SPEAKER_00: Hey, everybody, welcome to Tuesday, the worst day of the week. Just kidding. We actually have a lot of really fun news today. We have a big conversation around Dahl E2 and OpenAI partnering with Shutterstock and the copyright implications for the future when it comes to generative art, IP, training data. It's actually a really good, solid back and forth between me and JCal. And then we talk about a new wrinkle from Apple, which is integrating smart home locks as Apple exclusives to go along with some new home key features in iOS 16. And that's all great, but it's also a smart home lock that Apple did not build internally. We have questions. Then I've got a little solo news on Apple again and that NFT reversal, some antitrust rulings against Google, some news about semiconductor talent shortages. It's going to be a great show. Stick with us.

SPEAKER_04: This Week in Startups is brought to you by LinkedIn Marketing.

SPEAKER_06: To redeem a free $100 LinkedIn ad credit and launch your first campaign, go to linkedin.com slash thisweekinstartups. Embroker's startup insurance program 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 embroker.com slash twist. While you're there, get an extra 10% off using offer code twist. CalShe just launched an election forecasting challenge ahead of the national midterm elections on November 8th. Go to calshe.com slash EFC, fill out predictions for each race, and if you get a perfect score, you'll win $100,000. If no one gets them all right, then the best forecaster gets $25,000. Sign up and play at calshe.com slash EFC.

SPEAKER_09: All right, everybody, welcome to Tuesday.

SPEAKER_10: How are you feeling, Molly Wood?

SPEAKER_01: I feel amazing. It's fall. It's finally fall. Like we had our weird fogtober or augtober. Have you heard this? I had never heard that some people in the Bay Area call it augtober. No, A-U-G. Like August, like because it's our second summer.

SPEAKER_16: Yes, it gets very hot here in September, October.

SPEAKER_00: Exactly. So we had our couple days of like irritatingly hot, and I don't know why. Ruin all the pumpkin patch photos. My kids too old for that, but you know what I'm talking about. And now it's just crisp and beautiful, and it smells like eucalyptus outside, and there's dew on everything. And the Mexican sage is going crazy in my yard, and I just am like, I'm just in such a good mood.

SPEAKER_20: We, my wife came home the other day, and she had a pumpkin so large that two people are required to carry it.

SPEAKER_23: And so, yeah.

SPEAKER_25: I really like her style. Yeah. She's pumpkin spice latte. I'm like, you know, these pumpkin things, they have no function. It's not like we're eating them or making pumpkin pie.

SPEAKER_29: They're strictly aesthetics, and my kids love them. My daughters love these pumpkins.

SPEAKER_00: I bought the like expensive, like heirloomy looking ones, like just for decoration. Because like my kid's kind of too old to carve. Like I'm going to have a little party. On Saturday, but we probably won't carve pumpkins. So I bought those purely decorative ones that were literally like $18 each. Just to look like a freaking heirloom tomato pumpkin.

SPEAKER_33: This is sending me into a Jerry Seinfeld term. Who are these people buying pumpkins? Like just for a look.

SPEAKER_39: And then I bought all the little tiny ones and I put them all around. We never make pumpkin pie.

SPEAKER_00: We don't make jack-o-lanterns. Five-foot skeleton. I am just, I'm like, I cannot believe how happy I am that it's holiday season. Halloween is the most inefficient. Holiday ever.

SPEAKER_19: Buy my 10-pound bag of Costco candy. Like I'm in it.

SPEAKER_46: There's a little bit of candy escalation going for trick-or-treat. You know, when I lived in the San Francisco area, everybody would descend on Broadway.

SPEAKER_25: If you look up like Broadway mansion, San Francisco, this is like, they call it billionaires row. Like a very notable billionaires have these $30, $40 million mansions that look out on the bay, Alcatraz and the Golden Gate Bridge. Quite nice. But you go there, and I lived south of this for a little bit in an area called Cow Hollow, below the Pack Heights, which they joke to as Pack Whites here in the bay area.

SPEAKER_51: Ha, ha, ha, ha, ha. Ha, ha.

SPEAKER_50: Ha, ha. Hmm? Specific wise.

SPEAKER_25: There's a lot of jokes around it. Oh, wow. But anyway, I take my kids to the first time. I kid you not, the first house we go to, they're not only giving out full-size candy bars.

SPEAKER_56: You know, like we all, as civilians, we buy mini candy bars. You get the little mini pack, right? You get a mini crackle, mini. Crack pack even, even like the mini mini. Sure. Snack packers, yeah. So you hand a couple of these over. I go to the first one.

SPEAKER_25: Not only are they full-size, they're king-size. Not only are they king-size. Come on. But this guy has lint chocolate bars. Premium king-size, $6 chocolate bars. This dude comes out with a huge bucket of them, puts it on the ground. I kid you not, the kids descend on it like they're zombies from The Walking Dead preying

SPEAKER_46: on like a, you know, a horse. My kids take two or three lint bars. I said, whoa, whoa, whoa, whoa, whoa. Uh-uh. You can take two. One for you, one for dad. Let's, let's go. It was crazy. The escalation is crazy.

SPEAKER_62: Trick-or-treat escalation is real.

SPEAKER_00: I was at Costco this weekend. I tweeted this even. I was like, Costco's just straight up moving the Overton window on Halloween candy. Like they have these big boxes of full-size bars. I was like, no, no, no, no, no. Do not normalize this. It's one thing if you do it in your specific whites neighborhood, but like, don't, I never want to say that again, by the way, that made me feel horrible. It does. It makes you feel terrible. I would like to just, I would like to retract that. But like, that cannot be, America has a sugar problem, a diabetes problem.

SPEAKER_69: Like you cannot be giving out the full-size bars. That's insane.

SPEAKER_25: All right, let's get started. There's some news to get through here today. Uh, and, uh, this first one, uh, is something we've discussed is very important about fair use and image use that I, uh, would like to get into with you. So let's queue up this first story.

SPEAKER_00: So we mentioned, yes, last week, uh, open AI reportedly raising more money from Microsoft. That was on episode 1592. And they are already back in the news because shutter stock is partnering with open AI to sell AI generated stock imagery. So already commercializing the images that are being created by AI, which is fascinating. Both parties evidently have been working together since 2021 shutter stock sold images and metadata to open AI to help train the dolly system. Hopefully that did not include like all the pictures that we uploaded from our class onto shutter stock, but I'm just not going to think about that. Uh, text to image generator dolly to integrated into the platform also, uh, or will be integrated into the platform in just a couple of months shutter stock stock has been down 33% over

SPEAKER_72: the past six months, but after they announced this partnership, it spiked more than 13% today

SPEAKER_00: after it responds to, I guess the idea that there is still a life in shutter stock or people remember that it still existed and AI can fill it with, uh, images, but shutter stock CEO released a statement saying quote, the mediums to express creativity are constantly evolving and expanding. We recognize that it is our great responsibility to embrace this evolution and to ensure that the generative technology that drives innovation is grounded in ethical practices. Now artists, I made that kind of snotty comment about where they using our class photos. Cause that's what we all had to use for uploading photos. But artists have complained that shutter stock did use their work to train this AI image generator.

SPEAKER_72: And then now we're going to profit off of selling those images as stock art.

SPEAKER_77: Uh, shutter stock's going to lose this lawsuit. I predict, uh, I think there's going to be a lawsuit.

SPEAKER_25: Uh, there is a hundred percent going to be many lawsuits on this. And here's why the use of, uh, somebody's original work to make derivative works. There's a long history of this. And so let's go through some examples here. If I were to give you the entire corpus of Marvel comics, every single Marvel comic ever made. And Marvel had built a delicate licensing arrangement with the people who created them. Now, in some cases they were what's called work for hire. You've got a salary. You came to work every day, whatever you built, like most companies. Uh, in fact, the people who work at open AI, people work at Google, any major company. You work for hire your work at, uh, marketplace or, uh, New York times. You were paid. They own it. You can't take it with you now. Okay, great. But in these instances, artists typically have rights and those rights are very detailed and they're very granular. And so in the case of people who created different characters, they might have licensing deals for those characters. Or in the case of I made, you know, ET, the extraterrestrial, I wrote the screenplay, you know, there's all these mechanical licenses and derivative product licenses, famously Star Wars, you know, he Lucas kept the rights to make the figurines and the toys, right? You know, and the sequels. So these are very granular, complex, specific. Yeah. And so now in the case of Shutterstock, I'm guessing that they buy the rights to resell these images. And the same with Getty. These were not work for hire. Now they might have a selection of them where they own all the rights explicitly. There might be some on the web that are what they call creative commons. If you don't know, creative commons is a licensing. I know, you know, Molly, but for the audience, you can look it up. Yep. Some people will let people create derivative works as long as they're not commercial, but this is clearly commercial. So let's say in the Marvel instance, I said, okay, now make me a character who's similar to Wolverine, but is a female hand is in an in the Frank Miller style of Batman, uh, et cetera, et cetera, et cetera. So now we've got this dark night ish female version of Wolverine. Your mind can imagine that and obviously AI is going to do it instantly. There is no world or situation in which no a jury, or the legal system would say that's fair.

SPEAKER_90: Mm hmm.

SPEAKER_25: Okay, great. Why would we not see that? Because it's pretty clear that this is a derivative product that is going to impede the person's ability to do commerce in the world. These are very complex tests here, copyright law. Mm hmm. So now you go to a bunch of less powerful photographers. Now somebody took a bunch of pictures of fruit, and you go there and say, Hey, make me a fruit basket. And I want it to be very dramatic with lighting. That's very colorful. And the AI decides, well, I'm going to take this Molly Woods beautiful photography. And I'll just never need Molly Wood again. Because we know what a pineapple looks like. She never happened to make pineapples.

SPEAKER_95: But now we got pineapples in the style of Molly Wood, this famous photographer. Yeah. You might have a harder time with that case. And Molly Wood, the pine out the fruit photographer might not have the wherewithal to go through these lawsuits. But this seems profoundly unfair. Of course, they are creating a contributor fund here. Try to reimburse people. But I don't see any details here.

SPEAKER_64: Yeah, yeah. I don't see any details here. And I don't see permission. Exactly.

SPEAKER_00: They're doing two things, primarily to try to combat these criticisms. One, so they do have, it sounds like I need a little more detail on the Dali integration that they already have. They have some kind of deal with Dali. And so they're banning the sale of any AI generated art that is not made using this existing integration. So they're sort of saying, look, if you use some other, like, we have an integration,

SPEAKER_01: we've trained it, presumably using specific art and works that we have acquired the rights to. And if you use something else, like I actually tried out this one that goes through discord

SPEAKER_00: last night, if you use that, for example, you can't sell that on Shutterstock, right? It has to be.

SPEAKER_102: You couldn't use stable diffusion as an example, right?

SPEAKER_00: You have to use just this Dali integration that only was trained on stuff that, in theory, Shutterstock has the rights to. So that's step one. Step two is this contributor fund that will, in theory, reimburse creators when the company sells work to specifically to train text to image AI models. So when they buy art in the future to train the models, then creators will get reimbursed. But you're, you're absolutely right. That does nothing to head off the lawsuits from the artists who will already say, who will say, you know, my art was already used here. I do think that's going to be hard for them to prove hard for who to prove the

SPEAKER_01: artists or if you were an artist who was going to sue Shutterstock now and say you included my art in your training data. Yeah. That might be hard to prove.

SPEAKER_25: I think actually I might take the opposite side of that. If you said to the person, yes or no, did you point? Did you? Did you upload my image to Dolly? They say yes. Okay. We're done here. Right?

SPEAKER_83: You used it. Take it out and retrain your algorithm. So that would be right. And get my permission.

SPEAKER_58: So now it is there. I mean, you get it in discovery. Yeah. And yeah, and just exactly what I was thinking. So you cut me off at the past there. Now you get to discovery. And they're like, Okay,

SPEAKER_84: what images did you use? We don't know. Really? You don't know? Okay, get us a developer. And

SPEAKER_25: let's see their slack messages. And oh, wait, it's, it's on a computer. Great. Show us the files that you uploaded. Okay, you can't prove it. Great. Start over, delete everything in Dolly, take your algorithm, and bring it back to zero. Now these things have probably, I'm guessing this software has already been trained on open images on the web, because if a developer is doing this for non commercial reasons, they probably just typed in Google searches, sucked stuff down, and they didn't think about it. So now where did the algorithm start learning? Where did the AI start learning is going to become the question? Mm hmm. And that's where I think this is going to be like a legal fight. For decades, this is going to be a legal fight that I don't think they're going to win. And other platforms I understand, like Getty, they have banned the use of AI art due to copyright licensing complications, because some of those like Getty, they would license very famous photographers, those very famous photographers would have a very granular license that did not anticipate AI making derivative works. Yeah. So this is where this whole thing is going to be litigated for a decade. And I think what they're going to have to do is start from zero again. And they're going to have to say to people, we want to buy from you collections of images for this purpose. Will you sell them to me, that would be the fair thing to do. So what shutterstock should do here, if they are an ethical organization, an open AI, if they're an ethical organization, what they should do to content creators is say, we are going to start a derivative works program, we're going to pay you $100 per image upfront, against royalties, that will be determined in this way, would you like to opt in yes or no, what they're doing here is they're not giving them the

SPEAKER_64: ability to opt in. And that's what pisses me off.

SPEAKER_00: Like they can opt in going forward. We don't know we should say here, we do not know if shutterstock did pay people in the past, right? We know that creators were upset about being included. We don't

SPEAKER_73: know if shutterstock has ever paid anybody. And they may have they promise they paid them to

SPEAKER_64: to sell images to license it originally to sell them, we don't know if they paid them any additional royalty for training data, they might have obviously they haven't because nobody anticipated this,

SPEAKER_25: you know, having been in the space and looked at these companies before, I never heard a stock image

SPEAKER_23: and I looked at investing in these 10 years ago, I never heard anybody bring up AI generative work. I think this thing came out of left field faster than anybody thought. Yeah. Hey, Tom Eschbacher is here

SPEAKER_131: with us again. He's a senior sales manager at LinkedIn marketing solutions. And we're talking

SPEAKER_62: about their amazing report today and startup marketing as well as how to use LinkedIn to grow your startup. What are some tactical things not big picture strategy, I'm talking tactics that

David Friedberg: founders can do today to figure out product market. One of the big tactics we see here is amplifying organic posting with paid advertising, you consider a startup that raises a seed round, they post the news on their LinkedIn page and see a bunch of likes clicks and follows come in. They follow that then with some updates about product and they see continued traction with, for instance, HR benefit managers at tech companies that have fewer than 500 employees. That's a signal and it becomes important to then get a larger sample. And to increase confidence, we've made it super easy to identify which audiences are engaging with your organic content, your LinkedIn company page, your website, and then extend reach into those segments with our best in class B2B ad targeting. So for early stage startups who amplify organic with paid, we see a 13x lift in unique reach. And those are meaningful insights to help inform product and go to market strategies.

SPEAKER_62: Such a great strategy. Head to linkedin.com slash this weekend startups and get the report now so you have an edge on your competitors. And as a little pot sweetener $100 off your first marketing campaign, thanks to calm and the team at LinkedIn, go get that report and get the hundy, which is why the right

SPEAKER_25: thing to do. And this is what I would do if I was representing and there are unions and groups that represent these and there are lawyers who represent artists, what they should say, what I would do what I advise them is to say, Yeah, okay, let's sit down from first principles and get permission first. This is the same thing I dealt with with blogs. RSS came out. And then people who had RSS players, RSS readers would pull down and gadget and auto blog, it's okay, great. And then some of them said, we have a free one, we're going to put ads into the in gadget feed, we're going to put ads on our RSS reader. And I said, Okay, you can't put it on our feeds. And they said, Oh, we didn't put in your feed, we put it in our RSS player. I said, Okay, when in gadget comes up, when auto blog comes up, you are not allowed to show an ad, you must get our permission. And a lot of them told me, Hey, Jake out. Sorry, pound salt. And I said, Okay, you know, I did I called up my pal, my friend and me, Nick Denton, and Nick Denton, and I said, we are going to create a competitive RSS to yours reader, number one, and demolish you and we are banning your RSS feed from all the Gawker and weblogs in feeds. And you know what happened, Molly,

SPEAKER_93: everybody changed your attitude real quick. And they said, Okay, when you pull these feeds up, you can't put ads in them. And we started putting ads in our feeds. And we said, You can show ours. And it's the same relationship we have with YouTube, right? If YouTube said to us, Hey, you can't put the baked in ads. We'd say, Okay, great. Yes. The only difference I think here is,

SPEAKER_00: we do own this content, and we have not licensed it. Like in that case, you own that content, you had not licensed it to anyone else. If you are an artist who, you know, Shutterstock clearly seems to believe and is saying, we, we trained these, we sold specific pieces of art, right data to train these models. And it's data we licensed, like if you license something to Shutterstock or to Getty, you sign a contract and and you give away probably a lot of these rights, certainly if you sell it to

SPEAKER_01: them outright. So Shutterstock is like, Look, we took content we already owned, we trained licensed,

Chamath Palihapitiya: licensed in most cases, yeah. And so the license, what does the license say is the devil's in the

SPEAKER_00: detail here. But I'm just saying they had licensed it. And they their position clearly is we have licensed this, we took data we had already licensed, we trained these models. That is why we're not going to allow any AI generated art that didn't come from these models that we trained on, on content we already licensed. And we're gonna, you know, pay people going forward. I personally think that we this is such an inflection point, that like, you're probably right about lawsuits. And Shutterstock is probably right about the future. And Getty is probably wrong. Like the idea of just saying, we're not participating in the AI and the generative AI art future. I think long term is gonna be a

SPEAKER_25: loser. I think the question is, so the point I'm bringing up is the begging for forgiveness, as opposed to asking for permission that we see in the copyright world. So we also had people who took this podcast, and they started clipping it. And I went to their website, it was like some pod clip something. And I said, Wait a second, you're taking Tim Ferriss's pod, and this pod and all in and you just took the chapter headings, and you made clips, you downloaded the you auto downloaded the mp3s or videos, you automatically cut them, you're hosting them on your site. And you never linked back to us, you never asked the permission. So I said, take it down or and the person didn't, you know what I did, I sent a DMCA to their hosting company. And you know, that company did, they said, you have 10 days or returning off your access. And you know, they did, they suddenly took our stuff down. And I said, Listen, if you want to use this, and I've done this with people who repeatedly cut clips, from all in and use it to build traction with their thing, I said, Listen, you know, with their social media, if you want to do that, just ask for permission, I'll give you a one year, non commercial ability to do that. So when people started clipping the show for like, fan stuff, I said, it's fine. If you do that, I'll let you do it for a year, we'll see where it goes. Just ask for permission first. I mean, I hear you like regularly, if you do one,

SPEAKER_23: that's different. But if you're doing a three, three per episode, that's, that's totally different.

SPEAKER_00: I hear you. I just think we're conflating two things like nobody knows copyright licensing better than Shutterstock and Getty. Right? They're saying so, so one, you're pointing out there is a culture of use and taking Yes, that 100% exists. And AI, like generative art is going to amplify that culture and already is. But to separately, like, I don't want to conflate Shutterstock with the people who are clipping the podcasts, you know, like, because again, they licensed this material. They trained their AI on material they had licensed. Like, I don't think this is a case of I just don't want it to. I am not of the opinion that Shutterstock is doing the same thing as like

SPEAKER_25: clipping a podcast without permission. Like I know they are not the same thing. But it is the same disregard. I do not for asking for permission in advance. I don't think anybody who licensed their photos to these services ever thought that they would have an AI that would build images that would preclude those artists from ever making money again. Because that's what we're on the precipice of if the if the people who are selling these images do not fight this now, they will never make money again is my prediction. Because once you have that, that's true. Once you have that AI trained, it's going to run the table 99 times out of 100, you are not going to need an illustration. You know, I saw this really ugly illustration they did in the New Republic of David Sachs, they did a profile of him. And I looked at that image. And I was like, that distorted image is gross. And like the illustrator made them look like really like kind of ugly characters almost like a mad magazine kind of send up. And I just thought to myself that I thought it was a dolly. And I was like, No, a human got paid probably 1000 bucks to make that or 2000 bucks to make those images, those magazines will pay something in that range. And then if they were offshoring it, maybe they'd pay to a 300 bucks for an illustration, that illustration will be made by dolly, all the illustrations in the New York Times, the art director at the New York Times, like give me an image of David Sachs, make it a little grotesque, make his teeth a little bit bigger, make him sweating, make his eyes buggy, and then show me 10 different versions of it. And now it make it sort of like mad magazine and the AI is going to know it, Molly, there are going to be

SPEAKER_83: no illustrator jobs. These are going away. They're gonna be no photography jobs is going away. I

SPEAKER_72: 99 out of 100 Nick in the interim went and found that photo in his horror for that illustration and it's horrified. It's unbelievably. But here's the thing, though. It's too late. And I'm not trying to oh, good God. Yeah, I did see these. Yeah. My favorite is actually is the teal one. Like,

SPEAKER_166: what is even happening there? His eyes are rolling back like he's turning into a zombie. Okay, anyway, I think we beat this to death here. Anyway, I think the cat is out of the

SPEAKER_164: way. Wait a second, I've been trying to make a point, but your slack was effing it up the whole time. Okay, make the point. Sorry. It's too late. The cat's out of the bag here. The cat's out of the

SPEAKER_00: bag here and Shutterstock is getting in front of it. And they're going to take a bunch of arrows on the way. But this everything about the art industry has changed in probably the last month. That's how

SPEAKER_01: fast this has happened. But like, yeah, it's too late. These AIs are already trained. They're not going to get untrained. Like, well, I think they wipe all the hard drives. No way. Yeah, I think

SPEAKER_62: that's exactly what will happen. I think they'll get an injunction. I think they get an injunction.

SPEAKER_25: Yeah. Because what's going to happen is this is going to go to the Supreme Court. I kid you not. This is going to go to the Supreme Court. I mean, this will be one of the just like the,

SPEAKER_01: you know, the Disney character stuff did. And then the legal actors like Shutterstock will get shut down and they'll get an injunction and AI everywhere will continue to proliferate. I guarantee. No, I'm not. You know what? You're right. You're right. But here's what will happen in America.

SPEAKER_25: They will be required to pay up front and get permission to train the AI with this data, which is what they should have done from the beginning. That's my predicted outcome. Anybody who is producing this stuff is going to be forced to show, show their work, show your work, just like in math class, show us how you solve it to make sure you're not cheating. Show us your work.

SPEAKER_00: There's gonna be personal information in there. One of the notice pointed out, you're not going to be able to show your work, you're gonna be like, I can't do that, because it will be revealing personal information. It'll I mean, it's gonna be messy, don't get me wrong. But I 100% will die on this hill. This cat is out of the bag. The cat is

SPEAKER_25: all the way down the cat cat is out of the bag. But to make a commercial product means you are subject to the law. Now, that doesn't mean that somebody will not make an open source one,

SPEAKER_181: just like you can still use contracts, you can say, you know, yeah, but people have signed all

SPEAKER_00: kinds of contracts over the years and then been like, I didn't think that when I granted you

SPEAKER_01: rights in perpetuity in the entire universe, which is what these contracts say, they literally say, like across the universe in perpetuity. And then people will be like, I didn't know you were going to be my image onto the moon. Well, that's what the contract. So that the contract says for you.

SPEAKER_25: Yeah, they usually do say that in any medium in the future. But it doesn't anticipate the use of these images to then preclude the ability of the person to make future work. Anyway, we'll monitor this. I think it's nice that shutterstock I'll give them credit for trying something. I think they should get permission. And I bet you that is what Getty Images is doing at this moment. They're going to go to Ansel Adams because they own his archive of the estate of Ansel Adams. And they'll say, what would it take for us to create an Ansel Adams? Is that the person right who makes this beautiful? They're going to go to Ansel Adams and say, we will any and then they'll say, here's we'll make it even fair. If anybody says in the style of Ansel Adams, we want this charge to occur. So if you if you use an Ansel Adams, you have to then use this version of Dolly that has the Ansel

SPEAKER_23: Adams archive in it. And it might be that Marvel or Star Wars will say, well, here's a product, we're going to sell for $100 a year, a product where you can create your own characters. And for non-commercial, just like Star Wars, let you make fan films.

SPEAKER_189: And is saying, and again, we don't get permission here.

SPEAKER_00: We don't know that they're retroactively doing it. We do not know that. We do not know that when they sold the data to train the AI, their integration of Dolly, we don't know if they didn't get permission.

SPEAKER_192: Somebody leaked me a contract. Show me the contract. Somebody leaked the producer's ad.

SPEAKER_177: Story where it says they.

SPEAKER_83: It doesn't say that they did that now. Anyway, this is my advice to the artists, hold the line, fight like dogs. Your life depends on it.

SPEAKER_139: I mean, that's true. Fight like dogs. Slash it's already over.

SPEAKER_197: And you will get paid and you will win. It just take five years, but you're going to get a huge payday because settling this is going to be easier than fighting it for a decade.

SPEAKER_95: These people want these tools in the world. A hundred percent.

SPEAKER_25: I'm going to quickly explain one of the crucial types of insurance that every startup needs. It's cyber insurance. Obviously this covers hacks, which happen more than you think. The world is crazy right now. We all know that cyber hacks are happening constantly. So if you don't have business insurance, you failed one of the first steps of being a founder and even startups need to get this insurance in place early because crazy things happen. It's not that expensive and it doesn't take a lot of time. Thanks to our friends at Embroker, their technology saves you so much time, so much money because prices are 20% lower and you get better coverage than the incumbents. You go from sign up to quote and purchase in just 10 minutes. Think about that. When you work with Embroker, instead of these large, slow incumbents, you're not dealing with big companies that want to talk to you on the phone for hours and hours, and then they forget about you. They never call you back. No, you got a professional, nimble organization. Sign up takes only days, not weeks, and the process is completely transparent. There's no peak pricing. No, they're going to treat you right. To instantly buy custom-built insurance for startups, go to Embroker.com slash twist. While you're there, get an extra 10% off by using the offer code TWIST. Thanks, Embroker. You do a great job over there.

SPEAKER_146: Let's keep going. Let's talk about, we got a little gadget news. It's been a long time.

SPEAKER_00: I love the good gadget. What's the gadget? And I am sort of mad about this because I just bought a similar gadget. In home safety news, so Apple included some new features in iOS 16 around smart home capability, including door locking and unlocking. And now they have a new lock. They call it the home keys feature. Now they have a new lock exclusive to Apple by this startup, made by this startup called level. And they are selling an app, a smart lock for your home. You know, like the thing that you use, you use the phone to unlock the door. In some cases they have a fingerprint reader. It looks like this one's even simpler. It's just literally like you just use the phone to like turn the deadbolt. And that has a traditional keyhole that can be used with a physical key. And they just sort of quietly started selling this smart lock for $330 that works only with their home key feature in iOS 16. What I think is really interesting about this is like one, I just bought a smart lock and I'm like, dang it.

SPEAKER_205: I have a lock or did you buy the August lock or just I bought neither of those.

SPEAKER_00: I bought some, I bought the wire cutter pick. I can't even a lock, I think, or something it's called. Yeah. But now I'm like, well, I haven't installed the X. I'm a little intimidated. So maybe I'm just going to return it and get the Apple one because delightful. And you can use it with although it is three times the price. But what I think is really interesting about this is that isn't this this is kind of this is sort of Apple's first foray into smart home, which so far has been like Google has made attempts to dominate the smart home, but pooped the bed as we know all too well nest. And then I mean, they have the Google Home devices like you can do Google Home integration with smart light bulbs and and outlets and things like that. But they they're like Google, so they're not that committed to it. Alexa is integrated in everything.

Jason Calacanis: And I apologize if I just say all your speakers off across the entire house. But here's Apple coming along like boop boop. We're gonna start with the lock and expand out from here. And I'm totally like I'm in because they're not gonna try to sell me anything

SPEAKER_56: or sell my data like that pitch is working.

SPEAKER_25: The only problem I have with this or that's confusing to me is they didn't build their own lock from the bottom up. Like this isn't an Apple hardware product. This is an Apple exclusive. Yes, I haven't seen them do this before. Have they ever done like an Apple sanctioned exclusive thing? Like I guess they approve the cases and stands maybe like this logitech is approved.

SPEAKER_211: And some of the chargers maybe are kind of yeah, maybe they're Apple certified. Yeah.

SPEAKER_64: So maybe this is like more like an Apple certified. You're right, though. Like you're pointing out a very interesting thing, which is that this is also a new hardware model for them.

SPEAKER_25: Yeah, maybe they're seeing what the interest is. It could be but then they would be competing against somebody they sanctioned,

SPEAKER_23: but I guess they've done that before they they have their battery pack and they'll have other battery packs in their stores, I guess. Yeah, still buy some other third party ones. It's kind of cool.

SPEAKER_25: I think the key to these locks is and I'm looking at the key. Yes, the key to these You know, the big unlock here, Molly, is when you can send a key virtually to somebody that has a time based thing, and then it gives you alerts. So that's what a lot of the other ones are doing now is like, okay, you can give a key, a virtual key to your housekeeper to the rug cleaner to the dog walker. And when they come in and out, you know, it's them and then hey, your dog walker, you fire your dog walk, you just go in your app. And hey, yeah, the keys burned.

SPEAKER_95: And now you don't have to have a key remade. So yeah, this is a key killer feature here. Obviously, Airbnbs love this kind of stuff.

SPEAKER_25: And so I'm all in on this and not this specific one, but in this category, and I'm going to be moving to some of these. I love these security systems. I think it's one of the things we're having now is I I'm noticing is the issue is not we no longer have this issue around catching criminals. Do you notice how they catch the criminals almost every time where like we have them on camera, and somebody's got an air tag on their bike, and they're like, Okay, I know where my iPad is. I know that this person like once in a while, you'll have the TSA agent take an occasional iPad home and they bust them like on the local news, or somebody steals a bike and the person's got an air tag in the seat somewhere hidden. And they're like, they tell the cops like my bikes in this place. And then it becomes do the cops have the wherewithal to lodge and the legal agencies to actually go knock on the door? Or does the person go knock on the door? Yeah. And you be careful folks like, you know, we stole your stuff. Yeah, be careful going to retrieve it.

SPEAKER_24: Because that could be a very dangerous situation for you. Yeah,

SPEAKER_72: shout out producer Justin, as Toby, one of our notice pointed out.

SPEAKER_00: All right, his bike stolen, remember, or he had his computer stolen. And he went and collected his work stuff. And he went and collected it. The cops were there. He like hid, he like hid behind a pillar, called the police was like, I know where it is. And they came in.

SPEAKER_25: Yeah, so like, do you arrest the person or not? My favorite is there's a person person on YouTube who makes a glitter bomb. Have you ever seen this person?

SPEAKER_225: Yeah, Mark.

SPEAKER_64: Oh, you know them? They're a Bay Area person, right? Oh, yeah. He's like a NASA scientist. He also does all the like squirrel stuff.

SPEAKER_169: His videos are unbelievable, like entertain the children for hours.

SPEAKER_116: So that he makes a glitter bomb every year and they get he has the

SPEAKER_25: voice from home alone, like, get your hands off of your dirty blah, blah, blah. And they also put stink bombs in it. And so he's making these like increasingly crazy ones where people steal a package. They bring it home when they unbox it.

SPEAKER_231: It goes and sends a stink bomb and glitter everywhere. And videotapes the person in live streams that it's bonkers.

SPEAKER_64: I mean, I'm so excited for you to be about to discover the entire Mark Rober catalog. Hey, you know, we have a we live in the future.

SPEAKER_24: I love these WL ITFs. And I've instructed my team to double down on these and find us those seed stage companies

SPEAKER_95: for us to talk about. I would really love to get to Japan in an hour. Because that would be two hours less than getting to Tahoe. And I can ski Hokkaido faster than get to Tahoe.

SPEAKER_235: That'd be pretty dope. Yeah. Tell us about this one.

SPEAKER_00: Well, yep. I want this one. This is the Venus Aerospace Mach 9 hypersonic space plane. I mean, the future is going to be awesome. Houston based Venus Aerospace is targeting one hour global travel with the space plane called the Stargazer trying to get passengers from Tokyo to Los Angeles in one hour. Right now, that flight takes about 12 hours direct. It's a hypersonic aircraft with the potential to reach Mach 9, i.e. nine times the speed of sound or almost 7000 miles per hour. I just want to leave. I want to put my headphones down and walk away and watch that scene in Maverick right now with where he's trying to get to Mach 10 in the plane. It'll be able to hold 12 passengers and fly at altitudes of 170,000 feet. It's first ground test like this is the future. Don't forget its first ground test is not expected to happen until 2025 at the earliest, but they're hoping in July. I mean, just in July, Venus raised $29 million at an $84 million valuation from a bunch of investors, like a party round and a combination of multiple series A tranches. And they hope that tickets for the Stargazer would be roughly the same price as a first class ticket on a commercial aircraft.

SPEAKER_25: I love this. I hope it happens in our lifetime. This would change the world. I always regretted Molly not going on the Concord. You know, it's a little bit just missed that window by a little bit. But man, the ability to get somewhere faster is just wonderful. Of course, the amount of energy it's going to take. Yeah, would I think be very, very expensive. They don't talk about what flying in a 7000 mile per hour, nine times the speed of sound plane would cost. I got to think this thing's going to burn a multiple of the fuel. And because it only has 12 passengers, I'm going to say when in today's dollars, this is going to be a $20,000 each way ride, which is insane. But if you're a celebrity going to do a show in Japan, and you're Lady Gaga, and you can get there in an hour, you can do three more shows in Japan every year. And you know, whatever, if you're if you're if you're making hundreds of thousands of dollars for an appearance, well, then it makes sense. Or if you're doing a deal that generates hundreds of thousands of dollars in some way,

SPEAKER_201: like a SaaS deal or something like I guess it could make sense.

SPEAKER_00: So maybe certain points out to that they fly so high, that air resistance is lower. And so maybe the fuel needs are not as extreme, which I hope because from a fuel perspective, it's kind of hard to write once you get justified.

SPEAKER_243: Yeah, when you get to take so much to get out of the gravity. Well, that's it. Yeah.

SPEAKER_25: So once you get out of the Yeah, once you get to 170,000 feet, there's no air up there. And I think that's why it can go so fast is that you don't get the resistance. But I mean, just for the thrill of it, it feels like a bucket list kind of item. So I love it. Totally. And I can't wait for this to get here. I mean, as an investment, this is going to be burning money.

SPEAKER_95: And I you know, the idea that startups could do this is kind of crazy. So also have very big military kind of applications.

SPEAKER_245: They're going to get a lot of non dilutive funding. Let's put it that way.

SPEAKER_95: Yeah, I think the government's going to buy a bunch of these. All right. Great job today. And look forward to seeing everybody tomorrow. We've got our crypto roundtable.

SPEAKER_201: And okay, boomer on Friday. So a lot of great stuff to come this week. Come on.

SPEAKER_00: And the news is not done. There's I got a couple solo segments coming up. Yeah, a little bit of a couple little updates that I knocked out because we divide and conquer on the show.

SPEAKER_25: So so much news for you. Enjoy. All right. You've heard me talk about Kalshi before. K-A-L-S-H-I. I want you to remember this name. It's a regulated exchange that offers financial prediction markets on everyday events. So you can trade on things like, hey, what's going to be next month's CPI or whether certain legislation will pass. Maybe what global temperatures will look like at the end of the year and more.

SPEAKER_00: But now Kalshi has an awesome new contest. They just launched an election forecasting challenge ahead of the national midterm elections, which are taking place on November 8th. Basically, they're looking to find the best election forecaster in the country. Here's how you join. Go to Kalshi.com slash E-F-C, election forecasting challenge, E-F-C. Fill out predictions for each congressional race. It's like March Madness, but for voting. If you get a perfect election call, you win $100,000. Are you following the polls? You can change your submission right up to the day before the election. If no one gets them all right, then the best forecaster gets $25,000. This is like crowdsourced predictive market right here. Super interesting.

SPEAKER_64: The contest is free. All you need is an email to play. The entire Twist team is on Kalshi now, and we are going to be talking about our bets

SPEAKER_25: during these ad reads. When you hear me talk about Kalshi, you want to listen and hear our bets. I want you to sign up and play at Kalshi.com slash E-F-C, Kalshi.com, K-A-L-S-H-I.com slash E-F-C.

SPEAKER_00: All right, I'm going to do a quick rundown on what is happening with the semiconductor economy because, of course, the U.S. has made this big push with the CHIPS Act to reshore semiconductor manufacturing and foundry building and really creating resilience in this specific part of the economy. And there is only one minor issue with this effort, and that is that we don't have enough engineers, apparently, to build semiconductors. In fact, the estimates are that the U.S. chip reshoring push is going to need 50,000 new engineers over the next five years. Now, obviously, we have a widespread STEM sort of shortage right across the board in the United States. And now there's this kind of need for these specifically trained engineers in this area around manufacturing. So, as a result, this act, as we know, let's just recap, is designed to boost competitiveness, innovation, and national security in the United States and increase domestic semiconductor manufacturing. The CHIPS Act did allocate $200 million for worker training, which is great. Companies like Intel and the National Science Foundation have increased training. And now Purdue, and I suspect that Purdue will not be the last, Purdue is one of the country's biggest engineering schools, and it's specifically expanding its semiconductor education, aiming to graduate 1,000 semiconductor engineers annually as soon as possible. This is, however, a big jump. First of all, it would be 1,000 a year. We need 50,000. I'm not trying to be a downer. This is great that this is happening at Purdue. But for scale, according to the Washington Post, Purdue currently graduates about 150 semiconductor engineers per year. And imagine how this happened. It's a pretty straightforward march, actually. Even when you look at, you know, you can kind of compare it to venture capital and the tech industry writ large, it all went software and it all went apps. And so, even students who are in STEM fields, and there are a shortage of those students, are primarily pursuing coding. And so, now we're trying to shift to a different type of engineering. In some cases, it's going to be physical. They will literally be chip manufacturing. It may be a different type of engineering and coding that, you know, people aren't doing now. So, it's a big overhaul in terms of curriculum and training. Purdue invited semiconductor experts to join a new advisory board to make recommendations on that curriculum and training. The Washington Post also reported that Defense Department officials went to Purdue to talk about workforce training because, again, this is seen as increasingly a national security initiative. According to salary.com, the average semiconductor engineer salary is $114,155. I would love to see, actually, how that compares to software engineers. And if that's going to go up, I would imagine that it is, as this becomes a bigger and bigger push for the United States. It also, I think, you know, we talk a lot on this show about the ways in which immigration reform is totally necessary to fill gaps across the economy, whether it's doctors, nurses, vets, semiconductor engineers, and I suspect that we will start to incentivize immigrants with new visas. We have some existing tech visa programs that, yes, everybody agrees are flawed in various ways, but I would not be surprised if there are expanded visas for people who want to work in this industry. And then it's going to take a bit, you know, we've had decades now, right, at least one solid decade of learn to code as a mantra within the U.S. economy, that if you get into tech, this is a great way to make

SPEAKER_253: a great living. Learn to painstakingly build and fabricate tiny little semiconductors is going to be

SPEAKER_00: a different, it's going to be a different kind of pitch and it'll be interesting to see how we work to incentivize these incoming students and workers to do this kind of, in some cases, like I'm saying, physical engineering. And I have to be completely transparent here. I don't know that much about what goes into semiconductor engineering. I know that it's incredibly painstaking and you have to work in like a clean room, you know, like those Intel commercials. And also in full transparency, one of our, one of our producers said that it sounded kind of boring, but hopefully it is not boring to everybody. Nevertheless, there's going to need to be a PR push. Thank you, producer Brian. The new guy looked up the average base salary in the United States for software engineers. That is 121,806 with usually additional cash bonuses of about 14,000. So total software engineer compensation, $136,260. And in order to begin a career as a semiconductor engineer, you have to at least procure a bachelor's degree in material science and engineering, or applied physics with emphasis, emphasis in fields like computer science and electrical engineering. So pretty different path forward.

SPEAKER_01: I mean, I think it sounds kind of cool. And I hope that other people do too, but it is going to, it's going to get interesting really fast. And of course, this push to increase workers is coming

SPEAKER_00: as Washington is imposing new export controls on the industry. Now, semiconductors made with US tech for AI, for high performance computing and supercomputers can only be sold to China if you get this very hard to obtain export license. US citizens and entities are now barred from working with Chinese chip producers, except with specific approval. And this prompted the Chinese chip maker YMTC to ask American employees in quote, core tech positions to leave, because it now has to comply with these new US export controls. So disruption is hard, changing the path that our economy has been on for the last few decades of globalization is incredibly difficult. And I don't want to say we have the cart before the horse, because both the cart and the horse have to get somewhere quickly. We, we have to do this onshoring push. I don't think, you know, anybody disputes that it's necessary. And given what we've seen in China and the Biden administration sort of hardening stance toward China, there's no evidence that we're on the wrong trajectory, but it is going to take a dramatic increase in workers to get us there. And, and I applaud Purdue for jumping on this immediately and making it a competitive advantage for that school. I hope that other schools will do the same thing very soon. All right, let's keep up with some international news here. I just get to, I get to do all the international news by myself. India has fined Google $113 million and ordered the company to permit third-party payments in the Play Store. So this is a big deal because this is a huge market and potentially because it could set a precedent for other platforms, uh, even in the U S who knows, but here's the news as we have it so far. The competition commission of India has slapped Google with this $113 million fine for abusing its dominant position of the Play Store. The CCI ruled that developers are allowed to use a third-party payments processing service for in-app purchases or, or purchasing apps. The case was open toward the end of 2020. It said that Google's in-app purchase policy constituted an imposition of unfair condition. So, uh, this Indian antitrust watchdog has directed Google to introduce a series of changes to the Play Store policies within three months, including the following. Google shall not impose any anti-steering provisions on app developers and shall not restrict them from communicating with their users to promote their apps and offerings in any manner. Non-interference, basically Google shall not restrict end users in any manner from accessing and using within apps. The features and services offered by app developers, AKA Google will just get out of the way and let you use the app the way you want to use it, including how you pay for stuff, presumably. And then finally, Google shall not impose any condition, including price related conditions on app developers, which is unfair, unreasonable, discriminatory, or disproportionate to the services provided to the app developers. Google also has to be transparent in communicating with the app developers, uh, about the services provided and the corresponding fees that will be charged and published in an unambiguous manner, the payment policy and criteria for applicability of any fees. Google, of course, like many companies has poured billions of dollars into the South Asian market over the past few decades. It's aggressively trying to find major untapped regions. Many companies of course are seeking access to the Indian market. It is gigantic and a little less problematic than the Chinese market. Although I think what they're finding is that the Indian market is not a pushover. Let's just say like Indian regulators pushed back hard against, uh, Facebook when it attempted to introduce that Facebook's Facebook basics service, internet basics. It was basically, they would offer free internet access within India in exchange, you know, for users using an app that Facebook had effectively programmed with its own, um, pre chosen websites, including, I believe its own service. And you see, uh, their antitrust regulators pushing back pretty hard companies coming in and kind of trying to do business as usual. I'm very curious to see the extent to which India may become a parallel to the EU a little bit in terms of, of regulation. This is by the way, Google's second antitrust fine by this commission, the competition commission of India in under a week, Google was fined $161.9 million for other anti-competitive practices. This was an investigation that started three and a half years ago after a statement was released saying Google was requiring device manufacturers to pre-install its entire Google mobile suite and mandating prominent placement of those apps. India, not having it. All right, speaking of app stores and app store rules, Apple is cracking down on NFT functionality and social post boosts with new rules in its app store. So I think on a recent crypto round table, we talked about the fact that Apple is now enabling in some ways, the NFT ecosystem. However, Apple has issued new rules with its software update in iOS 16.1. By the way, if you're like me and you've been waiting to install iOS 16, you now can. I always wait for the 0.1 update because I have been burned before. Anyway, with the 16.1 update for iPhone and iPad, as well as macOS Ventura, the app store introduced some new rules, including limiting features that are unlocked through NFTs. It also mandates apps to ironically use Apple's payment method to purchase boosts for posts on social medias. First of all, we'll get to what that means in a second. But second, uh, did they roll this out in India? Because I'll be very curious to see how they respond to that. Um, what social boosts are social media boosts. These are normally boosts that you'd be able to purchase inside an app, right? Like, do you want to boost this post on Instagram, for example? Now those have to be processed in the app store, presumably meaning that Apple will get its 30% cut of all of those transactions. This would only apply to apps that offer in-app tools for promoting your posts. It would not apply to back-end sponsoring through an ads manager, because that presumably that's less app based. And then when it comes to NFTs, Apple is really cracking down on payments being processed outside its ecosystem. Until now, NFTs may have been used as a workaround to Apple's app store fees. We even talked about that on the crypto roundtable, that it could be like a little bit of a Trojan horse where you would offer an NFT for free. And then that NFT itself could be used as your ticket or your kind of purchasing opportunity outside of the ecosystem. A little bit of a hack that will not be allowed anymore. I blame ourselves. I think it must have been because we talked about it on the crypto roundtable. And now Apple has shut that loophole right down. It said that apps are allowed to list, mint, transfer, and let users view their own NFTs. However, the ownership of NFTs should not unlock any more features within the app. Womp womp. Users can only purchase NFTs through Apple's in-app payment system. And then the apps may not include buttons, external links, or other calls to action that direct customers to purchasing mechanisms other than in-app purchase. This is aggressive. And coming on the heels of that ruling in India that we just talked about, it might be aggressive in a way that backfires a little bit on Apple. Certainly, there have been a lot of complaints already. And this, you know, at a moment when it seemed like Apple was enabling a richer NFT ecosystem, this is 100% an attempt to like take the toothpaste and shove it right back into the App Store tube. Apple generally, of course, once again, receives a 30% commission on purchases through the App Store that has been lowered to 15% for smaller users, but that's on all transactions made through its platform. The reduced rate was only introduced at the beginning of 2021 because there was so much pressure and pushback on Apple about these fees and percentages. Apple has argued that these commissions are necessary to keep and maintain safety and security on its platforms, which is one of those situations where two things can be true at the same time. On the one hand, that has been true for a long time and developers prefer to be in the Apple ecosystem because it is more likely that those users will pay for apps and users feel like they can trust apps that they get from the App Store. However, this is already backfiring from at least some expected sources. Tim Sweeney, the CEO of Epic Games, creator of Fortnite, which of course has been in a lawsuit with Apple over this very issue, commentated on the announcement tweeting, to cryptocurrency enthusiasts, this means Apple is now adding a 30% tax on your so-called true ownership of digital goods. To cryptocurrency detractors, this shows Apple's motivations are only money. For digital items, they support the NFTs they tax and ban NFTs they don't tax. Of course, as I mentioned, this is not the first time that Epic Games has voiced its opinions about Apple's App Store tax. It blasted Apple back in 2021 after Apple unveiled those decreased fees for smaller developers, saying, by giving special 15% terms to select robber barons like Amazon and also now to small indies. Apple is hoping to remove enough critics that they can get away with their blockade on competition and a 30% tax on most in-app purchases. The two companies, of course, went to court. At one point over violations of antitrust law, Apple was found not guilty. Spotify has also complained about this cut, saying that Apple's quote, tax cut, illustrated the fact that the company's App Store policies are arbitrary and capricious. The reduction of App Store fees to 15% for small developers does nothing, Spotify wrote, to level the playing field to ensure Spotify can compete fairly with Apple Music. I guess the only good news here, if you can call it that, is that now NFTs are playing by the same rules as every other app on the App Store, but certainly from the perspective of the potential of NFTs themselves, that I think is going to be a big bummer. We will definitely make sure we ask Sunny and Vinny about this in the next crypto roundtable. It will also be curious to see, again, I want to highlight this contrast between what just happened in India with the Google Play Store and Apple's continued aggressiveness in terms of funneling all users through its own payment system where it gets a cut every time. I have a lot of questions about where this is all going to end.