SPEAKER_00: Hey, everybody. Welcome to Friday. You know, we have a great variety show for you today. We love Friday variety. First, producer Rachel is going to hop on and we're going to chop up some news. We have some great stories to run through today. Apple blocked an update to an

SPEAKER_01: app on its app store for having generative AI capabilities that Apple determined are just not safe. They want either content controls or age restrictions. That is a big deal. And we talk about what that might mean for this budding industry. Salesforce announced earnings. Its stock jumped as much as 16% after reporting those Q4 earnings and playing the Wall Street game. Very similar to Meta. We're going to break down how tech is now being forced to play ball

SPEAKER_02: with Wall Street. Then we have interviews. I talked to Dylan McDonald, the CEO of Foodini, a startup that connects people with dietary needs to suitable restaurants and menu items. And then finally, producer Rachel is back with OK Boomer and spoiler alert,

SPEAKER_03: her first ever sci fi book. This is incredible. It's going to be a great show. Stick with us.

SPEAKER_05: This Week in Startups is brought to you by Mbroker. Mbroker's startup insurance program

SPEAKER_00: helps startups secure the most important types of insurance at a lower cost and with less hassle. Save up to 20% off of traditional insurance today at mbroker.com slash twist. While you're there, get an extra 10% off using offer code twist. Vanta. Compliance and security shouldn't be a deal breaker for startups to win new business. Vanta makes it easy for companies to get a sock to report fast. Twist listeners can get $1,000 off for a limited time at vanta.com slash twist. And org space. If you're a startup and you are not building a performance culture, your competitors will eat your lunch. Get $2,000 of credits on pro plans with a 30-day free trial at orgspace.io slash twist. All right, Rachel, it's been a minute. Let's chop some news up.

SPEAKER_09: It has been. It has been a while. Well, I'm super pumped to join you today for the news because this first story I'm really interested in. When you send it in to our group chat, immediately I was like,

SPEAKER_10: we have to cover this one first. So Apple's delayed the approval of an update to chat GPT powered email apps on its app store over concerns that it could generate an inappropriate feed of content for children. Dun, dun, dun. I know. The big question here, should chat GPT powered apps be limited to

SPEAKER_12: those 17 and up, even if they aren't using search? Because that's what Apple said, right? So they blocked

SPEAKER_01: to this update. It's an app called Blue Mail, and they wanted to integrate the chat GPT technology from OpenAI to automate writing emails. And then Apple said, you need to set your app permissions to 17 and up because from what we've seen of Bing or chat GPT, there's a chance that this could generate inappropriate material is how I

SPEAKER_10: understand it. Exactly. And Bing is already up on the app store too. So we see them and they actually do have that

SPEAKER_18: function that says people should only be 17 or up when using this. And the reason they have that is because it is a search tool. So your kids can look up something inappropriate. But this app is just for mail. It's not necessarily something where kids are going to be searching on it. Apparently, Apple wanted

SPEAKER_01: Blue Mail to either set move up its age restriction to 17 or include content filtering. And I mean, you know, just because it's generating email, if it's a sort of a similar thing as a chat GPT, like a prompt, you know, Katie bar the door, you can totally put in your email prompt, that's like, write me a, you know, death threat with bomb making materials. Like there's, if there is no content restriction, there's certainly the option to do that. But I feel like this is interesting for a couple of reasons, not least of which the most of which in fact is, is Apple's gatekeeper role here, right? It's like, okay, parents just came in to the AI world. And sometimes you want Apple to exercise its

SPEAKER_00: gatekeeping authority. And sometimes you don't. And I wonder if this, which one this is.

SPEAKER_23: Yeah. And moving forward, then, do you think that all apps using generative AI should have a

SPEAKER_18: restriction to make them 17 or older, even if that might make it a little bit harder for their app

SPEAKER_00: to grow? I mean, I think this is going to be a fascinating conversation. And I, I find it,

SPEAKER_01: which is a dodge, I know, to the, to the question, but it gets to this kind of fundamental question of how these things are operating and operated. There was a really interesting piece in the Atlantic that was about this woman who I think in 2020 wrote a whole long piece about the dangers of making AI sound like people like, remember, you might not, but when, when Google, I say that cause like I'm old, but when Google unveiled, this is somewhat recent, the bot that has the ability to call out and make a restaurant reservation for you. There was a huge freak out about the fact that this chat bot, literal interactive thing that would talk to people, wasn't identifying itself as not human. And there was a big moral debate about like, why are we designing things and designing AI to imitate humans? And we clearly have, and that's why you get these sort of hallucinatory responses potentially from, you know, Bing or chat GPT when you want it to write a play about something horrible. On the other hand, there's the argument that we've heard about, like, if you restrict its content, are you censoring it? Is it, you know, or are you, are you imposing your values? And Apple has imposed its values on us for years. Apple is super freaking prudish. Like it's, you know, they have always wanted to create this kind of sanitized playground on the app store and on in the Apple universe where there's like no sex and no violence. And there is really, and there's always been a question about like, is that your right, Apple? Like, just let me have what I want to have.

SPEAKER_18: Yeah. And it's funny that you mentioned how everybody was kind of like up in arms when they let people use Google to order from restaurants. Cause one of my roommates, it feels like you can't scroll through their TikTok page without hearing a generative AI, um, like bot over a video game that sounds like a president. So a lot of them are a Barack Obama. This is such a thing. And I swear, yeah. So there's like a lot of people that will use generative AI to like sound like Barack Obama or, um, other people that are politicians. Barack Obama tends to be the one that comes up most for him. Um, and he's like, Oh, these are so funny, like starting off, but it really does. Like when you start thinking about it, I guess it's like a little bit more critically, um, obviously Apple's thinking a lot more long-term here. It starts off as something that is just like an email app or just a funny video online. It's like how much further can, can it dive, um, dive into, I guess like another

SPEAKER_39: question I have for you is how much of a headwind do you think that Apple's guards, I guess, around AI, right? How much of a headwind could this be for the industry?

SPEAKER_01: Yep. I mean, it, it's a big deal, right? Like, I think that it was kind of a slow burn for me, this story. Like at first I was like, Oh, okay. And then I'm like, wait a second, this is a huge deal because it is Apple at a time when there's all this scrutiny on these, on tech companies and Apple and its app store still just basically saying like, no, we think this is dangerous. And, and that immediately puts the question of whether this is dangerous at the forefront and into everybody's minds. And all of a sudden you cannot sidestep this question of the content. Everything is content moderation. I've been saying this for years, right? And so is AI and you can't now sidestep the conversation about what should it deliver and how, and how should it be implemented in ways that can safeguard society for God's sake. Like I saw some stupid story going around today about how like Sam Altman, who is simultaneously going to make billions off of this and is quite certain, certain that it's going to kill us all was like telling people like, Oh yeah, I have a whole like bunker situation and like land in Big Sur that I can fly to. And I have like a bunch

SPEAKER_02: of guns and, and somebody was like, what are you going to do? And it was like, in case AI comes to try to destroy us. And it was like, okay, what are you going to shoot? Like your internet cable?

SPEAKER_48: Like, it's just this, it's like, what are we supposed to be hiding from? Like we need to know.

SPEAKER_32: Right. And also if you're designing this thing that you think is going to kill people,

SPEAKER_00: maybe design it differently. Like you're in charge. This is not, there's not open AI, like AI bots were not born sprung full fledged from the forehead of Zeus with like eternal rights to not ever be censored. Like make it safe. But we, but now we're going to have this

SPEAKER_01: like bizarre, huge conversation about what safe means and like who even knows. But anyway, all that is to say, I think Apple putting in their thumb on the scale here and saying, we think this is a

SPEAKER_51: really harmful, potentially harmful, um, interaction for people is a huge deal and a big headwind.

SPEAKER_18: And blue mail does have some restrictions and content filtering around it. I guess the content

SPEAKER_10: filtering in place is just not good enough. Uh, like for Apple's standards, do you think Apple should be the one to create this standard? Or do you think like AI companies can be like, okay,

SPEAKER_18: listen, we tried to put guardrails up. Um, this is as good as it's going to get. And then they should be allowed on that platform.

SPEAKER_01: I just think this is going to be a huge firestorm. Absolutely. Right. And it's just going to lead to it more and more, uh, questions about whether Apple is a monopoly, whether they should have this gatekeeper power. The truth is that the, the truth is that we're like, in some ways have always been thinking about this wrong. Like not everything, first of all, has to be an app. Like if Apple

SPEAKER_57: doesn't allow an app on the app store, that seems fine. You could just go to like a mobile website.

SPEAKER_02: Like, like, I don't understand why everybody's just like, this is the only framework in which we can ever operate is the app store. Just whatever happened to the mobile web, like create

SPEAKER_01: a cool mobile web interface for whatever it is you want to do. And then nobody has to go through the app store and they can't gatekeep you. So I sort of feel like I would really like the brightest minds in the entire world to be a little more creative about how they deliver their products to us

SPEAKER_02: than to sit there and go like the app store will let me have it.

SPEAKER_53: Yeah. I'm blanking, but there was a really cool program that fresh use, um, on the launch

SPEAKER_18: team that can create any like web page, um, and make it look like an app. I don't remember what he used, but I believe it was used for like our app for the all in summit, which was really cool. Yeah. But, um, you're right. Like moving forward, people are, I feel like are going to try to find loopholes and I think that would be one of them. So if you're somebody that can't be on the app store would be cool. Um, I guess we have to find that post show, uh, the platform that he used, but it does seem like people are making loopholes. Exactly. Like the web is an

SPEAKER_48: open platform. The app store is not, it doesn't seem that hard to me for people to just like work

SPEAKER_01: around it. But if they're going to just sit there and accept that this is a huge headwind, because this is the only way you can ever deliver new technology. Well, then I guess you're just not, I guess you're not that creative.

SPEAKER_65: I've been dealing with business insurance for three decades. I am on the board of a bunch of companies. I watch people who don't have insurance, get themselves into trouble all the time. Switching providers has always been a nightmare. It's too expensive. It takes too much time. And often it doesn't even guarantee better coverage, but now you can make switching radically simple with in broker. Yes. And broker is the perfect destination for industry tailored commercial insurance. It's business insurance specifically for startups in broker single application help startups get four quotes, one, two, three, four, four, four lines of coverage in just 15 minutes. They connect you with one of their expert brokers for unmatched service that goes beyond your policy. And listen, broker is such an amazing product. I use it. A lot of my startups use it. It's so easy to use. So try and broker today with code twist and get 10% off their startup package at in broker.com slash twist. That's E M B R O K E R.com slash T W I S T. And use the code twist. So you get that 10% off. It's meaningful. Every dollar counts right now. We love you and broker. Thank you so much for supporting this podcast for so many years.

SPEAKER_23: And how has, I know you've mentioned before, like how TikTok kind of like how approach the app

Jason Calacanis: store in general, um, how did TikTok approach this approach being listed on the app store? I don't know, actually, does it have a 17 and up? I don't think it has an age restriction, right?

SPEAKER_71: Cause isn't that kind of crazy? Cause you can really see, like I said before, there are people saying

SPEAKER_18: some pretty heinous stuff, like as politicians sometimes, uh, playing Minecraft. So when you're

SPEAKER_10: looking at the screen, it does look like something that is age appropriate, but it's definitely super

SPEAKER_02: not great for your kid to be listening to you. Right. It is a little like, I mean, YouTube is on

SPEAKER_01: the app store, right? Like it is a little absurd. These sort of, these sort of choices like that seem kind of random, but I do think that what it really is about is Apple putting a flag in the sand and saying, we're not, we're not going to let this technology just be released roughshod when, as we have pointed out in this show, it's pretty much an alpha. Like it is not necessarily ready for human consumption. And maybe this is really Apple being like, uh, don't think you're just going to flood the app store with things that can, you know, create horrible interactions or give in accurate information or just be like generally unsafe or destructive. TikTok, by the way, side note, just introducing a quick headline as TikTok tries to make itself more safe. It apparently is setting a new automatic one hour screen time limits for teenagers for anybody under the age

SPEAKER_79: of 18, which actually I think they should set that for everyone. Like I need it. There's like the

SPEAKER_09: screen time that you can do, but I always end up bypassing it. And I also saw in an article that if

SPEAKER_18: you were under the age of 13, even it goes one step higher where like the parents have to input the code. And I think that's a really smart decision as well. So if you're under 18, you can kind of be like, okay, yeah, this is like not great. You can use some of your, um, like decision-making skills at that point and make your own adult decision and be like, okay, yeah, like I probably spent too much time on this today, but if you're under 13, your parent can get the code. So yeah, which is the,

SPEAKER_00: that exists built into iOS. I know. Cause my son used to complain about it for years. Um, but it is,

SPEAKER_01: it's very interesting to see TikTok do that, to try to make itself a more, a safer product. Like this is the, this is the conversation and I bring up the TikTok thing because that's the conversation that's happening right now around technology is like, is it safe? Is it harmful? How damaging is it? Yeah. And who, and who is in charge, you know, and Apple again, stepping in to be the, like the dad here who controls the passcode is a failure of all kinds of other governance from, you know, regulation to

SPEAKER_18: just companies releasing things when they're ready. And we've seen like, it's interesting how TikTok kind of already had these, um, functions if you use the app over in China. And it's also crazy to think that this isn't something already on Instagram after all that stuff came out, um, about how horrible Instagram is to, especially like young girls, mental health. You'd think that this would be something where I'm talking about the feature where the parent has, um, those controls

SPEAKER_10: put them on for like a certain amount of time. I feel like that should have been like, this feels late.

SPEAKER_09: Like this feels like something that then might've already been, we should have already had this, but yeah,

SPEAKER_01: that's what I'm talking. I mean, that's what I'm saying. Like it, if we're to the point where Apple has to be like, this could be damaging. And, and frankly that puts Apple in the position. If you really like the, the slippery slope from the Apple perspective is that if nobody else for one thing cares about our health and safety, right. They're just out here like, ah, we'll make billions on open AI and chat GPT, but like, we don't care if it kills us all. We got our guns in our bunker. Or, uh, we don't care that Instagram is causing like massive rates of depression and suicidal thoughts among young women in particular and teenage girls, or we don't care about this and that. Then Apple starts to become the defacto dad or mom for all of it, where they can just be like, no, this isn't good

SPEAKER_16: for you, but that's not a, that's not a healthy ecosystem either. They shouldn't have that much

SPEAKER_18: control. It's just a hot mess. It would be interesting. Like I feel like if we're, I'm so not in the world of creating apps, but I feel like now there's a lot of people maybe that are creating apps in the generative AI space or the AI space in general, where they have to create an app that is suitable for the app store on an apple in mind, um, before even launching. So there's like this whole

SPEAKER_10: thing where there's like zero limits as a, as a founder creating anything. Right. But the limit for a lot of these people is becoming the app store. And it's kind of sad, I think on, on that point where it's like, you're almost stifling, like the potential of the technology, but at the same time, who's going

SPEAKER_18: to give it like the safety limits. If it, if it was an apple, like who's it going to be? Exactly. And

SPEAKER_01: that's the weird, I think that is the weird push pull that we find ourselves in now, if not Apple,

SPEAKER_67: then who, then whom? Yeah. Listen, it's 2023. The macro picture is a little shaky. It's uneasy out there and tech is getting hit super hard. As such, you cannot afford to lose sales for silly stuff, like not having your SOC 2 right now. If you are unsure about your SOC 2, you need to check out

SPEAKER_65: Vanta. Vanta makes it incredibly easy to get and renew your SOC 2. On average, Vanta customers are SOC 2 compliant in just two to four weeks. Compare that to three to five months without Vanta, huh? And they partnered with over two dozen audit firms who have been trained to file SOC 2 reports directly within Vanta. This is a total no brainer. A bunch of my portfolio founders have used Vanta and they've had amazing experiences. And if you don't have SOC 2 compliance, you can close major customers. One major customer, that could be the difference between your startup thriving or going away. So get it done right now. Vanta's going to give you $1,000 off because you listened to this podcast. Think about it. $1,000 off Vanta.com slash twist. You got to write that down. Put it in your

SPEAKER_102: notes, V-A-N-T-A.com slash twist for $1,000 off your SOC 2. Well, speaking of a changing environment,

SPEAKER_08: let's talk about Salesforce. Yeah. So Salesforce shares jumped as much as 16% on Thursday after the

SPEAKER_09: company beat expectations and provided strong guidance for its next fiscal year. So basically, Salesforce is juicing its stock price with the same strategy that Meta just used. And it seems like Benioff is listening super closely to his new activist investors, and we can get to the strategy

SPEAKER_18: in a minute. But first, let's go over some of these highlights from the Salesforce Q4 earnings.

SPEAKER_01: So let's see. Q4 revenue was up, was $8.4 billion. That was up 14% year over year. They beat on revenue, and we'll get to that in a minute because they had lowered some expectations, but they did end up $400 million over their expectations. Q4 operating cash flow was $7.1 billion. That's up 19% year over year. Q4 net loss, and this is somewhat worth pointing out, was $98 million. But for a company of Salesforce's size, that's like effectively

SPEAKER_03: break even. It's crazy. They reported, I know it's crazy. And those losses did quadruple quarter over

SPEAKER_01: quarter. And yet, you know, when you're talking about $100 million in your Salesforce, like in Q4 revenue was $8.4 billion. It's kind of nothing. They did report $828 million in restructuring costs due to layoffs. In Q4, which is a lot of money, they did that 8,000 employee riff in January, which is a lot of people. And then in Q4, and this is what's really a big deal. Salesforce bought back $2.3 billion worth of its own stock for shareholders and announced that it was increasing its share buyback

SPEAKER_57: program to $20 billion going forward. Any of this starting to sound familiar? Meta?

SPEAKER_09: Right. And you know, this is like I said, at the top, this is really following that same strategy that Meta did, which is when your stock collapses, and Salesforce shares dropped about 60% from its COVID peak, which was back in like November 2021, to its bottom in December 2022. And the strategy of

SPEAKER_10: Meta, number one, is cut spend, number two, lower expectations, and number three is increased share buybacks. And for Salesforce, that cutting of spend really did come with, unfortunately, that 10% cut of staff that you mentioned, that happened back in January, and lowering expectations. It did beat across the board in Q4. And then for increasing share buybacks, it just announced it was increasing its buybacks program to $20 billion. And the results, shares are up 11%.

SPEAKER_01: Yep, they're playing the game. Yeah, here is I mean, really, right? Like, it's, it's like the Scotty from Star Trek game, like, Oh, I just, you know, I under promise and I over deliver. And also I make investors happy. So here's a direct quote from Mark Benioff on Salesforce's earnings call, we immediately put into place an accelerated transformation plan in four areas. One, short term and long term restructuring of the company. Two, improving profitability and productivity. Three, prioritizing our core innovations, continuing to do what we do without getting distracted. A little translation there. And then for this is really the key, a deeper and even stronger relationship with our shareholders, you. And what I think is really happening, I mean, this is this is to me, exactly like the meta strategy, like you mentioned, which is, you know, just play Wall Street like a fiddle, just get that fiddle out and start, start playing. Because I think for a long time, the tech industry could afford to ignore Wall Street, they did not have activist investors coming for them, like Salesforce had activist investors trying to take stakes, right? Like Meta didn't have that they had Brad Gerstner writing that open letter, which was a big deal. But Salesforce literally had active investors sniffing around when Jana Partners comes and starts leaking dish about your internal stuff to the media. And, and calling for that kind of change and forcing it from the inside. Like, you got to play the game. And I think for a long time, big tech didn't have to they could just deliver for shareholders, they were the untouchable category of investment. And all of a sudden, now, they have to change. And they, and they are like, you see which companies are like, turn on a dime, start playing. See, also, by the way, I think that this trend will start to extend to regulation. Also, like a lot of tech companies were very, very slow to get policy teams, public policy teams, if they had them at all, and policy teams that could interface with Washington or do lobbying. So you're seeing like, gradually, the lobbying spend from big tech just go up. And I mean, talk about a hockey stick. Yeah, up into the right. And they're not going they they have had the luxury of ignoring investors with their preferred shares and, you know, founder control, and the luxury of ignoring regulators. And I think that those days may be changing, at least in the short term.

SPEAKER_23: And where do you see, I guess, going all the way to just Salesforce? Do you have any predictions for maybe

SPEAKER_18: over the next quarter, the next two quarters, as they start playing this, this game with Wall Street?

SPEAKER_01: I mean, I do think, you know, it's interesting, Benioff talked about profitability, specifically from the earnings call and said, profitability is truly our number one strategy. And that's my number one strategy. That's what I've been focused on with the management team. That's the number one thing we talked about at the start of every meeting. A key part of what we're doing is making sure that every executive in this company knows that profitability is our highest priority, right? Like, Zuckerberg talks about efficiency. Benioff is talking about profitability. I mean, for years, this has been a company that just seems to print money without even breaking a sweat. So I would imagine that what we're hearing Benioff say is, okay, we're just going to stay focused on. I think that it's like a little bit of jargon to say,

SPEAKER_48: prioritizing our core innovations, it is kind of meaningless sounding. But it also speaks to

SPEAKER_01: focus on like, when things are going well, and companies are trying to figure out how to grow, they start this octopus strategy of acquisitions, let's say, right, like you buy Slack, and you start to experiment with different parts of the business or different spin offs, or this or that or whatever. And when, you know, the ish hits the fan, you have to sort of come back to the core business. I've worked at a lot of places where I was like, right outside of the core. And, and that's like a fun and innovative place to play at a company. But, but it's not core. Exactly. When things get really serious, you actually want to be core. Yeah. And so I do think that there's actually some real strategy contained in that statement. I still know, I know they had a bunch of restructuring costs, I am going to be keeping my eye on, you know, you might think it's a drop in the bucket now, but quadrupling losses quarter over quarter net loss, like, I'm just gonna be watching that.

SPEAKER_138: Just gonna still gonna put that out there a little my internal Jana partners is like,

SPEAKER_18: let's just keep an eye on that. Yeah, so next quarter, we're gonna have to follow back up and check out how that is. Because you're right right now. That's like break even, like you said, so

SPEAKER_108: definitely metric to be watching out for. Yep. Let's talk about ramen for our startup of the day.

SPEAKER_10: Absolutely. I'm a huge fan. My favorite ramen place in New York, actually not my favorite.

SPEAKER_09: My favorite one is called Lemongrass. It's also in New York. But there's one called Ivan Ramen that I used to live near. I don't live there anymore. But I used to live near one. They made a whole Netflix documentary about it. So I'm very passionate about ramen. And today's startup is in me. It's a plant based ramen company. It raised $10 million in a series a from people like celebrity investors, including Usher. So we have to talk about it.

SPEAKER_03: I'm like, this is crazy. Fascinated by this. I know the bubble is not over people. The bubble is

SPEAKER_01: not over $10 million for ramen. The ramen. Actually, the ramen bubble will definitely never burst. And ramen is like having a moment. Ramen is back like ramen hack. It's a thing not to shout out. It's never

SPEAKER_148: been gone. I graduated college like three years ago. So I feel like it's I'm still too close to like

SPEAKER_18: the bagged ramen like era. So it always feels like it's going. But for me, now there's like ramen

SPEAKER_151: hacks everywhere on TikTok. Like all the ramen hacks have been huge. You put an egg in it and you're like,

SPEAKER_18: whoa. We're not talking about nice. We're not talking about, by the way, like fancy ramen. We're

SPEAKER_39: talking about this is a bagged ramen brand that is plant based, which is really cool. A global instant

SPEAKER_02: noodle product. Yeah. So according to TechCrunch, this is so amazing. The market value of the global instant noodle sector in 2020 was nearly $46 billion and is projected to be about $66 billion

SPEAKER_01: by the end of 2027. So the ME company, its latest funding round brings its total investment to $15 million and its flavors include spicy beef. I'm doing air quotes because plant-based black garlic chicken and Tom Yum shrimp. But the funding is going to let them create nine more flavors. They have launched in retailers, including Whole Foods, Wegmans and the fresh market. Amy grew six times year over year in revenue

SPEAKER_18: and sold out seven times since 2021. That is insane. I don't know if anybody else has like ever done this,

SPEAKER_10: but I remember growing up, my brother was obsessed with top ramen that was dry and he would crush it up and put it in a plastic bag. And it was like nuts. Like this kid, my brother eats dry bagged ramen. I, to this day,

SPEAKER_09: I don't know if that's a thing, but he is like one of those people that has like a top ramen sweatshirt. Like it's very brand loyal to top ramen. Wow. Like incredibly. So I'm interested in, um, first

SPEAKER_18: off, I want to know who is at TechCrunch, like crunching, um, no pun intended, these numbers about the instant noodle sector, like good job. Love it. But I am excited to see more plant-based options

SPEAKER_39: being added. This is kind of fun because my brother's favorite is chicken. I actually think that's great.

SPEAKER_01: Yeah. And I do like that. They're starting to be that kind of, cause, uh, cause the veggie top ramen

SPEAKER_156: is disgusting. It's called the oriental. Definitely not the best. Also it's called that. Yeah. Like it's just like, that's not great. I know. And it tastes like butt. It's awful. It's

SPEAKER_148: terrible. I hear the chicken one is just like the way to go. But for this one, um, right now we are

SPEAKER_09: looking at a, a, I inserted in our show notes, a, uh, an image so we can kind of see the packaging and it looks like the spicy beef one is actually like one of the top flavors over it. I mean, so if you have ever tried this or have ever tried dry ramen, go over, let me know it at

SPEAKER_10: underscore Rachel Braun. I'm interested. I want to know, or send us an email at producers at this

SPEAKER_108: week in startups. I mean, crumbled up dry ramen is a staple topping for salads in the Midwest.

SPEAKER_161: No way. Yeah. Oh, absolutely. Like a, like a crouton, like, like how a crouton works?

SPEAKER_01: You use it like a crouton. Yeah, exactly. You put that in your, you put that on your, and I can't believe the O word orient, like chicken salad, but there's like a whole salad that you can make in the Midwest. That's like got oranges and chicken and peanuts and cabbage. And then my mom

SPEAKER_164: used, um, um, no, she used, uh, these like low meat noodles that were like crunchy. Instead,

SPEAKER_33: they came in a container or you can do crumbled up ramen. They're like, and there are a million, even like way worse salads, quote unquote salads. Like there's just, I think I've seen people do it

SPEAKER_10: in like jello salads and stuff. Oh no, now you're pushing it. I think TikTok is going to go absolutely insane with Emmy. Um, I will definitely be trying. It says it's sold at Whole Foods. I, um, now because my packages get stolen all the time and I can pick up my Amazon packages from Whole Foods have been going there more. So I will be on the lookout and I will be reporting back, um, and sharing my thoughts on

SPEAKER_01: this. I want to try it. And also it's important to note here. I think that as we, uh, talk about ramen innovations, we have to talk about, I think it's Momofuku is doing noodles that are like protein

SPEAKER_174: heavy, like lower carb. Have you tried their, um, packaged noodles? They're like, they're a version of

SPEAKER_01: Top Ramen. I want to, but they're really disappointing. It was so disappointing. Don't,

SPEAKER_18: don't try them. I like regular old ramen with an egg in it. Momofuku is great. I've been to the, opened one. I think it's in Toronto. And then I tried like the instant ramen packets from them. I don't know. It's not, if it's, if it's not under like a dollar or two, a package,

SPEAKER_178: and it's getting a little like, I don't know about that one. That's kind of how I feel too. I'm like,

SPEAKER_179: this instant noodles, man. Like, I don't fancy it up. Running a startup is like being a small

SPEAKER_65: market team and you're trying to compete against somebody with unlimited resources like the Yankees and the Dodgers. Well, if you've seen Moneyball, one of my favorite films, you know that using data correctly can help you compete against those big incumbents who have seemingly unlimited resources. One thing that startups haven't really had access to until now is detailed scenario planning. This is stuff that like big companies get to do. There really haven't been tools that are affordable or elegant enough for us in the startup crowd where you can do this easily with org space, O-R-G-S-P-A-C-E. It's people software for software people. Basically, it lets you create plans for deploying the capital that you just raised and money you're making, and then adjusting headcount based on different future scenarios. For example, what if you raise your series A? You get 10 million in the coffers,

SPEAKER_67: right? And you got to deploy that. What if you can't raise right now and you got to make it work with your $3 million seed round? What if your revenue goes up 20, 30% next month? With org space,

SPEAKER_65: you're going to be able to plan for hyper growth. You can plan for rifts and everything in between. Things like cost, skills, DEI, all that is in context so you understand the impact of your decisions. This is the thoughtful way to do it, folks. Twist listeners get $2,000 in credits on org space's pro plans with a 30-day free trial at org space.io slash twist. Org space.io slash twist.

Jason Calacanis: Get those $2,000 in credits. Congratulations to Emmy on the raise. This is actually too a perfect segue to today's series A startup and series A interview, which is all about food.

SPEAKER_148: It is. You spoke to Dylan McDonnell, right? Of the founder and CEO of Foudini.

SPEAKER_01: Foudini, get it? F-O-O-D-I-N-I, like magic, like Houdini. I love this. We spend a long time at the top of this interview talking about that good name. Foudini connects people with dietary needs to suitable restaurants and menu items so they can, you know, go out to eat without stressing. He was diagnosed with celiac disease, which means he's profoundly gluten intolerant. Like it can make you, celiac is terrible. I know people who have it and the constant story is the years and years of suffering before they figure out that they cannot ingest or digest gluten. That inspired his founder journey, had a background as a lawyer and use that to become a better founder and be really careful with this idea of, um, telling people, you know, that they're giving them offers with safer food. They have dietitians on staff. It's a really interesting idea with, as it turns out, a huge,

SPEAKER_70: huge potential audience. Yeah. Excited for them to come to the, uh, the States. Yeah, exactly. They're in

SPEAKER_01: Australia for now. Um, have a listen before we, before we ruin the whole thing, have a listen.

Jason Calacanis: Dylan McDonald is the founder and CEO of Foudini. What a clever name, which connects people with dietary needs to suitable restaurants and menu items. Welcome to the show. Thank you. Thank you

SPEAKER_191: very much for having me. I appreciate it. Foudini. That is such a good name. Like when,

Jason Calacanis: tell me the origin story about the name and then we'll talk about what the company does.

SPEAKER_191: Sure. Um, well, it all, I suppose it, it tracks back to me being diagnosed with celiac disease when I was about 10 years old. So I think that's, I suppose the origin, I suppose, personal connection to this problem. Um, and when I started looking into this first, I was looking at probably specifically through the lens of gluten-free because that was obviously what was impacting me. So there was a few other names at the start that were more in the, in the gluten-free space, but the more I looked into it and the kind of solution we hope to create, the more I realized this was something that affected so many more people than just the gluten-free people like between the allergies, the preference diets, then tolerances. Uh, and so we kind of wanted a word that I suppose encapsulated all of food, but also kind of spoke to the magic of the solution that we are creating, because I think a lot of the people that, that use us do kind of see it as magic and that it is something that shows them exactly where and what they can eat in a way that was never really

SPEAKER_197: done before. So the Eenie is kind of the play on Houdini and the kind of magic around that and

SPEAKER_05: kind of put the two together. And that was the name. Yeah. I love it. Very clever. Um,

Jason Calacanis: so tell me how the actual platform works for people and what, how many different sort of food

SPEAKER_197: needs are encompassed here. Sure. Of course. So how it works very simply is, uh, users either download our app or, or use our web app, which is accessible on our website. They create a dietary

SPEAKER_191: profile that can select from over 50 different diets and allergens. So that's everything from your big 10, your gluten, your wheat, your dairy, your eggs, your nuts, et cetera, true to vegan, vegetarian, true to low FODMAP diets and the kind of different food groups within that down to keto, down to coriander, garlic, onion, you know, would you be amazed the amount of people that literally

SPEAKER_203: just use us for coriander? It's one of these genetic things that people either love it or hate it.

SPEAKER_131: Yeah. So the soap thing, do you have kosher? Exactly. Just say that again. Do you have kosher?

SPEAKER_203: We, we, kosher is a category within there. We actually haven't created as a dietary profile,

SPEAKER_191: so you can, uh, filter by that, but we're trying to actually build out the, the dietary profile element of that at the moment. So we'll get there.

SPEAKER_204: So that's what you build out first dietary. Got it.

SPEAKER_191: Correct. And I think just on that point, it was the key around this was that it had to be customizable, like in this world of, of, of, of everything on demand and, and like the way Spotify and, and all Netflix have the personalized recommendations. We really wanted this to be the dietary profile that lives with them through the entire experience. And so once that dietary profile is created, the app then automatically shows you the restaurants that are most suitable for you. And also an exact breakdown of what they can and can't eat on the menu. But the crucial part around that as well, is that it's all dietitian reviewed and approved. So we have a team of dietitians who literally review every single menu item. I think we've over 200,000 menu items on the platform at the moment. And all of those have been manually manually reviewed by our dietitian team over the top of the technology, which kind of does a lot of the data entry.

Jason Calacanis: Wow. And then it's my understanding. It also contains information about the current kind of cross-contamination possibility.

SPEAKER_197: Correct. Yeah, correct. So that as part of when we sign on new partners, our dietitian team work

SPEAKER_191: directly with the chef or whoever owns the menu to go through again, down to the level of, you know, what, what seasoning do you use on your fries? Is it cooked in oil or butter? What are the cross-contamination procedures in venue? And all of that information is filtered through on the app as well. And I suppose for us, it's obviously we're trying to drive positive change in industry to make restaurants more aware of just how many people are impacted by this and all the benefits that come with catering to this audience. But it's also just about reflecting their reality. Because we have some users who are super, super sensitive, and even a low risk of cross-contamination is not good enough for them. And that's fine. Like those people just need to go to certain places. But equally, we have a lot of preference and intolerance people who can tolerate a certain level of cross-contamination. And just by providing the kind of, I suppose, reality of each venue, it allows users to make informed decisions on what level of risk they're

Jason Calacanis: willing to take. So how do you think about the size of audience for this? I mean, certainly everybody has some kind of food issue ranging from it will kill me to I'm just picky about this thing. Yes. Like, is that how you imagine, you know, that at least it's a two sided marketplace.

SPEAKER_00: But when you imagine the potential TAM, are you thinking like everybody who has some opinions about

SPEAKER_191: what they want to eat? It's a funny one. Because again, most people who are not familiar, I suppose, with this area always underestimate the size of how big this market is. Like, there's a few stats around, I think they estimate about 68% of the global population is lactose intolerant, right? So if we were to try and take stats like that, our TAM would be, you know, two and three people on lactose alone. But the way we've kind of defined it, we're looking at kind of stricter food allergies, intolerances and kind of lifestyle diets that people actually follow. And we have TAM at one in three people in terms of the Western population, at least, of the Western world. So, and then I suppose, when you actually build on to that, it's not just even the one in three, it's actually the partners, the families, the co workers, the friends of those people as well. Because another point we try and make to a restaurant is if, you know, I've been in working situations where there's 10 of us who are going to work lunch, I'm the celiac. So we have to go to somewhere that caters to me. So if you don't cater to me, you don't just lose me, you lose my nine co workers as well, right? So on the restaurant side, I would actually argue it's a much bigger market than just the one in three. But in terms of the direct

Jason Calacanis: it's one in three. Wow. And then tell me about the dietitian part of it, because that feels pretty

SPEAKER_224: crucial in terms of being able to reassure people. But I wonder if it also creates like a sense of a guarantee that could be kind of problematic for you. It's a fair point. And sorry, just as well,

SPEAKER_191: backtracking to that last point for a second, there's 32 million Americans who have food allergies as well, by the way, that's a that's confirmed stat by by fair who are the allergy organization there again. So again, yeah, it's there's those numbers are backed up by by government organizations. But no, to answer your question, when we first looked at this, you know, we were we were really trying to drill down into how can we actually provide people with information that they'll trust, because that is the key point here. And it's crucial that that is the case. So we through our processes and QA we have, we take every effort to make sure that the information that we have is accurate, right. But at the same time, we're not preparing food, we're not in the kitchen, we can't stop a chef getting outside the wrong side of the bed and accidentally putting the wrong ingredients, something these things, unfortunately, happen, like, a lot less often now than I think they once did. But at the same time, they happen. So and I'm a lawyer from a past life as well, right. So in terms of our disclaimer that, you know, it just has to be on there is that we can't guarantee, we just simply can't. And I don't think anyone ever will be able to. But we are disclaimer makes it clear that here are our processes, we make every best effort in terms of initially onboarding restaurants, but also in terms of the ongoing QA, we do our best to ensure the accuracy of the information. But if you have a severe food allergy, you always need to flag it. And that is something that will never change. If you have a nut allergy or anaphylactics or anything like that, no matter how much good data you get, you still have to tell the restaurant that I have this allergy because that puts them on notice that this is a thing and to make sure they take every precaution. So even for those people, I suppose we still help them pick the restaurants that they can know what a certain confidence have options and are suitable for them, but they will always need to flag it. And that will never change.

SPEAKER_57: Um, I want to ask you about the origin story, you were a lawyer, and you became a founder. And

Jason Calacanis: clearly, there's a personal connection celiac can make people incredibly sick for a long time and not even know why I would imagine. Um, or I am told, but like, what, tell me this story.

SPEAKER_235: Sure. You know, lawyers, not like a crappy career that people usually just

SPEAKER_203: no, no. Uh, well, I was actually born in the States. So both my parents are Irish,

SPEAKER_191: but I was born in Philly and we moved back home, uh, when I was about four. And so obviously accent portrays, uh, I grew up in Ireland and went to school there, et cetera. And I had been sick for a few years, say between that kind of four and 10 age. And again, no one kind of had a clue what was wrong with me. Celiac gluten-free weren't even words in the dictionary, hardly in the States, not a mind, you know, in rural Ireland, you know, 20 odd years ago. So eventually someone kind of pinged it, did the blood tests came back positive. And I think at that point, and anyone listening or watching who, who was eating gluten-free food 20 years ago, will remember that it was awful. Like it was really, really bad. You wouldn't feed it to your worst enemy type stuff. And obviously that kind of improved a lot, um, over the last 20 years, but it was just the dining out issue that remained. But I moved to Australia, I qualified as a lawyer. I, uh, I worked in a top tier corporate law firm and then moved to Australia where I worked in house over here. Um, and I think it was kind of traveling on the way over. I went to Japan for the rugby world cup and I went to Bali as well. And even just there, like trying to navigate food with language barriers. I was just, I couldn't get my head around how in this day and age, this was still such an issue. And I got to Australia, COVID hit a little bit of extra time on my hands, probably at home. And I just started doing some research. And the more I looked into the size of the market, the more I looked into the existing solutions that, that were out there, the more I was like,

SPEAKER_216: this just doesn't make any sense. And that was kind of the genesis for, for starting to, to deep dive deeper and, and try and create the solution.

Jason Calacanis: Wow. Um, okay. So tell me about the business model. You're a two sided marketplace who pays whom.

SPEAKER_191: Yeah. So we, we've monetized the restaurant side at this point in time. So our restaurant partners pay us a flat subscription fee to be listed. Uh, that includes, I suppose, everything from the app listing and driving them new, new and loyal. Cause that's a key point with this audience. And anyone listening who has a food allergy will know when they find somewhere that they like, they tend to keep going back there. Um, so driving them new loyal users. We also have an in venue asset set up, which allows people who walk in off the street to scan a QR code. And in five seconds, see a customized version of the menu to them. And from a restaurant's perspective, that also, uh, it saves the staff a lot of time dealing with a lot of the kind of nitty gritty questions that also reduces the risk of errors happening. Cause that was an issue, particularly here in Australia post COVID, there was a massive shortage of hospitality staff and a lot of kind of younger untrained staff who just weren't really across this issue. Um, and it's just a better customer experience. And then we also give them the data on top of it in terms of these are the people coming in. These are their dietary profiles. These are the items they're looking at. And that can also inform, you know, menu creation decisions as well. So yeah, I think that the restaurants tend to get a good bit of value from the subscription, which is, is priced pretty low at this point in time as well. How much does it cost the restaurant? It's $49 a month. Yeah. So we, we, we priced it in quite low and that's, I suppose, again, we're, we're trying to aggressively scale through Australia. We we've, you know, we've, we've expanded now into nearly every state here and we, um, we just wanted to make sure price wasn't a barrier to entry at this time to really just, just get as many on as

SPEAKER_57: possible. Right. And then on the consumer side, it's free, completely free to download, completely

SPEAKER_191: free to use. And that's the thing. It's, it's funny, kind of the most rewarding thing or the most enjoyable sometimes for our team is doing events and you kind of, you, you, you see people walking past like at a celiac expo or, or a food allergy expo and they're kind of eyeing up foodini and they're like, I wonder what that is. And then you kind of hit them with the kind of the tree that shows you where you can eat, what you can eat dietitian reviewed. And then the, and it's, and then they're like, and how much, and then, and it's free. And they're kind of like, all right.

SPEAKER_245: I can imagine that from the perspective of like the person at the conference, I will tell you as

SPEAKER_191: an investor, you should probably charge people. Yeah. Yeah. Look, it's, it's one of these things we've, we have, there's so many revenue streams here. We could look at right in terms of freemium models. And we have a roadmap full of extra features to add on here. Like we've a group search feature that allows a family with four different food allergies to create a group profile and it will spit back the restaurant more suitable for the whole group, for example. And that can obviously be applied around. So there's a lot of stuff that we could at a point in time put behind a paywall. But for me, I always want the core functionality to be free. Um, where again, at this point in time, it's, it's really just about scaling, getting as many people using it and, and, and I suppose, improving their lives and getting back out there and, and, and eating and enjoying the kind of hospitality industry, uh, as opposed to worrying about putting too much of a paywall in front of them

Jason Calacanis: at this point in time. Yep. Great point. Um, how, tell, talk to me about the dietitian part of this.

SPEAKER_195: How many dietitians do you have? How long does it take them to review a menu?

SPEAKER_191: Sure. Yeah. So we have a team of six dietitians. Um, and I suppose a big part of this at the start was everything was done manually. That entry, the whole chevang, it was, it wasn't a scalable solution, but we spent about two years building out, I suppose, proprietary technology and using AI to enable us to do a massive proportion of this job completely, uh, automated. So we have tech that can scan menus, auto tag it, auto populated in our system. And for the dietitians, then it's, it's a lot. And the, the, you know, the success rate or the correct rate in, in, in that process is, is 95% plus or something like that on average. So the dietitians then job is to come over the top of that and ensure that it's correct. So the data entry is gone. All that side of things is gone. And it's, it's more about the conversation with the restaurant about the things that can't be on the menu about like cross contamination and seasoning and stuff like that. And on average, it takes between 30 to 60 minutes per restaurant onboarding. So it's pretty quick. Again, that's come down a lot from what it was at the start. So I think that the kind of operational efficiency here was such a crucial part of this business to allow it to be scalable. Um, and that was one of the reasons my, I brought in my, my co-founder who Timo Kugler, like he had 20 plus years experience in, in operational scaling in both corporate and startup world. So he really drilled into how can we make this super scalable, super quick, but also crucially still correct. And of course, there's some menus, you know, if it's an Indian restaurant with 200 items, it's obviously more than an hour to, to, to run through that. And then equally there's some small cafes where it could be 20 minutes. So the average is between 30 to 60 minutes per restaurant.

SPEAKER_195: And then in terms of evaluating the cross contamination, is that, is that an actual

SPEAKER_57: in-person visit? It depends. Whether they call or, you know, it sort of depends on the restaurant.

SPEAKER_191: It can depend. So we, if they, if they guarantee a very high level of non-cross contamination risk, that usually triggers an in-person of an in-person inspection to confirm that. But if it's, if they kind of, which is 90% of our restaurants would say, look, we have a process in place, but we can't guarantee that there's no cross contamination. I think that's the kind of most restaurants go with that because that is just the reality of most kitchens. In that case, we, we, we, we confirm that over the phone and we flag that on, on, on the app, but we don't need to conduct a physical inspection. It's only if they say we guarantee none, that's when we need to confirm that because we won't put it on the app otherwise.

SPEAKER_245: Smart. What, I wonder just like, what are the hardest allergies or illnesses or food

SPEAKER_57: intolerances to work around? Like who gets the shortest list of restaurants to go to?

SPEAKER_203: That's a good question. I think like the hardest ones in my mind are all still the anaphylactic

SPEAKER_191: ones, right? Because they're the ones with the highest level of risk. So anytime there's anything, not shellfish, you know, egg, anything that can cause anaphylactics, that's always on our end, the one that we're always the most cognizant of, because it's the one that can trigger usually the, the, the most severe response. But in terms of what gives people the least options, I think it's probably the FODMAPS diet, right? It's the low FODMAPS diet. I think when people do a full low FODMAPS diet, I think that takes so many foods. Are you familiar with what

SPEAKER_203: that is? What is that? It's essentially, I wasn't not too long ago either. Don't worry. My dieticians had to spend a while. Yeah. FODMAPS. Yeah. So that's, it's kind of like, it's a diet that a lot of people go on if they're suffering from IBS or other kind of gut health

SPEAKER_191: issues because it's essentially a food elimination diet. So it essentially, you eliminate pretty much 90% of foods and then slowly start reintroducing things to essentially figure out what it is that is triggering the gut health problems. So it's one that's actually risen in popularity quite drastically over the last year or two, especially in the same way keto exploded in the US. I think there's something like 13 million people on a keto diet in the US at the moment. It's everybody I know. I can tell you. Yeah. Well, there you go. All right. So yeah, keto does take a bit out of play as well, but I think, yeah, the low FODMAPS diet, when they go on

Jason Calacanis: the full elimination, takes a lot out. FODMAP is what it is. Correct. Yeah. And my producer put in Slack, by the way, oh my God, those kids can only eat potatoes. I said it better myself. Exactly. So talk to me a little about that roadmap. Like what, what do you imagine happens in the future? Could people order directly from the app? Can they filter by? Do you include food groups? I sort of

SPEAKER_57: interrupted you at the top when you were talking about actual dietary restrictions, but do you also

SPEAKER_203: have like kosher or halal or? Yeah, we do have kosher and halal, but built in as categories within that we haven't yet. We're working towards getting them at that kind of initial dietary profile point,

SPEAKER_191: but you can still filter within the app by kosher and halal restaurants, right? So you can still do

SPEAKER_266: that. So you'd be like gluten-free first, kosher second. Yeah, exactly. It's kind of,

SPEAKER_203: there's essentially a screen where there's like 50 different options and you just pick as many as

SPEAKER_191: applied to you. But at the moment, halal and kosher are actually more within the experience as a category in the same way you'd filter by burgers or American food or pizza, you can filter by halal and kosher kind of within the experience itself. But in terms to answer your question, the roadmap, well, for us, you know, we see this, we're, we're, we're super focused right now on the hospitality industry, because I think it was the one that, that we saw needed the most help, the quickest, but in terms of how we see Foodini long-term, we just see it as the dietary profile that helps you navigate life. Whether that is in a stadium, on a cruise ship, doing your groceries, you know, you know, at a, at a birthday party, at a, at a friend's birthday party, you know, wherever that might be, that, that dietary profile will help you navigate life. So I think, you know, we could, I could talk for hours in terms of all the ideas and the roadmaps and the verticals that, that we have on the, on the horizon to target. But I think it's one of these ones where if we, we don't want to get too distracted too early, you know, we're doing a really good thing right now in hospitality, and we really want to nail that and get that right before we start focusing too much on the other verticals. But I, we think that, yeah, the sky is the limit in terms of the opportunity of, of, of incorporating this into where food exists, which is in a lot of places.

SPEAKER_57: Amazing. And then tell me about the funding history.

SPEAKER_191: Yeah. So we, we did a family and friends round at the very start. And then last year, we closed over a subscribed pre-seed round. It was led by Antler, who are global VC funds. Some people might be familiar with them. And we were actually, you know, I think it's, it's both in terms of our team. And in terms of the advisors we took on and investors, you know, we were very, we really wanted them to be aligned to the mission that we have. And so the cap table, we ended up with, you know, in addition to be just being very strong, generally, I think nearly everyone was quite aligned or had a personal connection to that mission. And like, even if you take our team, for example, the latest team member, Jason, well, he was actually the former head of marketing for Google, Australia, New Zealand. So I think he also launched Etsy and eHarmony in Australia, but he's a kid with like, very, very severe food allergies. They've been to hospital like 15 plus times with EpiPens and stuff like that. Yeah, like they've had it bad. And Timo, similarly, he's a kid with two, two kids, two boys, both have food allergies and a vegetarian wife. Like he always jokes when they're going out to dinner. It's like a bad start of a bad joke, trying to talk to trying to explain what's happening to a waiter. Our marketing lead is a tier one agency background, but also has a master's in dietetics as well. And it's also a registered dietitian. So everyone on both the team and also kind of the cap table board of advisors, all we think are very aligned with our mission. And we were very deliberate in making sure that we chose people who were that way. Because everyone is just so much more, I suppose, aligned, dedicated, willing to help when that is the case. So yeah, we closed that round last year. And we will be doing a seed round this summer. And that's just very much to kind of drive the growth that we're already seeing.

SPEAKER_195: Amazing. When are you coming to the US?

SPEAKER_191: So we're coming to the US later this year, we've already been accepted into and actually, if there's anyone who from Australia who is thinking of launching into the into the US, it's called the OzTrade landing pad. I'd highly recommend this program. It's essentially an Australian government initiative, which helps companies get set up in the US. So we're going through that program at the moment. And you know, we have a base in San Fran through it, which is amazing for when we're on the ground over there. So yeah, top tip for any Australian people trying to get into the US that that program has been excellent so far. But later this year, we're going to launch some test balloons, and essentially start testing the US go to market strategy. We have our Australian playbook. But as everyone will know, the US market is quite different. And we really want to make sure that we suppose approach it in a smart and deliberate way. So we'll start with a few test balloons in a few locations that we think make strategic sense and then and then

SPEAKER_57: scale it out from there. That's great. Sounds like also a good tip for investors who are trying to scout hot startups coming from Australia. Liz, you said it, not me. Yeah, exactly. Dylan,

Jason Calacanis: thank you so much. Dylan McDonald is the founder and CEO of Foodini. It is a pleasure to talk to you.

SPEAKER_203: Congrats on all you've built. Likewise. Thank you so much for having me on the show. I really

SPEAKER_00: appreciate it. All right. And then we're not done yet. We're not done yet today. We have

SPEAKER_10: OK Boomer. Who did you talk to today? I talked to one of the coolest people ever, my friend Michael Tan. He is an author, which is really cool. I've never had an author on my segment before. And his book is called The Death of Uncertainty. The first time I read it was back in Miami Hack Week, which I also did an episode on where I talked to my friend Eric Button about a year ago. And almost a year later, Michael Tan is on. And The Death of Uncertainty is a book about basically this world that is dictated by an AI algorithm that can tell people how to make the best decisions for their lives. So everybody has these perfect lives, except for one person who realizes there's a suboptimal way of life. And to understand the highs of life, you have to understand the lows. Overall, great book. We didn't just talk about his book, though. We discussed his thoughts on AI, predictions about AI, how he was pretty early to the trend, early enough where he could write a full book about it, the predictions about the future of AI. Like I said before, especially within, I feel like this weird generative space is something that he was really interested in. And overall, great discussion. This is probably one of my favorite ones.

SPEAKER_170: It's a novel?

SPEAKER_89: It is.

SPEAKER_170: The Death of Uncertainty, yeah. Sounds like it.

SPEAKER_89: It is. And I don't really read a lot of sci-fi, to be honest with you. So I thought it was really

SPEAKER_10: great. But I would love to hear other people's thoughts who read maybe a lot of sci-fi books. I think it's a genre that I'd like to dive a little bit deeper into now, especially with AI, I guess, getting bigger. But this was my first sci-fi book I ever read.

SPEAKER_299: Oh, my.

SPEAKER_03: I know. I'm still late to the game. I have a whole, I have a million book lists for you. I'm so excited. Welcome, Rachel.

SPEAKER_266: Your life is about to change completely. All right, let's listen to this interview. I'm excited.

SPEAKER_10: Thank you, Michael Tan, for joining me today on OK Boomer. I met you in a really cool way. But before we get into that, Michael is the first person I've had an OK Boomer, I think, that has ever written a book, which I think is insane. I met Michael when we were in Miami for Miami Hack Week. I was actually at a like a hacker house that was made by another former OK Boomer guest named Eric Button. So shout out to Eric Button at an event created by Jadon of Miami Hack Week, who again, another person that has come on the pod. So Miami, I have a lot of I have a lot of hope for, I guess, as a city. So thank you so much again for coming on. Super excited to break down your thoughts about AI, which is the topic that really encompasses your whole book.

SPEAKER_303: Yeah. Thank you so much, Rachel. I'm well, that's almost exactly a year ago.

SPEAKER_305: I'm very glad I ended up going to Miami and meeting you and Eric and everyone else.

SPEAKER_308: Yeah, I'm super duper pumped to have you on, too, because I was nervous that it was like, oh,

SPEAKER_10: my gosh, I wonder if too much time has passed or if you even remember. So pumped that you answered my Twitter DM. But I read your book pretty much immediately after Miami Hack Week because we were talking about it. I think like right then and there I bought it on like my Kindle. Your book is called The Death of Uncertainty. It was really interesting. So basically, without taking giving everybody too many spoilers, an AI algorithm in the book determines like what the best choices are for people's lives. So nobody makes mistakes and everyone's living like these ideal versions of their lives. But the book follows the man who finds out about like suboptimal choices and what it's like to actually face like adversity. And then for a lot of the book, he reflects on the system and explores what it means to like actually have free will and everything that comes along with it, like film it, human nature, things like that. Overall, phenomenal book, really interesting. I can't believe you wrote this in college, you're only a few years out of college, actually. So I guess this book is fairly recent. How did you start writing about AI before AI got cool? What got you interested in this?

SPEAKER_313: Yeah, I like to say that I published the book a couple years before ChatGPT came out. But

SPEAKER_305: yeah, it was really my second semester of my senior year in college. And you know, that's a semester where people have a lot of free time. It was really spontaneous. I was having a dinner conversation with a friend. I was always interested in like AI, ML, I was a stats major. And yeah, we just had like a dinner conversation about like, what if there was this AI that could just like tell you the right thing to do? You know, we people do ask like ChatGPT for advice these days. So it's kind of cool to kind of see it manifest in reality. But we were just having a conversation. And the very next day, I got introduced to a professor at Georgetown, Professor Eric Custer, who helps students write books. And he came across my profile, I've done some writing online. So, you know, we got set up and he's like, I think you would, you should try this and we can help you become an author. And I was like, well, there's an idea that's in my head that I want to spend some time exploring. And this might be a great avenue to do it. So incredibly spontaneous timing. I just happened to talk to the book professor. I had a really good dinner conversation and we're off to the races.

SPEAKER_70: That's awesome. That is definitely a unique way, I guess, to getting to write a book. Jason,

SPEAKER_10: the original host and now co-host of The Speaking Startups before Molly even came on, he wrote a book. So he wrote a book called Angel. So quite a while ago, and he was much further along in his career than you were. So hearing about writing a book in college must have been a really difficult thing considering that exposure to AI couldn't have been that high, even if it was your interest. Like how did you have this much insight into AI? Was it really just like your interest that fueled you or were you really like, it was just something that became like encompassing?

SPEAKER_305: Yeah, I think, you know, it took some ML courses, just like talk about AI because tended to hang out with people who were interested in it. We were a little nerdy group and so we were just like randomly have little musings about like what the future could hold here. I think everyone was always excited about AI. I think really just in the past few months, it's gone from an excitement to something that feels more tangible and you can really see that excitement manifesting. So I wouldn't say it was like, not popular or hard to conceptualize. It was just like, not as much as maybe, you know,

SPEAKER_70: very recently. Yeah. And with the rise of AI, as we're seeing with chat GPT and other generative AI tools,

SPEAKER_322: um, do you think that your book or any aspects of your book are starting to form into like the current

SPEAKER_305: reality that we live in? Yeah, I think, uh, it's really interesting to see because my book, the AI chatbot essentially is called HAPOC. It's like an acronym, but, uh, it's, uh, it's like a, you know, extrapolate chat GPT, a hundred, 200 years. And then you have a chatbot that can literally have, like do anything that, you know, people ask it for advice. It gives those people, you know, the perfect answer. Um, every time it tells them, you know, what choices to make, they are always directed towards the path that's optimal towards, you know, their goals. Um, so you can kind of imagine that kind of world. It's, it's more of like a thought experiment where, you know, if chat GPT becomes so good or, you know, LLMs become so good that they can just tell you what to do, you know, and be correct in that, you know, advice without fail. Like what would society look like? So, uh, it's cool that, you know, uh, you know, two years since after the book's come out, kind of the first popularized instantiation of, you know, a chatbot that people go to, to ask questions to, to ask for advice has come, um, to fruition. Um, and as that improves, we can only imagine, you know, how good that will become and what, what the, uh, ramifications could

SPEAKER_09: be for, for society. Yeah. And we're not there yet. Obviously we're still starting to see, uh,

SPEAKER_10: chat GPT, um, competitors, maybe not produce the most accurate types of content. Uh, like Bing, I know had a lot of slip ups with its, uh, new Bing latest AI chatbot. Do you think you'd ever write a

SPEAKER_70: book about the opposite of your AI chatbot? So maybe it gives people like the worst advice?

SPEAKER_309: Like a evil Sydney gun, gun, rogue. An evil Sydney. Oh, Sydney, by the way,

SPEAKER_10: for people that don't know, Sydney is what Bing's AI chatbot was internally named. And then it like, very creepily started telling people like identifying as Sydney. Um, even though that was

SPEAKER_70: something that was supposed to be internal, we think weird, weird, weird. I think we covered it

SPEAKER_308: already on this weekend startup. So if you want to like, go, go back to a previous episode, but that to me, that's creepy. That's super creepy. Yeah. Yeah. I certainly thought about it. I mean, you

SPEAKER_305: could, it could weaponize itself, you know, it could, you could give bad advice, but I was, I was more so thinking about like this law experiment. It's kind of similar to like, it was a little bit inspired by Brave New World, this other kind of like dystopian fiction novel that's, um, and the idea is like, what if you had like the most benevolent instantiation of AI, you solve all the liven issues. The AI just, you know, is fully directed towards optimizing human flourishing, helping people make the right choices. Um, what would that look like? Uh, even at the, the most, you know, benevolent instantiation, um, there are some like interesting themes on, on free will, adversity, all the things you alluded to earlier. Um, um, so just like giving AI every benefit of the doubt, um, that was kind of like the, the, the framing of this kind of like ball experiment by

SPEAKER_70: explore. How do you think we can be changing our approach to AI now as we're starting to see

SPEAKER_10: advancement? It kind of feels like we're in this J curve, right? Where we're kind of at that base, where we're starting to see like the acceleration happen, kind of like we did with like crypto and web three, not that AI, um, or crypto and web three have, um, necessarily like the same trajectory or the same outcomes, but what, what could we be doing differently to approaching AI as a tool that they're releasing to the public right now? Um, is it too soon? Could we be doing something else?

SPEAKER_344: Yeah. I don't pretend to be an expert on like alignment or anything, but it seems like with, with

SPEAKER_305: things happening like Sydney alignment might, it is, you know, a really important point that we should invest in, you know, you know, the book talks about a benevolent instantiation and, and it has its benefits. Um, but I can only imagine what a, uh, malevolent, you know, instantiation of AI could do. Yeah. Yeah. And, and so, um, I think solving that alignment issue, you know, making sure that our AIs don't become these scary Sydney bots taken to the extreme. Yeah. Um, that, that sounds like an existential risk. And, uh, I feel like that's like something that because of how big that downside could be is worth, uh, worth trying to solve. And Jason and Molly have kind of talked on the show about this,

SPEAKER_10: about, um, putting guardrails in AI and is that limiting freedom of speech and open AI had guard rails, obviously, but pretty strict ones where, you know, you couldn't say like certain things about politicians or else you'd get kicked off. You did it three times, things like that. How do you feel about guard rails being used with AI right now? Yeah. I, uh, I don't have like a very strong

SPEAKER_305: philosophical, dogmatic view one way or the other. I think there, you know, there are times when guardrails make sense. Obviously, there are some really bad things you can ask an AI about. Um, and I genuinely think a general principle is just to kind of have them align with the, uh, the norms around speech that we've, you know, we, as a society have, have already set up for ourselves, um, in the U S those norms, you know, probably aligned to what's outlined in, in, in our laws and in our constitution. Um, so I don't know if there's like a role for the creator and AI to, to mandate a certain, you know,

SPEAKER_355: guardrail. Um, but I think we can use like what already exists as an, as an, as a norm for speech guardrails in our society, um, as a, as a good starting point. Yeah. And you said you don't have

SPEAKER_70: a really strong point on that. What do you have a strong feeling about philosophically in terms of AI?

SPEAKER_305: Yeah, I think it's, uh, it's, it's really revolutionary. I think, um, the main like societal longterm, you know, I guess prediction, if you want to call that is just like, there's so much of an effect that can happen, like productivity. Um, you know, you can think about, you know, the new generation who grows up with AI may probably be like way more productive than, than ever, just as, you know, our generation grew up with the internet and Google, you know, from an economic productivity standpoint is probably more productive than the previous generation who didn't have the internet. Um, you know, you have the creation of creative assets from books. Like if I had Chachi Petit, this book would be written probably a lot faster, books, paintings, you know, you know, movies, films, like the creation of creative, creative assets can be like more efficient, faster, better by an order of magnitude. And also just the creation of like economic value, you can imagine, you know, like in theory, you could, you could scale a software business as like a one or two person team now, because you, you have all these tools that can automate your, your sales marketing, even some of your engineering functions. Um, so you can really see this like explosion in like overall productivity. Now I think it remains to be seen whether that kind of overall productivity growth is like evenly distributed across society. You could imagine like median productivity actually coming lower, you know, exasping kind of like the tails of the bell curve, um, could have some effects on wealth inequality. Um, and then you also see kind of like, I think demand for like knowledge workers could likely decrease, you know, you, you can see in a future where, uh, one consultant or lawyer or salesperson or data scientist can now do the work of 10 because they have these like empowering AI tools. You saw this like literally yesterday being the consulting firm announced a partnership with open AI. We'll see how all that stuff pans out. But I think like the implications on like overall societal economic productivity and knowledge work are pretty

SPEAKER_70: interesting. I completely agree, especially on that productivity side. Some one use case that I used,

SPEAKER_09: um, an AI chat bot for was I asked it to like elongate bullet points in my resume and like paragraph format to update a portfolio, um, that I wanted to have. And I was like, Oh, the, the bill points don't look that

SPEAKER_10: good. And I, yeah, I had to edit it a little bit and write it through grammarly maybe, but overall it was like such a great way to save my time. Um, and my mental energy and that decision fatigue that comes along

SPEAKER_09: with it. And do you think that, or I guess you see AI more of a tool than a replacement, or do you

SPEAKER_305: think that it's a, a replacement? I mean, I think, uh, at the, at the current moment, you know, it's, it seems like a tool. If you have something that is 80, 90% right all the time, I think that's very useful as, as a tool to augment someone's, uh, productivity and you can like look over what it generates and check it if you like it. Yeah. I think as, as AI progresses, gets closer to that, you know, it's a hundred percent accurate. That's when you can just like trust it to, to replace, but I don't think we're, we're there yet. I think, you know, right now. Yeah. It's, it's, uh, you know, it's like if, if I, I wouldn't count on chat GPT to like be my lawyer, but I think a lawyer could count on chat GPT to make their work, you know, 10 X faster. Oh, I believe we actually saw that. I believe

SPEAKER_10: there is like an AI lawyer, um, coming up in court to try to like help eliminate bias. But yeah, I, I guess that, that is something that we might not be seeing for years and years and years to come, especially as people become a little, I've actually started seeing people like this weird

SPEAKER_09: phenomenon happening mostly over on like the internet, um, with people around our age where in the beginning when Dali was around, people were like, wait, this is like, so cool. It's so

SPEAKER_10: interesting. And now you're starting to see people be like, oh, AI, like, I don't trust that. How do you think we can get people to trust technology more?

SPEAKER_70: Or I guess should, should we be trusting technology more? That's another question.

SPEAKER_381: Yeah, I think it's a, should be a function of how trustworthy the technology is. Um,

SPEAKER_305: I think, you know, when technology proves itself to be, to act, to be accurate, to be helpful, people trust that you see this, you know, a lot of programmers trust GitHub copilot and what's generated, but that's only because they've seen that what's being generated is, is accurate and, and helpful and creative to, to their working lives. So I think, you know, as people continue to use chat GPT or, or other, you know, you know, AI powered chatbots, uh, as they find it adding value to their lives, they naturally, you'll naturally tend to trust it more. I think I've seen that myself, you know, the first time I ever heard of it, you know, you're, you're a bit skeptical, like, can this really do this? And then you like use it a few times, you're like, wow, this is actually helping me improve my resume or whatever it is. Um, and the more you find, feel that value, um, I think naturally you, you trust it more. Yeah. And by the way, the robo lawyer, if anybody's

SPEAKER_10: interested, again, we've covered it before on the speaking startups, it's called do not pay. Um, and there are some pretty crazy stuff and allegations happening to it on Twitter as well, as well. So interesting thing to check out, um, when you're done listening to this recording, of course, um, and Michael also, if people are interested in learning more about, or thinking about AI more on like this philosophical level that you really bring the table with your book, um, how would you recommend that? I know you said that this was just like a good dinner conversation, but for people that aren't necessarily surrounded by others that know a lot about the topic, how can people start really

SPEAKER_305: thinking deeply about this subject? Yeah. I think, uh, I mean, shameless self-plug, uh, you could check my book out. Yeah. Uh, yeah, it's on Amazon. You can search for the death of uncertainty. It's linked in my Twitter bio. If Rachel, you want to share that afterward. Um, what's your, what's your app handle to the self-promo at, at the man, Mike tan. So. Oh, I like that. The man Mike tan, because it's my name and man in tan, right? I like it. So perfect. Yeah. Is it only available,

SPEAKER_70: um, on ebook or is it available as a physical one as well? You can order a print or ebook print, print looks good. You can't see it in the virtual. Put it in front of yourself so we can see it better. There you go. I only have it as an ebook. I think you need to do it as an audio book. I'm ready.

SPEAKER_308: I'm ready for it to be an audio book. Okay. I think that would be cool.

SPEAKER_344: I'll, uh, so yeah, that's a good idea. Well, we'll explore. Yeah, I know. Um,

SPEAKER_70: no, but in all seriousness, I think that's a good, I think you're right. I think that's a really, really good, um, way of learning more. I also think like you said, you mentioned already the brave or excuse me, not, not the brave new world, brave new world. Um, which is written by an author with the last name Huxley, I believe. Um, books like that are really interesting, especially cause that book was written like in like the, I think it was published maybe in the 1930s,

SPEAKER_391: something like that. 1920s, 30s, something like that. Really a long, long time ago. Um,

SPEAKER_70: but it was a book that actually came up after I read your book. I was like, Oh, I want to learn more about this. So it's funny that you mentioned that one, but yeah, definitely would recommend

SPEAKER_322: checking out Michael's book. And, um, where in terms of AI do you think we're going to be in the next

SPEAKER_305: five years? Wow. That's, that's tough. I feel like it's really hard to, to predict anything that's, you know, pressing exponentially, um, tend to underestimate exponential growth. But if what they're saying is true, if, you know, GPT four is going to be trained on, you know, 10 X more data points in GP three, um, or whatever that, that order of magnitude is, um, you know, you can, you can see, uh, these chatbots just, you know, getting better and better. Um, I think on the chatbot side, as everything we talked about, the, the utility, the accuracy, that the value it's, it's providing to people, um, hopefully we'll just like get better and better. You know, you might get in five years to that state where Rachel, as you said, it's, it's less, it can actually become a replacement and not just a tool. Um, and then I think on, on the people side, um, the users, like, I think we'll see some really amazing things and we already are seeing some really amazing things built just like primarily using AI or, or chat GPT. Um, I read, I think I saw this like crazy stat that like 50% of the code written by programmers on GitHub copilot is written by GitHub copilot and not themselves or something like that. Wow. Um, okay. So yeah, seeing it, seeing

SPEAKER_70: creation happen directly on the platform. Exactly. Yeah. I think you'll, you'll see a lot

SPEAKER_305: of things created and they'll look great. And a lot of that is, you know, uh, a chat bot behind it.

SPEAKER_70: Yeah. Well, I'm super excited again, everybody. This was, this was Michael Tan at Mike, the man, right? Mike, no. At the man, Mike Tan. At the man, Mike Tan, link in his bio, you can find, uh,

SPEAKER_385: the death of uncertainty, a story of AI and free will super excited. I was able to read it before

SPEAKER_305: the, the boom happened. Yeah. Hopefully I, I, uh, hopefully you enjoyed it. Hopefully it's, uh, it didn't get stuff too wrong or, you know, things that are happening in the real world

SPEAKER_381: or aligning to, uh, to what the book is saying, but I'll leave that up to, uh, for you to judge.

SPEAKER_09: Well, we're gonna have to have you on again in five years to, uh, to rehash this all again, again, Michael, thank you so much for coming on. Thank you, Rachel. It was a pleasure.

SPEAKER_01: Okay, everybody. That is it. Thank you for joining me, Rachel. It's always fun to talk news with producer. Rachel will be back Monday with a lot more this week in startups email

SPEAKER_16: your thoughts and your story ideas to producers at this week in startups.com and have a great weekend.