where he helps buyers and sellers find each other. And so we had a really candid talk about when it's time to sell. It was a really great, great interview. Yep.

SPEAKER_59: Amazing. We're going to have that and a little bit of news about the investing space overall. It's all coming up on Monday, even if you're off. Yeah, we're off. We are, but we'll still have a show for you.

SPEAKER_36: We're always there for you guys. Yeah. Enjoy the long weekend. Bye-bye. Have a great weekend. Bye.