Chamath Palihapitiya: Okay, everybody, we are back on a Tuesday. Thanks to my guy, Andrew Gazdegi of Acquire.com for joining yesterday.

SPEAKER_01: If you missed that episode, I know some people were out volunteering and taking the day off like we did and observing MLK Day. If you missed that episode, it's really worth going backwards a day because a lot of our founders are in the process of doing M&A because maybe funding is not available and it's time to get some returns for you, your team, your investors, no shame in the game. If you got to sell your company, do it right.

SPEAKER_02: And this is a great, great episode for you to go backwards and listen to yesterday's. But today, tons of news, Molly.

SPEAKER_04: Yeah, exactly. We're back with a bang twist going up on a Tuesday.

Jason Calacanis: I'm sorry, I didn't want to do it, but the way you said it, I just had to. Okay, we got a bunch of generative AI news. Of course, Getty preparing to sue stable diffusion creator stability AI, Meta partnering with Shutterstock to train its AI, Microsoft CEO Satya Nadella announcing that Microsoft will integrate OpenAI software into Azure, its cloud computing platform.

SPEAKER_08: We're going to kind of cover our, the debate, the controversy, the potential. It's a good conversation.

Chamath Palihapitiya: It's a great conversation. Then we do some startups and we have a long, thoughtful talk about hardware startups. We got a pair of hardware startups today announcing funding and products. One of them is building an AI enabled lighting rig and solution for indoor farming. Super important, depending on the crop you're growing to optimize that. And then Mill is a startup by the former Nest co-founders making an automated compostability

SPEAKER_01: trash, which is super interesting. And Molly and I were both enamored with that. But we talk about the lessons for other startup founders around hardware.

SPEAKER_11: Yeah, definitely. And P.S. Bring me your climate software companies.

SPEAKER_12: Absolutely. Scalable, please. High margin businesses. Hello, ones.

SPEAKER_16: I'm here for you.

SPEAKER_17: I'm here for you. It is going to be a great show. So stick with us. This Week in Startups is brought to you by Squarespace. Turn your idea into a new website. Go to squarespace.com slash twist for a free trial. When you're ready to launch, use offer code twist to save 10% off your first purchase of a website or domain. And Agetech Collaborative. Startups. Your go-to market team is waiting. Agetech Collaborative's cutting-edge accelerator program connects you with investors, test beds, like-minded innovators, and industry expertise. They're taking Agetech to the next level.

SPEAKER_20: Join them at agetechcollaborative.org slash twist.

SPEAKER_25: Okay, everybody. It is not Monday. It's Tuesday. It's that Tuesday. It feels like January is the most stutter start month, isn't it?

SPEAKER_05: Like, you're kind of like, you're starting to creep back, but like, people aren't totally at work. And then there's like the post-holiday sickness. And then there's the random three-day weekend, just when you were starting to get a rhythm. Like, I find January very frustrating. I'm just like trying to get back into it. Yeah. But you're not quite there.

SPEAKER_30: And then like, now my kid's sick. And it's like, oh, yeah.

SPEAKER_29: And I'm still sick. And you're sick. I was, had that. My daughters got sick a week ago.

SPEAKER_31: And I got sick. Like at the end of the week, I was sick all weekend.

SPEAKER_32: But, you know, I'm looking out my window over here and I see sunlight. And so the rains have stopped and the sun has come back out today. So thank the Lord. You know, I was thinking about this like sick thing and my wife's like, oh my God, you need to take a day off. And I was like, do I, because I have like a tickle in my throat. You may hear me cough maybe during the show, which I find terrible during podcasts, by the way, I, my advice to other podcast hosts for my listening pleasure is do not show up with a scratchy throat.

SPEAKER_37: But for me, I have to be here.

SPEAKER_39: You really don't. We could have called in, Sonny.

SPEAKER_25: We had a plan. You don't sound bad. I, what I hate is if you did sound bad, I would have been like, get the hell off my

Jason Calacanis: Zoom because it, it is sort of just like, like, there's a lot of work you can do when you're sick, but if you sound like crap, nobody wants to hear that. And I feel like there's a tendency to, for people to just be like, it's fine. I know I'm like.

SPEAKER_32: When I can tell the person's in pain from their sore throat, I'm just talking about sore throats for people in our profession.

SPEAKER_46: If I sense you're in pain, it makes it painful for me to listen to the podcast and I want

Chamath Palihapitiya: to change the podcast. Right. No matter how good the conversation is, if you have that raspy, I can tell you're having a hard time speaking.

SPEAKER_50: I, I actually have changed the podcast when I've heard. Don't do it. Right. Yeah.

SPEAKER_51: I mean, this is a performance, right?

Jason Calacanis: Like we don't want to give people 70 or 80%. Like when you sound like crap, it's just not that I think is like a hard line. If you sound bad. Okay. Don't do it.

SPEAKER_53: Well, I mean, and here's an opportunity for AI, like touch up my appearance.

Chamath Palihapitiya: That's what we were getting right now. We don't need makeup on podcasts anymore. I used to have makeup artists come to the studio when we did live tapings because the HD was so unforgiving, but on zoom, it touch up my appearance at 30, 40, 50% is the same as makeup.

SPEAKER_01: So makeup artists, I used to spend $10,000 a year on this podcast on makeup artists. It was a hundred bucks. Every time a guest came in, then it became like 200 inflation. And, uh, the show used to be two or three, two guests a week, maybe. So it was 400 a week, 50 weeks a year. Yeah. It was even more. Yeah. Maybe 20 grand a year. And so we had a makeup line item, no more because of technology.

Chamath Palihapitiya: Now, if we can make ourselves look a little bit better here on the margins with the touch up, why can't we do that with somebody who has a cough or a tickle in their throat in audio? I wonder if for the producers, if there is a filter for a raspy voice, I was thinking, Molly, I, the, one of the benefits of remote work is I can work when I'm sick.

SPEAKER_32: Yeah. Which maybe is a sickness in itself.

Jason Calacanis: A little bit, a little bit. It's well, it depends, right? If you, cause there are definitely times when you are too sick to like get dressed and shower and commute and you don't, you'd be infecting other people if you were like on a bus or Bart or something.

SPEAKER_05: And it's like, you know, like that's really real, but you might not be too sick to work.

Jason Calacanis: Like it's pretty rare. And even if I'm too sick to talk, for example, like couldn't be on the pod, it's pretty rare for me to be too sick to like email. I think the last time I can think of in the last 10 or 15 years that I was legitimately too sick to do anything was COVID where I just was like, oh yeah, I can only sleep. Like I couldn't even watch TV. It was too much input. It was awful. But yeah, I don't know. I mean, it's true.

SPEAKER_68: Like you can power, but it, it does make me wonder if then does the sickness last longer?

SPEAKER_25: Cause you should just be resting. I don't know.

Chamath Palihapitiya: I don't know if resting, how much that actually helps. Um, but I, you know, I'm always incredibly frustrated.

SPEAKER_31: The only thing that four to five times when I'm sick that I don't want to get people contagious and I wouldn't have gone to an office. I still would like to get worked on and I would work from home.

David Friedberg: And now that everybody's working from home, it works just fine. So anyway, that's interesting.

SPEAKER_74: Um, but if you sound like crap tomorrow, let's just keep an eye on it or an ear on it. If you will. Yeah. Haha.

SPEAKER_76: Listen, if you want to be an entrepreneur or you want to do a side project, square space is an amazing place for you to start. Why? Well, it's the platform where you can build or sell anything. I've been talking about square space for a decade because it is the absolute best place for you to take that first step in being an entrepreneur or starting a project or even putting up a portfolio of your work. We'd love it at launch. We use it for all our different projects. When we have to put something up like remote demo day.com, man, we can get it up and running in minutes. The feature train from a square space just keeps coming. The product velocity at that company is exceptional. And when they ship, they ship it. Beautiful, beautiful templates, inventory management, APIs, advanced analytics, 24 seven, 365 day a year award winning customer support. And Hey, it's going to look great on any device. They also have appointment scheduling. So let's say you're a trainer or something, or you're a coach, and you need to manage appointments. Great. You want to take payments for content or for a product. You can do that. You don't need to hire an army. You just need to go to square space.com slash twist and start a free trial. That's all you need to do. Square space.com slash twist. And make sure you use the offer code twist for 10% off your first purchase of a website or domain.

SPEAKER_77: You can get your domain there at the same time. We love you, square space. Thank you on behalf of all the, the startups, capital allocators, and tech enthusiasts who listen to this podcast for being our longest running partner. It means the world to us. Square space.com slash twist. Use the promo code twist, please. All right.

SPEAKER_32: I, I think I predicted for the last year that there would be a ton of lawsuits. I did not predict they would happen this quickly, but there is a clip of me a year ago talking with you about Dolly where I said, if you're a photographer, I encourage you to call Getty if you sold them your photographs and to say, what are you doing about our IP being exploited? How am I being paid for this AI training to call, you know, your stock photography, because this is going to remove the need for stock photography in the world or stock illustrations.

SPEAKER_83: And this morning I saw, uh, across the news headlines, um, and the wires that that's exactly what's happening.

SPEAKER_05: Yep. Getty images is suing the creators of the AI art tool, stable diffusion for scraping its

Jason Calacanis: content saying that, uh, stability AI, the tool in question, unlawfully scraped millions of images from its site. It is being described as a significant escalation in the kind of, as you predict the, the, on the long incoming copyright battles between, uh, copyright holders and some of these tools,

SPEAKER_05: the AI firm is arguing that the amalgamation of data for products like stable diffusion is covered under the fair use doctrine, however, there are some glaring signs here, including an analysis of the stable diffusion's dataset, which found that it contained a large amount of content from Getty images that is watermarked and that when, oops, exactly. And that there are examples of images and this is amazing that have the watermark right

SPEAKER_90: on them, well, so stable diffusion created images and the AI produced the Getty watermark.

SPEAKER_94: Yeah, because clearly those images are in the dataset, right?

SPEAKER_05: And so when stable diffusion, you know, it was like, oh, uh, give me a look at two, uh, misshapen trolls playing football, one with a freakishly long arm. I don't know. This is just what I'm looking at here.

SPEAKER_98: Then it was like, okay, cool, here you go. And then included in that image, because clearly this is the origin, the dataset is the Getty watermark.

SPEAKER_102: Now that feels like a pretty easy one to have solved ahead of time.

Chamath Palihapitiya: If you were trying to steal, uh, and that would show a ton of intent. What this shows is they did not have the intent to steal because they didn't try to cover it

SPEAKER_32: up, right? Cover up is greater than the crime. Um, I think they were not thoughtful and they didn't consult lawyers when they created this technology. Any lawyer would tell you this is a derivative work and I'm not a lawyer. And I told you this and that you're infringing on the creator's right to monetize their content in the future and in future mediums.

Chamath Palihapitiya: So when you make something, you have the right to it, not only in the mediums that exist, but in, in future mediums, et cetera.

SPEAKER_32: Um, so, um, this is, uh, a serious violation. It will result in a settlement of millions of dollars. I predict, uh, we may not get that settlement. It may be done quietly. Um, but yeah, they're gonna have to remove all of this. And this would be the equivalent of like, I stole your design.

Chamath Palihapitiya: Like I took a design of a dress or something from Gucci and I stole the design of the dress

SPEAKER_83: and I included the Gucci logo on it. Right? Like, you know, the person's guilty, right? When you're selling the purses that you, it's just a straight up knockoff.

SPEAKER_107: Well, and this is a, this is an interesting, this is just such an interesting.

Jason Calacanis: And I want to do a small correction here. The lawsuit, uh, has not officially been filed.

SPEAKER_05: Getty has commenced legal proceedings in the high court of justice in London. So it might be a little bit of a linguistic difference, but just to be as specific as possible here, which getty was, which, uh, clearly the stable diffusion folks were not, so this gets to a couple of things, data and data

Jason Calacanis: integrity and data cleaning. So like there's the saying garbage in garbage out. And in this case, what we're seeing is copyright in copyright out.

SPEAKER_05: It should have been a fairly trivial task to filter watermarked images out of your data. So I do that. Like almost every, like for the last couple of years, I've done this presentation about data to graduating, uh, the graduating class of Notre Dame state of science school. And they offer like a master's and literally one of the things I say to them is like, Hey, FYI, for the first few years of your career, as a data scientist, you're going to be a janitor because cleaning data is the key to providing usable data to models.

SPEAKER_98: And so obviously whoever trained this AI did not, or could not be bothered to just filter out images that are watermarked. Like that is the most basic level of care. And it makes me like very worried about what is going into these models on every level, right? Whether it's like, I don't know, massive troves of racism or stolen or personally identifiable information. It's like, okay, what is the, what are the data sets that these things are being trained on?

SPEAKER_39: Who's in charge of cleaning them up? Even the tiniest bid you'll left watermarked materials in there.

SPEAKER_32: I mean, it's pretty crazy. Like where did they source their, and this is my citations as we've been talking about the last couple of months. And I'm saying like, well, where are the citations and chat GPT and Google is putting them in.

Chamath Palihapitiya: So Google identified this issue, the disruptive new companies, part of, and this is a bigger startup lesson, part of being a new startup is to move fast, break things, and to beg for forgiveness instead of asking for permission.

SPEAKER_31: So as offended as I am, as a content producer, or as offended as you might be, I'm not sure the degree to which you, you find this offensive. I suppose people in the content business always find it a little bit, yeah, I think if you're in the content business, you find it offensive.

SPEAKER_120: It's watermarked, I mean, for God's sake.

SPEAKER_32: Yes. And so this is, uh, stealing. This would be like taking your photo library off of your phone and then creating this AI based on your personal photos. These are another group of people's private IP. They own this.

Chamath Palihapitiya: This is stolen material. Let's call it what it is. They train their AI on content they stole from another party who does commerce with it. They should break the rules. Let me explain. I am not encouraging you to this, but this is what we've learned in commerce and in industries.

SPEAKER_124: Go ahead and break the rules because what has, how much money has stability, stable diffusion, uh, or stability AI? One of them is the open source product. One's the commercial, right? So this company, how much money have they raised on doing this stolen material demonstration?

SPEAKER_05: At least, uh, their most recent raise appears to have been a hundred and one million dollars. That was in 2022. And I think they raised before that. Oh, and they're valued at a billion dollars post money.

SPEAKER_124: Okay. So you have a billion dollar valuation. You had a hundred million dollars in cash at a minimum. Who knows what else has happened? Yeah. Um, pays for a lot of lawyers, uh, pays for a lot of settlements. Right. So message to the market, if you're Airbnb, if you're Lyft or Uber, uh, whoever you are, and you're doing things in the world, go ahead and break the rules, raise the money. And, and I'm not saying I agree with this or endorse it. So don't clip this in a funny way, aggregators. I'm telling you what capitalism has learned.

SPEAKER_133: As capitalism has taught us. This has trained the, the data that founders have been trained on. Yes. Yeah.

SPEAKER_134: Exactly.

SPEAKER_124: So I am not endorsing wholesale stealing of other people's IP to raise at a high valuation, uh, and then go spend that money on lawsuits and settlements after the fact. But the fact is getting images deserve that a hundred million dollar investment.

SPEAKER_32: Uh, the majority of what was just demonstrated by, uh, stability AI, uh, was based on, if it was based on their work, what we will learn the second order impact of this. Is the people who should own this business are the people who own the content and the IP, and they should go take the stability AI open source project.

Chamath Palihapitiya: They should go higher and steal away people from stability AI. And they should make a product that creates images on the fly when you have a subscription to getty images, which people do. And it should then pay royalties to the original source material and getty should then make sure that they're, uh, they have permission to do so.

SPEAKER_138: Hey everybody. It's time for a special interview with an old friend of mine, Rick Robinson. He is the GM of age tech collaborative, which is brought to you by our friends at a ARP, which I think, correct me if I'm not at 52. As of last week, am I able to be in a ARP now? You are, you are welcome to join. Oh man. You and I got old, what happened? This is amazing. You guys are really excited, uh, about engaging the startup community in building technology for folks who are getting up there in age.

SPEAKER_145: Yeah. It's really exciting for us developing the age tech collaborative to try to put a focus on what we call age tech. And you might be wondering like, what is age tech? Well, it's, uh, it's the intersection of longevity and technology. Really? These are health tech companies. These are fintech companies. These are wellness companies. Essentially, it's going to be almost every company because the market 50 plus is becoming so enormous that they can't be ignored. And then you've got a lot of people who are supporting that market who can be any age. It's kind of a white space because not a lot of product developers and marketers and startups and investors have put a lot of focus on this, but it's huge and it's growing. In fact, it's around eight and a half trillion dollars in terms of economic value in the U.S. right now.

SPEAKER_142: All right. Thanks, Rick. When you're selling into the 50 plus market, having a relationship with AARP gives you a bunch of credibility.

SPEAKER_76: Of course, in the meantime, you can go learn more about the age tech collaborative at age tech collaborative.org slash twist. And join us later in the program. You hear more about the age tech collaborative and how they help innovative startups succeed.

Jason Calacanis: So I have a little bit of a hot take around this because there are a couple other lawsuits and a partnership. What's your take, Molly? Well, OK, so as context for my take, GitHub, I'm sorry, another lawsuit has been filed against GitHub's co-pilot, which is sort of the code creation tool. Yes. There's an announcement that Meta and Shutterstock have partnered so that Meta can train its AI on Shutterstock's catalog of images. And so Shutterstock is going to make some money here. But what I would say is no matter how this shakes out, and I think you're right, that there will be a lot of lawsuits. Getty didn't create any of this stuff at once. Neither did Shutterstock. Like creators are going to get screwed once again. This is exactly reminiscent of the digital music wars back in the day. And in fact, Getty's CEO likened this whole thing to Napster and Spotify and saying, well, Napster was just stealing, but Spotify negotiated these deals. And so Spotify figured out a legal way to do this, and that's what we want these AI tools to do. And they probably will. And there will be a Shutterstock model. And then Shutterstock will make money and Getty will make money. So when I was in college, I covered the Unabomber arrest for my school newspaper. Like I was there on the mountain when they brought him down, there was a team of four of us, three photographers and myself, and a white Bronco like mysteriously came down and we were like, well, that's weird.

SPEAKER_05: What's that car about? And so our two, two of our photographers were like, we're going to follow it. And I was like, great, go do it. So they follow this Bronco all the way from Lincoln, Montana, this little town to Helena. And they proceed to get the first photo of Ted Kaczynski post arrest. Wow. A couple of college kids. Good get. At which point they start getting calls from Getty. Sure. And companies like them. And the AP being like, you have to give this to us for free.

Jason Calacanis: Otherwise, you'll, you know, be dead to us in this industry. And Getty being like, we're going to offer you the lowest possible amount that we possibly can and da, da, da, da, da. And they negotiated a whole thing and it was great. And I think they got college paid for. And it was awesome. I think they ended up selling the photo to Getty.

SPEAKER_05: But when that photo ends up in the AI output. And Getty gets a royalty. They will not. So, like, I just, this is a really bummer, dystopian take. Well, they sold the image. But the content aggregators. Sure.

SPEAKER_98: They sold the image for, like, as low.

Chamath Palihapitiya: They signed a contract knowingly selling all rights to the photo for all time in perpetuity.

SPEAKER_98: Yes, and every artist who signs with a record deal because they want to get famous signs away the bulk of their potential. Like, I'm just saying. Why don't they get royalties?

SPEAKER_05: That the labels and the royalties are tiny. And look at what artists have to say about what they get paid from Spotify and YouTube.

SPEAKER_159: I'm just saying the creators are going to get screwed again. Like, these aggregators of images will make the bulk of the money.

Chamath Palihapitiya: If people sign a contract and they have an attorney, I would say that's, you know, much different magnitude in terms of getting screwed than having somebody steal your content, obviously. Right?

SPEAKER_05: Yeah, sure. I'm just saying, like, if you, like, look at where the money is going to flow, it will be the exact same thing as a musician saying we're making pennies off of Spotify streams. Yeah, I mean, I'm under no illusions that we're standing up for the, like, little guy here.

SPEAKER_159: We're standing up for Getty and Shutterstock and the people who have the money to come along and say, like, I know you're, like, a starving artist and you'll do anything you can, sign this contract.

SPEAKER_74: And they do.

SPEAKER_174: I'm not so sure. I mean, the value of taking a picture today. I know, but that's because you're, like, an uber capitalist.

SPEAKER_175: Yeah, I mean, I think there's a trade.

SPEAKER_124: I'm not, I'm putting aside what happened, like, you know, decades ago when musicians didn't have representation or photographers didn't have representation, but there is a value, you know, to the content that they provide. And if they sell the rights away and they had representation, um, hopefully they had good representation, you know, I have less of a problem with it. And, you know, the value of certain things in society.

SPEAKER_75: I mean, how many, yeah. What's that? I'm just saying that's my hot take. Yeah, no, no, I, I, I get it.

SPEAKER_124: I, I always look at, like, did the person have representation? Like, did these people who sold their images or just a professional photographer have it? And then there's another thing that's happening.

SPEAKER_186: Usually no, but maybe.

SPEAKER_124: Yeah. I mean, and a lot of times people get screwed on their first record deal and they screw the record label on the second or they, you know, they don't get, like we talked about last week with, you know, getting 5 million for your first picture.

SPEAKER_125: And then getting 15 or 20 or making a back end of 75 million on the second, and then you have them over a barrel because you're Robert Downey Jr. How do they do Iron Man without you, right? The whole franchise is, you know, based upon you.

SPEAKER_124: So, um, but really, you know, the other hot take here is I think it's, if we believe this is going to be a commodity, this technology, right? And they're based on open source projects, which you can fork.

Chamath Palihapitiya: And I think Sam Altman kind of said that last week when you went to that thing, that this would be commodified in some way. Kind of, yeah. And there are many projects working on it. You know, I, I think the choice will be either stability chat GPT and, um, you know, Facebook and other Google, other places, if they want this technology, um, they will do partnerships with those companies or those companies will embed it in theirs. Those companies used to, if you remember, like, they would partner with, um, content people would partner with Yahoo to host their content. And then eventually they're like, wait a second, I can put up my own website. So this will be part of the process. I think the next shoe to drop will be somebody will create software to sell to the Gettys of the world to make their own derivative products. So they'll sell that content and then Wikipedia will make their own, uh, AI tool to write content based on what's in the Wikipedia already.

Jason Calacanis: Would that mean that somebody created the tool to like that middleware layer?

SPEAKER_05: Like, here's how to turn your data into a chat GPT. Like, isn't that, wasn't that sort of the original goal of something like open AI was to open source the technology to do that to secure the bag?

SPEAKER_74: And I think that was like, one of the things that stable diffusion said too, right?

SPEAKER_194: Well, stable diffusion is based on an open source.

SPEAKER_74: Or stable AI, right.

SPEAKER_75: Stability AI.

Chamath Palihapitiya: Stability AI is based on the open source. So it's the WordPress of that. Right. You know, and here's the other rub. Um, eventually this stuff will make it to the desktop and what you do on your desktop is your business. And isn't subject to the same rules as another company providing services. So if you load a webpage to your computer and you take the ads out and you do that on your computer and you don't republish it. Um, or you cut and paste some content and you make a document with, you know, the hundred stories on, you know, Superman, uh, that you found in the New York times, you're allowed to do that for personal use. You can do all kinds of things that break copyright law, uh, that would be breaking copyright law in a commercial venue when you're doing it privately. So here's another possibility. I have my own version of stability AI or chat GPT running on my computer. It downloads sources of information and it makes that content on my computer.

SPEAKER_124: I'm not paying anybody to do it. They're not profiting from it. I'm just doing it personally. Number one, people will not know that you did it. And number two, uh, so just practically, it's kind of hard to enforce number two, it's just one person. So how is it causing a problem?

Chamath Palihapitiya: You know, for, um, you know, the original copyright owner. And so that would be the equivalent of making your own remix of a song in the privacy of your own home and playing it for yourself.

SPEAKER_01: These tools are so powerful. I get the sense that eventually you'll be doing this on your personal computer. And then I think that's very hard.

SPEAKER_05: Well, and I, we are, we have not updated. I mean, I've been ranting about copyright and intellectual property laws for 20 years. Like, yeah, we are not prepared for this conversation. Like fair use remains somewhat uncodified. And it sort of argues that if it's a derivative work, it should be fine. Like the reason Getty is likely to be successful in the short term is because there are freaking watermarks on the images that Dahlia is producing. Right. So like the data, the lack of care with the data is incontrovertible. They took images that are incontrovertibly owned and put them in their data set. However, fair use does provide for derivative work.

SPEAKER_98: Yeah. So in the absence of codifying that really, really specifically into law, which we never really have, everybody just kind of guesses at what a derivative work is. But that, that is the stuff that's going to have to be worked out as we enter the era of AI, because there, there, there will be lawsuits that are lost, but there will be lawsuits that are won by Gettys and Shutterstocks and whatever, because derivative works is pretty vague. And that's why, like, when you declare, like, you know, when you're always like, this is it, this is going to be the lawsuits and everybody's going to lose and whatever. I'm like, not all of them are going to lose because derivative works is protected.

SPEAKER_209: Yeah.

SPEAKER_210: It's just that we don't know what that means yet. I have extensive experience in this. And what I will say. What did I just say about 20 years?

SPEAKER_122: I have a lot of experience in this.

Chamath Palihapitiya: What I'll say is it's what has been codified is a test, which is to interpret, to be interpreted by a judge and a jury. Now that test very rarely gets applied because the chances of winning when you're doing something commercial are so low that people settle. And so the history of this is that it never goes to the mat, it gets settled long before that because it favors the content producer so much in a commercial arrangement. So if something was non-commercial, then, you know, there's very little reason for a lawsuit to be brought because there's no money to be made or lost. And so where you see this is in reaction videos and on YouTube. So YouTube, you know, has been an amazing proving ground of not having to go to court to mitigate or to reconcile these debates, right? Uh, to manage them and what typically happens is if money is turned on or off, if monetization is on or off on the video becomes a deciding factor when you're using our content. If you were to take this discussion and to make a reaction video to it and depending on how much you use the percentage use, um, you can kind of short circuit the entire discussion simply based on if monetization is on or off. Right. That's the key.

SPEAKER_01: And so here, uh, and then the further test of it's, is it going to make it impossible for that person or impact that person's ability to make money in the future? So it's a four part test. You can read about it on the Wikipedia. It's pretty straightforward, but in practicality, um, if you use other people's content on YouTube, uh, the fair use comes in when it's educational, the percentage you're using.

SPEAKER_05: And if you're, uh, monetizing it and what you can see now, if you're watching it is all of the ways that people have figured out how to play huge chunks of that movie.

Jason Calacanis: Megan, that we were talking about, uh, in the pre show, I think, um, by saying that it's a reaction, right? So what I'm saying is this has yet to be everything.

SPEAKER_05: Yes. To everything you said, this has yet to be litigated to the mat because in most cases, what you had is a giant, right? Like you had the DM, you had, um, the MPAA or Sony or Getty coming and saying you infringed my copyright and people were like, I'm tiny, I'm settling. Or yes, they just took it down. Now you have Godzilla versus Mothra and you have an area of the law that is still unsettled. So you got open AI on the one hand, right? Or stable diffusion on the one hand, you've got a company, you've got chat GPT. You've got a company that's worth 20 billion, $29 billion. Now today coming up against Getty or coming up against all of these copyright holders. And what I'm saying is this is gonna go to the mat now. And this area of law is still not totally codified. Fair use is still somewhat vague, vague enough that I think we're gonna see this fight.

SPEAKER_40: And I don't think it's a slam dunk. Yeah, I'll disagree. I think all these companies are gonna fold. I think they're not gonna be able to.

SPEAKER_220: I'm not mad. No, no, I just, yeah. You think open AI is gonna fold? Like they're just gonna be like mad.

SPEAKER_199: A hundred percent. A hundred percent. They will move towards only scanning data that they have access to and they'll get permission and they'll pay licensing fees.

SPEAKER_32: And just like Napster, because I, I heard this exact argument like Napster had a lot of money and you know, the music industry, uh, was ruthless about it. The same thing with YouTube.

SPEAKER_223: But now the act, the information they have access to is the entire internet.

SPEAKER_98: Like once you turn on Google and connect one of these things to search and the repository of human history. Yeah. Like how do you have permission stop that?

Chamath Palihapitiya: Um, with lawsuits. Yeah. Lawsuits become the backstop. And once you, once people see the possibility of the existential, uh, basically, uh, demise of

SPEAKER_83: their business, which YouTube did, that's, I mean, they would have been an independent company had they not had those lawsuits. Lawsuits force you, uh, to settle and you will lose them.

SPEAKER_32: I think that it's, I don't, I don't think it's that that the law isn't codified is that the law

Chamath Palihapitiya: is designed to not give a bright line so that there's less abuse. And so that the market works it out. I think that's why it's like that, but it also depends on which market you're in. There'll be, I think this might be filed in, uh, the UK for a reason. They may have started there because they know they're going to win there and they know content protectors, just like people will jury shop here in the United States and they know California juries are super anti-capitalist and Texas ones are, and Delaware ones are super pro-capitalist. Yeah.

Jason Calacanis: And so on the plus side, if that happens, then the data sets that the AI is trained on will be necessarily way smaller, which could extend the life of humanity.

SPEAKER_05: So win, win, win, win. Also it will just takes longer for them to get smarter than us and kill us.

Chamath Palihapitiya: So I, I think there is a question is, you know, would the results be as impressive? Exactly. They didn't have, I think that will limit our data sets.

SPEAKER_05: I 100% think that limiting the size of the data sets will make the AI's dumber. Yep. And that's a win for us. So I just talked myself into the lawsuit side.

SPEAKER_237: Yay. Let's go.

SPEAKER_238: You know, it's, it is unprecedented.

SPEAKER_239: Uh, that's for sure, uh, the scale of this.

SPEAKER_138: Hey everybody, it's time for a special interview with an old friend of mine, Rick Robinson. He is the GM of Agetech Collaborative, which is brought to you by our friends at AARP. So how does the collaborative work?

SPEAKER_77: How do you help companies and investors kind of access these companies and these markets?

SPEAKER_145: So essentially what we do is we look for companies, we incubate them, we invest in them, and then we bring them into this new environment. We call the Agetech Collaborative community. So yes, we have pitch competitions that we run throughout the year, themed, and some of them are open mic style. And it's a way for us to source and find great early stage companies, usually pre-series A. We invite some of them into our accelerator program, which is extremely high touch, eight weeks, four times a year, where we bring in aging experts. We help get them best prepared to deliver their product or service to the market. And as I mentioned, we often invest in these companies, and then they graduate into the Agetech Collaborative community, which is an online platform that makes up an ecosystem that we're developing that includes, of course, the startups, investors, testbed organizations, enterprises, and business services, all in this one online environment where they can support and draw from one another.

SPEAKER_138: Great. So there's an online community people can go visit, they can go visit that at agetechcollaborative.org slash twist, agetechcollaborative.org slash twist. And so if you want to build in that market, if you want to sell into that market, if you want to invest in that market, this is a great way for you to partner with AARP, correct?

SPEAKER_145: Absolutely. Yep.

SPEAKER_01: But you know, the technology is real. And now we've proven that it has some value. And so Microsoft seems to be the primary beneficiary of all this in terms of taking these research projects and actually getting them to consumers as a finished product, correct?

SPEAKER_05: Or at least out loud, which I think is interesting. So the news today is Microsoft CEO announcing at Davos that the company will integrate open AI capabilities into all of its existing products and said it will give customers access to build on top of open AI software. So the thing that you just said about, you know, take all of Wikipedia's information and turn it into a chat GPT. In theory, that is a bit of what Microsoft is doing. And it's doing that through Azure. It's a cloud computing platform. It means businesses will soon be able to use open AI's models in their own apps. And those tools include GPT three and a half, Codex, and DALI. We have, and I asked our producers to just pull a little bit of information on what some of the other big competitors are doing, because Microsoft, like, appears to be the leader here. And it could be true. And it could be because they're the ones integrating the fastest. Like, if you look about, if you look, you know, we know that Google has been working on AI that's so advanced that one of its own scientists thinks it's sentient. And we know that Meta, I think even Brad Gerstner has described them as, like, effectively a huge AI company that's, like, gotten really distracted by VR. They've invested hundreds of millions of dollars in AI and so have Google with DeepMind. Um, so Microsoft is taking the opportunity to leap out ahead, but I think the rest of the year is going to start getting a lot more interesting as some of these companies that maybe have been building behind the wall. Yeah. Like, let those monsters out in the open. I'm really into the monster movie analogy for all this today.

Chamath Palihapitiya: I, I, yeah, I mean, if this becomes like IBM Watson, which is fascinating to think that, like, six or seven years ago, IBM realized, like, AI was going to be the future.

SPEAKER_32: Yeah, having tools available, and they provided it to a lot of startups, they were very active, they sponsored a lot of our events, they were really active in providing Watson to startups, and they were just a little bit early.

Chamath Palihapitiya: Um, and now here we see it, open AI has essentially given Azure, their own Watson. And since some of these things are open source projects, if you look at the top 10, most active open source projects right now. Um, most followed, you know, most edits, all that kind of stuff, participation, the majority are in the AI machine learning space.

SPEAKER_31: So I think that this is going to be the future, just like cloud computing for storage, you know, or CPU and for CDN content delivery networks. Yeah.

Jason Calacanis: Yeah, yeah, it's good or even like LTE, you know, broadband availability, like this is such a this is going to be the layer of a vast majority or a huge explosion in products and services revenue.

Chamath Palihapitiya: This is then becomes the issue that we've been talking about with round tripping around tripping when one big company invests in another company. And then that other company gives them money back, according to reports, the revenue being booked by the work being done by chat GPT is Microsoft cloud revenue. Um, and so I think that that is now become codified to, to use the word from the previous discussion, uh, we're now getting some, um, uh, some feedback on that. So you have like Microsoft paying open AI to license this or investing in open AI, but then you have open AI hosted on Azure. Um, and that revenue is going towards Microsoft, so these like round tripping cloud deals, um, I think is going to result in some serious sec scrutiny, um, or just some scrutiny just to make sure it's clean, um,

SPEAKER_05: because let, just to put a finer point on it, what might be happening here. Um, and again, we don't know, but what might be happening here is that Microsoft is paying open AI to license the open AI software for Azure and that 75% of that profit then goes back to Microsoft because open AI AI then has to pay to use Azure's cloud computing to perform each.

Jason Calacanis: Function.

SPEAKER_124: Well, open AI, the question is open AI has compute costs, right? Right. A lot.

SPEAKER_25: We've heard this like 2 cents, 7 cents. So they pay whatever it is. And then Microsoft gets 75% of Microsoft opening as profits. Sorry, that was to close the loop, if you will.

SPEAKER_124: Yeah. So this weird investment deal combined with the fact of where is open AI hosting and where are they spending $3 million a deal, uh, $3 million a day, reportedly. So if they're spending $3 million a day on Azure, that means they're shipping 3 million from their bank account to Microsoft's bank. And Microsoft is licensing chat GPT software and investing in the company and getting some of their profits. 75%. So if, but there's no profits right now, but eventually there will be profits. Right. So this is just one of these co-mingling of money going back and forth that makes one wonder, you know, uh, if you were to, this is the, I'll, I'll give the least charitable and the most charitable interpretation. The least charitable is chat GPT is a massive compute resource, addictive product that could drive cloud revenue at Microsoft through the roof, which would drive their valuation at a multiple of that.

Chamath Palihapitiya: And that might be incredibly high profits, uh, for Microsoft. So chat GPT is the worst interpretation. Chat GPT created something to goose Microsoft earnings.

SPEAKER_249: Microsoft then invested in chat GPT on terms that none of us can understand.

SPEAKER_31: A 29 billion dollar valuation for a company that's losing, I guess, a billion dollars a year. Mm-hmm.

Chamath Palihapitiya: So why would they do that? Well, it's obvious in the, the explanation, the, the most charitable would be gotta be hosted somewhere. And if they hosted it on AWS, it'd be the same cost. And Microsoft invests in companies and people buy software from each other.

SPEAKER_01: So here's the 20 other companies Microsoft has invested in and bought and yeah.

SPEAKER_268: So Microsoft has, I mean, the middle ground interpretation is that Microsoft has made a lot of investments to, to beef up its Azure offerings.

Jason Calacanis: So the, the right, like, I sort of feel like the middle ground here is yeah, definitely. And they wanna make some money off of open AI, but the ultimate goal is what Satya Nadella announced today, which is integrate these tools into Azure. So make a little money from open AI, spending money on Azure. Sure. Make some money from open AI, whatever it becomes as a company, but fundamentally goose the product offering that is Microsoft Azure so that you can catch up with the competition and keep making money in the biggest money making space that exists in computing right now. Yeah.

SPEAKER_272: That'd be a kind interpretation. Yeah.

SPEAKER_274: Well, I think people are extremely jealous. That's just like the business, right?

SPEAKER_276: Isn't that just the business interpretation, like make Azure irresistible. Yeah. And if you happen to make a little incremental. Yeah.

SPEAKER_122: Feeding features to Azure.

Chamath Palihapitiya: Beefing up the offering is great. Yeah. The problem comes in when money is, you know, making a circle, right? So suspicious. And that's where. So suspicious. Get involved and say, is this, are we, are you increasing the velocity of this just to, you know, increase the value of both organizations in a non-organic way? Mm-hmm.

SPEAKER_283: And I, you know, I don't, I don't know how you prove that. Yeah. You're so suspicious about this.

SPEAKER_284: Is there like a little birdie that made you suspicious? Do you have some like inside reasoning? No.

SPEAKER_199: I mean, it's the talk of the industry. Yeah. I mean, there's a number of, there's, I think there's a number of factors at play here. Um, the non-traditional structure. Yeah.

Chamath Palihapitiya: And the flip-flopping of open AI from a non-profit to a for-profit. Right. Right. Then raising money on an insane valuation. Yeah. And raising all this money on, uh, Azure cloud. And I think then there's a whole level of jealousy of Sam Altman and the team over there for creating something that's worth $29 billion after presenting it as a non-profit. Yeah. And raising all this non-profit money. Sorry. Or maybe there's some resentment. I, I'm just hearing this constant non-stop back channel of why is this, what happened here? Mm-hmm. To go from a non-profit to a for-profit and then what's happening with this weird deal structure. Whenever you see a weird deal structure, everybody in Silicon Valley is like, what's going on?

SPEAKER_122: For good reason. Yeah. For really good reason. Because there were some weird deal structures going on in crypto too, right? Right.

SPEAKER_83: So I think right now there's just a heightened sense, uh, or people's, and listen, I'm not saying that there's anything funny going on here. I suspect there's not, um, overwhelming, overwhelming, I do think there's a heightened Theranos FTX climate right now.

SPEAKER_124: And if you present a really odd set of facts to the market, the market is going to be like, well, that feels more like this bucket of weird things that we've seen occur as opposed to standard deal structure.

SPEAKER_125: It doesn't feel like a standard deal structure. Yeah.

Jason Calacanis: It's not a standard deal structure. Yep. Right. Explanation and good insider. Cause it. Insider stuff.

SPEAKER_05: It's true. It's true. Jealousy. Non-standard deals are a reason for a red flag. And certainly, yes, there's a lot of jealousy. And I think also just maybe some disappointment, right? When you set some, whenever you set something up and say like, this is going to be, this is going to be the thing to save humanity. And then you see everybody like running toward developing it as fast as possible. Humanity be damned. And you literally have Sam Altman saying the worst case here is lights out for, for all of us. Yeah. It's like, isn't that what, but so you're just doing that though and cashing out like.

Chamath Palihapitiya: It does seem like this is the negative back channel is that this is a bag securing maneuver. Right. Um, and so, but again, I think people are highly jealous of Sam Altman because he's brilliant and he's made a lot of money.

SPEAKER_01: And, you know, this seems to be like the next Uber or the next Google or the next Facebook or the next Airbnb, right? Um, it's certainly evaluation trending in that direction.

Jason Calacanis: Well, let's go back quickly to the Satya Nadella genius, the goat factor of this, which is.

SPEAKER_05: That, uh, as a reminder of what's at stake and why Microsoft wants to juice Azure the product and has been doing whatever they can to do that over the years. AWS is Amazon's entire. Profit center. It's cloud computing business. Even on e-commerce, right? The only lose money I think on e-commerce now was like flat at best. And then there was some analysis that came out recently that was like, actually, I think Amazon's retail business might be like terrible. Um, Microsoft is profitable in multiple areas. So they're able to lose money on Azure, AKA make investments to make it better. But it does give them a serious advantage to, to beef up that offering and say, I mean, who wouldn't want to flip to Azure? If, if all of a sudden it's like, yeah, it can do all the, it's got chat GPT built in like, oh, okay. I'll take that for my startup. Yeah.

Chamath Palihapitiya: I mean, this is the classic, your margin is my opportunity moment. So Google's, uh, revenue comes from search.

SPEAKER_01: Microsoft doesn't need to be profitable on bang. They can be profitable on office and the windows franchises and their software platforms. Google launched their version of office with Google docs. And then you also have Amazon makes their money from, uh, you know, their profit center is obviously cloud. So yeah, this is the, um, crazy like risk board, you know, like the board game risk, you know, like where you can be aggressive in certain parts of the game and other parts of the game. You can be conservative and be defensive. Um, I think it all becomes a commodity real quick.

SPEAKER_80: I think it's gonna be like storage.

Jason Calacanis: Which honestly, I mean, if, if Sam Altman is as smart as he is being given credit for, then he clearly saw that coming. And maybe that's why he decided to secure the bank.

SPEAKER_305: It's like, now would be the time to sell.

Chamath Palihapitiya: Get the bag now. This might be the peak moment to sell before it becomes commoditized.

SPEAKER_32: And if it's open source, yeah, like, you know, why buy this if it's open source and you could fork it. Um, boy, does this make Google look smart for buying deep mind when they did in 2014. Right.

SPEAKER_01: Um, and incompetent, uh, for not getting more of the product out there to the public. They really have lost the PR, uh, race here.

SPEAKER_75: They really need to let the deep mind people start releasing stuff.

Jason Calacanis: Um, and that, that, I mean, there've been a couple, like, I mean, that whole arena of Google's business has just been a mess, right? Like, you know, they keep firing ethicists and then they had obviously that big drama with Blake Lemoyne's, the AI scientists, data scientists who said it was sentient and then got fired. And like a lot of weird stuff has happened in Google's AI division. And maybe that's why they have failed to capitalize with like a consumer facing product, but it's certainly, they might not be behind technologically, but from a commercialization standpoint, they definitely are. Yeah.

SPEAKER_309: And Facebook's got, uh, I mean, this is now I think going to be the dialogue for the next three or four years is the forget about crypto and web three.

SPEAKER_32: I mean, use case matters is another thing I take away from all of this.

Chamath Palihapitiya: I was out over the weekend and somebody said to me, uh, Jake, I was a very prominent billionaire, uh, venture capitalists said, I remember a dinner with you six or seven years ago. We were at this thing and you said 99% of these crypto things are griffs run by incompetence and, um, everybody attacked you and, uh, you held your ground.

SPEAKER_83: He said, it's the bravest thing I've ever seen. Like, uh, and you turned out to be a hundred percent. Right. And, uh, I said, you want to know why that is. And he said, yeah, actually, I would like to know why you saw that. And I said, it's very simple. That wasn't an estimate. I had met with a hundred companies. Exactly. And I wrote down a piece of paper, which ones I thought were frauds and which ones I thought were real. Yeah. The number was actually 99. I was reporting a statistic. You thought I was being hyperbolic. And it was a fact. And he said, is that true? I said, it wasn't exactly like a hundred on the nose.

SPEAKER_124: I had literally met though, with hundreds of companies, certainly over a hundred over a number of years, you know, two or three years and had them on the pod.

SPEAKER_83: And it was clear to me that 90 high 90 percentage were absolute incompetence and or grifters.

SPEAKER_124: And I think that's what we're seeing now, which is the reason I thought those things were not real was because when I said, show me the use case, explain to me what the user does. Mm-hmm . And how does this save them time, save them money? And, uh, it's one of the great luxuries of my life of being an outsider.

SPEAKER_125: Uh, I was an outsider in journalism. I had to create my own magazine really to get published. And, uh, I made my own way into the tech business.

SPEAKER_136: I didn't go to Stanford or Harvard or, you know, MIT.

SPEAKER_124: Uh, and I made my own way into venture capital, right? Well, all three aspects of my career. I was an outsider coming in. I, I didn't, I never came to it with some assumption that like, uh, I understood it. So I always just asked very basic questions. And the very basic question you have to ask yourself is like, well, what can I do with this? Right. When you ask what you can do with Dolly and ChatGPT, you can't stop thinking of use cases. Right. You can't stop seeing Tik Toks where somebody says, I took this book review. And I, I took these five book reviews. I asked it to make me a review of a book. Then I took it and made an image, uh, a series of images with Dolly. And then I exported it and put voiceover technology on it. And I made a YouTube video. And then I wrote a script to make me a hundred of these a day. So now I can publish my own reviews of books and summaries of books and publish them to YouTube and make money off of it. And you're like, okay.

SPEAKER_319: Yeah.

SPEAKER_320: And that's just one. Last week we killed law school.

SPEAKER_322: Law school. Right. Yes.

SPEAKER_05: Like literally last week, we got rid of the first year of law school. We were like, find me all the case law on this and show me the most winning strategies for beating a parking ticket.

Chamath Palihapitiya: We did it in our chat. Tell us remember we were doing like how many subscribers do Verizon. Yeah. And, uh, you know, T-Mobile have for their wireless networks and, you know, and asking it these resource questions. And we're like, okay, yeah, we now have like a producer chat GBT. Right. That can do some research for us.

SPEAKER_05: And so then not only did we come up with a bunch of use cases, but we came up with a bunch of use cases that have revenue slash money savings attached to this.

SPEAKER_25: Yes. Yes. That have economic impact there. We came up with products. Yes. Like it's two questions. It's what do you, what can you use this for?

SPEAKER_05: Yes. And how does it make money? And so far with AI, we've been able to answer that question or save money. Right. We've been able to answer that question almost every time that we have come up with something. Whereas for all of those years, I remember being at like collision in 2015 or something and asking God to explain the blockchain and being like, uh huh. Great. Yeah.

SPEAKER_98: I think I understand what you're saying from a technology perspective, but like, what's that for? And how's it, you know, like none of those questions could ever be answered. Yeah.

Chamath Palihapitiya: I mean, the immutability of it, uh, and the proof of work, proof of stake, all this stuff. It was like very interesting with very little use case. Yeah. Oh, for voting. And I'm like, you mean like if I wrote on a piece of paper, like who I voted for, for president and you had a copy of it and I signed it. Yeah. Like that. Okay.

David Friedberg: How is that different than the piece of paper that I signed that was digitized? It's immutable. It's immutable. Well, yeah.

Chamath Palihapitiya: And it's like, well, that's a, that's a really fancy word. Uh, that means you can't change it. Right.

David Friedberg: Okay. I can't change my signature on the paper either. So what are we talking about here? Yeah.

SPEAKER_334: My question Jason was when people would be talking about like an immutable decentralized Twitter.

SPEAKER_335: And you're like, what if someone posts like child porn or something? And they're like, do you really want that? Oh, uh, well, you know, the, the algorithm would probably just put it to the bottom, but it'll be there forever. Like, oh.

SPEAKER_73: Okay. Revenge porn, uh, doxing. Uh, yeah. Okay, great.

SPEAKER_335: Awesome. Yeah.

Chamath Palihapitiya: Yeah. There's an immutable website that can never be taken down with everybody's stolen photo rolls, camera rolls.

SPEAKER_341: Like, oh, great. What a contribution to society you've made. Yeah.

SPEAKER_32: Uh, you know, I saw this, um, you know, hardware companies are hard, but I saw one today and I was like, I'm so rooting for this company.

Chamath Palihapitiya: I I'll buy it. I would never invest in the company, but I love it.

SPEAKER_344: You know, this is like the siren song of hardware. You just want it so badly to work.

SPEAKER_08: Yeah. So second bubble of 2023 sustainable startup of the day.

SPEAKER_05: Yeah. Here we go. I mean, it really is too. It's super interesting.

Jason Calacanis: Actually, that was the last anecdote I have from that event. Seeing Sam Altman speak was that someone asked him, um, you know, you said it's a great time to start a company.

SPEAKER_05: Like my friend really wants to start a company. What should he do? I was like, I'm probably like an AI company. And I like scampered right up to him after. And I was like, climate tech starting climate tech company, like one or the other, right? Be intentional. There's going to be a lot of money for both, but like both of them are solving real problems that aren't going to go away. And I was like, you don't have to do nuclear fusion or carbon capture. We need software. Like we need measurement. We need all of this stuff. But it turns out we also do need, um, smart, uh, trash cans.

Jason Calacanis: Ah, turns out. We have, uh, this interesting press release. We don't know that much about funding or anything like that, but we know that there has been a release, uh, about a new product from the co-founder of nest. Matt Rogers, who founded nest with Tony Fidel back in 2010, Google bot nest for $3.2 billion in 2014. And in 2020, Matt went on to co-found another startup called mill industries with a guy named Harry Tannenbaum. And they just released mills own home device, which is a smart trash bin that decreases food waste.

SPEAKER_98: And it doesn't even just decrease it. It like compost it. It will dry the food out, neutralize the odor with charcoal, grind food scraps overnight.

SPEAKER_268: And then when the bill, the bin is full, they mail you a box and you send the waste back to the company and they use it to feed chickens. Like this is crazy.

SPEAKER_01: Yeah. There's a video. It's circular composting. There's a video on the website, uh, that I looked at this morning when I saw it come across the producer chat.

Chamath Palihapitiya: Yeah. And what's clever about this garbage can is that it's more like a compactor. So they're calling it a garbage can, like a compost garbage can. It's not that this is a very sophisticated mechanical appliance.

SPEAKER_32: This is more like an appliance. Yeah.

Chamath Palihapitiya: What this appliance does is what happens in the video is you dump in a bunch of, you know, I don't know, banana peels. You have your egg salad sandwich is the example they use and they show an egg salad sandwich being made. You throw the shells in, you cut off the crust, you put it all in there. There's a grinder, uh, that looks like an ice chopping machine. I, my dad had one of these in his bar. He liked to have chopped ice. And so I would take, uh, the buckets of ice, put it through the ice grinder and have little spikes and it would just rip it up and then we'd have shaved ice. And he put the pine glasses in the ice. It was very nice to have a frosted pint. Yeah. And so as you're seeing in the video, um, you then at, in the middle of the night, I guess it will grind everything. Mm hmm. And it grinds it so finely, and I guess it gets rid of the smell cause it's sealed and has charcoal. Mm hmm.

SPEAKER_124: That it takes up less space and unlike a compactor, which pushes it together because this is banana peels and whatever it grinds it. And then I guess it compresses it in some way. Mm hmm. And then you put it in a box and ship it to farms. Yeah.

SPEAKER_362: Then they take that milled product and they let chickens eat it to make more eggs. It's not, not simple, you know, like let's like, this is genius.

SPEAKER_363: Yeah.

Jason Calacanis: You have to, you do pay a monthly fee that includes the kitchen bin. It's $33 a month or $396 a year.

SPEAKER_05: Mm hmm. Um, if you actually, if you choose the monthly payments, the surface is $45 a month and the bin costs $75. They have a reservation system now for the bin. It has to be plugged in. You have to download an app so it can be connected to wifi. And then they're still working through. I will say though. I mean, I have like a little compost dealie that I got at Ikea because not very many cities have this. We have curbside compost pickup and green bin pickup, which is great. Hardly any cities I think have that. This you average you. The average household will empty the bin every two to three weeks, which is amazing. They're not quite weeks.

SPEAKER_368: Wow. Yeah. Like you could just be putting scraps in this thing for two to three weeks.

SPEAKER_05: That's amazing, but they are still work quote working through the necessary scientific and regulatory processes to establish a safe and nutritious chicken feed ingredient that we can distribute commercially. So they're a couple of steps away from making sure that they can feed this to chickens. However, amazing.

SPEAKER_374: Just as a grind or composting thing. That's amazing.

SPEAKER_01: You could just, if you have a backyard, like a compost pile, you could just throw it in the compost pile, depending on how finely they grind this stuff.

SPEAKER_25: Um, or if you even have a compost bin, you could put it in there every two to three weeks. Like I would be stoked about this. Even if even, even pre chicken feed. Yeah. This is freaking amazing. Yeah.

SPEAKER_249: I've never seen anybody come up with the idea of putting a grinder that shreds the stuff and makes it into like, you know, basically sawdust.

SPEAKER_01: That seems to be the, um, cool part of this.

SPEAKER_126: So congratulations. I think it's gonna be great. It's pretty amazing.

SPEAKER_05: And like, it's a pretty shocking, like the amount of, you know, if you think like when we interviewed John Doerr, he talked about thinking in gigaton terms, like food waste is a gigaton level solution. Yeah. If they can figure out how to turn this into compost that actually sequesters carbon and all of that, or even feed chickens or whatever, like that's a major impact.

SPEAKER_238: Not to mention just making life a little bit more awesome. Yeah.

SPEAKER_83: I wonder also when you use your, um, trash compactor, I'm sorry. Um, food, what do they call the thing in your sink that spins the food? Uh, garbage disposal. I guess. Is that a garbage disposal?

SPEAKER_385: There's some other word for it, but yeah, that's what I call it. There's a word for it.

SPEAKER_83: I forgot what the name of that thing is that we all have in our sinks here in California. We never had them in New York. They were illegal in New York city.

SPEAKER_25: Um, and those, uh, I'm really distracted by producer Brian, who lives in Canada, who calls it a garburator. It's a garbage. Yeah. Is that a thing?

SPEAKER_387: Is that a thing you call it in Canada?

Chamath Palihapitiya: I've never heard that word before. Anyway, garbage disposal. It's ripping apart, you know, your eggshells and we will put fruit in there or banana peels or whatever.

SPEAKER_01: I think you can put a pretty decent lemon peels or whatever.

Chamath Palihapitiya: You know, that in and of itself, um, I think puts all those nutrients eventually into the ocean, um, to be eaten by fish.

SPEAKER_341: I think. I think so.

SPEAKER_392: Um, but yeah, I guess it is called a garburator. Um, food waste.

SPEAKER_98: Food waste. It's called a garburator. Food waste accounts for eight to 10% of human greenhouse gas emissions every year.

SPEAKER_05: Like this is the stuff that gives you like an individual composting is a huge part of solving for that because it means that you reuse waste instead of have waste go to like landfill and be trucked somewhere and be this like big, huge mess. So it makes it smaller. It has a circularity aspect. The food itself is not wasted. It's not rotting in landfill. It fills and like off gassing. So yeah, there's a reason why.

SPEAKER_08: It doesn't fix it all, but composting is considered like a very large part of that food waste equation.

SPEAKER_01: Yeah. And if it doesn't have to be shipped anywhere, that clearly helps. Totally. And there's a reason why when you have these giant landfills, you see seagulls like by the thousands over them as they're picking it apart to try to find food. Mm-hmm. So yeah, better to just use it to feed more chickens. And keep them out of landfill.

SPEAKER_05: Landfill is like a big, landfill's a problem. I think in my austerity, my next.

SPEAKER_404: You should get some chickens. Eggs are really expensive now.

Chamath Palihapitiya: Yeah, I think that's kind of my next step in austerity measures is actually, wow. Actually, wow. Thinking about that, because of some influenza or something that happened, eggs went up to $5 a dozen, I heard, or something like that? For flu, avian flu, yeah.

SPEAKER_01: That, which is significant, I think, because I think we're paying three bucks a dozen eggs or something, two or three bucks.

Jason Calacanis: Somebody posted paying like $22 for 18 eggs. What? It's like eggs.

SPEAKER_410: Oh yeah, organic ones, good eggs, yeah. Yeah.

SPEAKER_08: You get the $5 eggs already.

Jason Calacanis: Huh. But yes, the average wholesale price of a dozen eggs in California was over $7 in January.

SPEAKER_415: Wow. 50 cents an egg. And I was at Costco yesterday.

Jason Calacanis: No eggs.

SPEAKER_98: Costco did not have any eggs. Zero. Crazy. Cause the cheap ones are like, you know, there's like a run on the cheap ones.

SPEAKER_01: I love the idea of people being more independent and having chickens on their property.

Chamath Palihapitiya: I, in the, in the peninsula in Silicon Valley, uh, many people as more of, I think a status object.

SPEAKER_124: Have chicken coops. Yeah. Is that popular in the East Bay as well?

SPEAKER_363: Yeah. Yeah. Chickens are big. Totally. Backyard chickens. It's a thing. NPR chickens.

SPEAKER_422: Backyard.

SPEAKER_423: Oh, NPR chickens is hilarious. It's hilarious. Like a virtue signaling chicken. That's so funny.

SPEAKER_124: Um, but the, the interesting thing about the virtue signaling chickens is we have a neighbor. Who has a coop and they can't keep up with their egg consumption. So every couple of months they will literally ring our bell and like their kids will just drop off two dozen eggs. Uh, because, uh, and I,

SPEAKER_427: Decentralization is the future.

SPEAKER_318: I think it's a very cool idea is to have. I would have chickens if I could. I have an HOA. Oh.

SPEAKER_08: Yeah. Wah wah.

SPEAKER_341: Chickens. Actually, that would be a really cool idea too, is if you do have an HOA and you do have 30, 40 homes.

Chamath Palihapitiya: In your homeowners association to put a, a common chicken coop and let people have access to it.

SPEAKER_430: And then. Community chickens. Community chickens.

Chamath Palihapitiya: Oh my God.

SPEAKER_293: I love it. Uh, well, community backed NPR chickens.

Jason Calacanis: Yeah. Right. Community backed NPR chickens. Exactly. Everybody has a mug and a tote bag and you carry your chicken feed in your tote bag.

SPEAKER_341: The problem with this startup is I know they're going for subscriptions here is if this works, somebody in China.

SPEAKER_01: It's going to just copy it and knock it off and make a better version and just sell you a $50. Amazing. Compost grinding.

SPEAKER_436: And you're just going to throw it into your compost and you're not going to go through sending it back for chicken feed or whatever.

Jason Calacanis: I mean, I gotta say this is the thing they need to nail in order to beat that coming competition is the chicken feed part. Cause that's a huge revenue stream. If they can actually get through all the hurdles to turn this into legit chicken feed or any other kind, you know, if they can make it an agricultural feedstock, then people will pay them to come and get it. But yeah, if it's just the trash can grinding up my food and compacting it, hell yeah, I would pay a lot of money for that. But like, probably not every month. I bet you there's one that exists already.

SPEAKER_437: I bet you there's a trash can you can buy right now that does this. Brutal.

SPEAKER_11: Matt, we're going to need you to come on the show and explain to us why this is going to be so awesome and why you can't be beat. Why you can't be beat.

SPEAKER_442: Why you can't be beat.

SPEAKER_11: Yeah.

SPEAKER_322: Well done. Is there another startup of the day?

Jason Calacanis: We have another startup of the day. Actually also literally accidentally in the sustainability space because bubble. Um, but the other one, and I'm just amazed by this is solemn technologies. S O L L U M. Like the sun. Solem is making movie moves in the greenhouse produce space. This is a like growing. This also is a growing part of the climate conversation because as like areas warm or are constantly flooded or are constantly in drought. Or, you know, so hot, you can't grow stuff. Figuring out alternative sources for agriculture is super important.

SPEAKER_05: And so this company solemn technologies just raised $30 million, including a $25 million check from a single private equity firm idealist capital that supports entrepreneurs going after energy transition and decarbonization. And I love this because it's basically grow lights. It's grow house lights, but apparently they've figured out some technology that makes the lights that they can just like remotely mimic perfectly the sun. Like they're powered by AI to reproduce the full spectrum of the sun's natural light.

SPEAKER_08: They have sun as a service, which unfortunately spells sun ass. Yeah. Yeah.

SPEAKER_445: So that was kind of a, maybe a naming miss there.

SPEAKER_05: Um, but it's a cloud platform powered by AI. This is like, this is actually merging the two bubbles with a dedicated support team of engineers, technicians, and agronomists. And then this support that includes installation maintenance, and they claim their solution saves 40% on energy produces better taste. They've been able to like, apparently increase productivity of certain crops by 40%, like strawberries, which are, you know, obviously not a year round crop in various places and can get super expensive and are hard to grow.

SPEAKER_08: And it's like modern green housing.

Chamath Palihapitiya: Um, fascinating, uh, using AI to make the lights more efficient. You know, that's something they do in solar, right? I think you had somebody on who was making technology to move the solar panels to capture the sun, right? And so this is kind of the reverse of that, you know, using software to optimize the sun on the plants.

SPEAKER_01: Right. And then eventually this will become full cycle. You'll be figuring out through AI and machine learning over time. So this is where the reinforcement learning, um, that I think Shamath was talking about a whole bunch on the last all in the, the, the learning part of this.

Chamath Palihapitiya: Well, what if you could take what the lights are doing with the eventual, uh, bounty? So, Hey, we went for it and we increased the lighting 20%. Okay. Did that increase? Was that energy worth it? And it's like, actually it turns out, eh, we got 3%, you know, more robust strawberries. This crop had 3% more strawberries and they were 1% better.

SPEAKER_124: Uh, but they didn't taste as good or, uh, it wasn't worth the, the squeeze.

SPEAKER_83: Wasn't worth, uh, the juice. Right. And I, I just love that expression. The squeeze is the squeeze worth the juice here.

SPEAKER_124: Uh, and I don't know if it is, you know, and that's what AI is going to teach us is optimization. Yeah.

SPEAKER_453: Whether it's for our bodies. Is the juice worth the squeeze? Oh, what did I say? Is the juice worth the juice?

SPEAKER_125: He says the squeeze worth the juice. I'm sick guys. So I, my voice may not be hard, uh, but my logic is broken. I like it.

SPEAKER_68: I like it. Cause this, I mean, either way it works. Cause the squeeze is like a lot of work and it's like, is the juice really worth it?

SPEAKER_457: Is the juice worth the squeeze? Is what we're talking about here.

Chamath Palihapitiya: And so, you know, that that's going to be awesome. And it might be some crops, each crop might be different. You know, if you're doing like a strain of cannabis, which is what these lighting systems, the reason a lot of investment went into these lighting systems was for crops that were worth a lot of money. Yeah. And so whenever we saw this technology, smart, any kind of smart growth technology, it was for the most expensive things. The two most expensive things I've seen are berries and weed.

SPEAKER_122: Yeah. I don't know which crop is actually more expensive.

SPEAKER_459: I don't know either, but that's so funny. Cause yes.

SPEAKER_122: Exactly. It might be. It's gonna be your girl scout cookies. I don't know.

Chamath Palihapitiya: It could be your OG Kush, whatever, uh, could be the more important one, but you know, it's interesting. We were talking about that bin, you know, Vitamix. Yeah. You know, Vitamix.

SPEAKER_249: Like to me, I have a Vitamix. Yeah, the juicer or whatever. The juicer. Yeah. The world's best blender. Yeah. The world's best blender. It's like every chef uses a Vitamix. It costs 300, 400 bucks.

Chamath Palihapitiya: I've had the same one for over a decade. They're hella nice. Smoothies in it, whatever. Check this out. That's Vitamix. Vitamix has a product called the food cycler. What's a food cycler you ask? Well, if you scroll down to the videos. Um, and you just press play on the video and you fast forward it just a wee bit. What you'll see is everybody in your household goes to the food cycler. And when you're scraping your plates, instead of putting it, uh, down there, uh, you will see, uh, they put a bunch of stuff in there.

SPEAKER_124: And once they dumped the stuff in there, it turns on. And what does it do? It turns it into. It's just like the garbage.

SPEAKER_125: Uh, sorry to be. Okay.

SPEAKER_468: To be fair.

SPEAKER_470: Thermostats existed when nest came out. Sure. And what happened. Nest.

SPEAKER_472: Yeah. Better form.

SPEAKER_98: Right. Better form factor, more functionality. Mm-hmm. Like you can still build a better mousetrap. And presumably that's what this company.

SPEAKER_05: However, I'm totally buying this Vitamix composter thing. This is amazing.

SPEAKER_341: I'm looking at it right now. I want this. Yes.

Chamath Palihapitiya: Yeah. I I'm looking when I do the Amazon reviews, there are, uh, if you go down to the bottom, uh, instead of doing the one from Vitamix, if you scroll down, uh, there's a review and demo of it from like a consumer.

SPEAKER_480: And they, they put like paper towels in it. If you go down to the bottom. Yeah. The first one on the left there.

SPEAKER_83: Yeah. DIY life.

Chamath Palihapitiya: I put the output quite nicely. Um, it's a good looking machine, but what it does is it actually, um, part of this process is heating it apparently. Mm-hmm. So you dry it, I guess, to get the moisture out. So you're not just making a slop. So you dry it and look at that.

SPEAKER_486: Oh my gosh, you can put your coffee grounds in there. Yeah, yeah, yeah.

Chamath Palihapitiya: But I guess because it dries it and look at you're putting paper towels in it as well. Yeah. So she's, uh, wiping down the, uh, table here and I'll describe for people watching. So she just put paper towels in there, eggshells, coffee grinds, the whole bit. And then when you look at what the output looks like, again, it looks like sawdust. And then to top off the consumer review here, they take it. And I guess the heat's the key part. Cause I guess hot air is blowing out of it. Uh, she takes it and dumps it into her pots to make, I guess, fertilizer out. Yeah. Pretty dope.

SPEAKER_492: Dude. I don't want that now.

SPEAKER_493: I'm gonna buy this.

SPEAKER_492: Sorry. I just cost you 200 bucks for watching the show.

SPEAKER_493: $322. That's expensive.

SPEAKER_164: That's nuts. Yeah.

SPEAKER_494: This should be a $99 version of this.

SPEAKER_164: Yeah.

SPEAKER_276: But that's the same as the, um, the one, the new one that doesn't make. Yeah. One year of the new one. But without a monthly fee, that's one year of the new one, but no monthly fee.

Chamath Palihapitiya: I was assuming the new one didn't have a price up front. So this is one year of the new one, I guess. Dang. Um, so this is an existing product category that we were not aware of. Yeah. Uh, this is exciting. Yeah. I think the key is the key part.

SPEAKER_276: I could get this for just $26.83 a month for 12 months, interest free.

SPEAKER_249: Oh yeah. Are you getting over your skis, Molly? Don't please. Buy now, compost later. Buy now, compost later. No, I will only buy it. Compost now, pay later.

SPEAKER_12: Compost now, pay later. Heyo. Squeeze that juice. Damn, that's cool. Well, congratulations to both of these hardware companies. Yes.

SPEAKER_500: Um, and let's save the planet. Let's do it. Let's do this.

SPEAKER_501: I'm kind of big on the waste thing.

SPEAKER_01: You know, I think, I don't know that this is a great investment category for venture capitalists, but I think it's a great category for humanity.

Chamath Palihapitiya: I've been so obsessed over the last year or two with austerity measures and waste. Mm hmm. And really just the less waste we have per household is just so significant. That's why I think making the cold brew coffee instead of me ordering all of those packaging. I really want to find a great sparkling water. This is the one I haven't solved. The sparkling water I have from what's the one everybody uses. We push the top down. Soda stream. Soda stream. Everybody hates soda stream. It doesn't come out good. It's too much work. Yeah. I want a tap. A soda stream tap. Like I have a hot water tap. It has to be convenient. So I want to be able to have just like when you would go to an ice cream shop in the day and they would just pull the, you know, sparkling water from a tap. I want to tap sparkling water. This exists.

SPEAKER_506: This definitely exists. I know I looked this up.

SPEAKER_83: It definitely exists, but I think it's very expensive. I see it at restaurants.

Chamath Palihapitiya: Like I just want it in my tap and then under my sink, have a giant CO2 thing. And I see hackers are making this, um, grow themselves.

SPEAKER_344: So you can buy, you know, but there's no. Yeah.

Jason Calacanis: Look, I got a whole, I got a whole situation that the DIY guy made. That's like a, it's like a brewer. It's like a huge stainless steel canister filled with carbon dioxide with a, yeah.

SPEAKER_164: And I mean, it works, but there's a company called grow. G R O H E. Is it German company?

SPEAKER_510: And they make water filter kitchen taps that include carbonation.

SPEAKER_164: Ah, that's exactly what I'm looking for upfront, but it's exactly what you're looking for.

SPEAKER_514: I want it filtered and I want it to have carbonation. Fantastic. Look at that.

SPEAKER_32: It's good looking too. See, I think you have to make this very easy for people.

Chamath Palihapitiya: If you want people to do this stuff, it has to be easier than ordering from Instacart or Amazon.

SPEAKER_25: And this is definitely easier. It's just, um, so it starts at 2,900 bucks.

SPEAKER_519: So it takes you a lot of LaCroix orders to recoup your money.

SPEAKER_386: Yeah. That sounds like it's going to take years. Yeah.

SPEAKER_521: If we're spending 200 a month or a hundred, 200 a month on cans. Can you put a price on the planet?

SPEAKER_403: Yeah. Maybe it's not. A hundred bucks a month on cans. If this cut it in half, 50 bucks a month, you would need to have 60 months to pay for it or something. That's not terrible.

SPEAKER_25: Yeah. It's like a lot of things where you still have to buy carbon dioxide. Like you have to buy the tanks and stuff. I would assume.

SPEAKER_315: That's why I put it at half. Like savings.

SPEAKER_25: Oh, half. Sorry.

SPEAKER_125: Yes. If we put it on cans, this saved us $50 a month on cans or something for the equivalent with the, to maintain it. Maybe it's worth it.

SPEAKER_98: And don't forget the value of smug is infinite. Yeah.

SPEAKER_60: This is a, yeah. The, this is the NPR carbonator.

SPEAKER_531: Yes.

SPEAKER_25: I mean, I came up with a new one. Somebody tweeted the other day and I was like, thank you for saying this. Cause I think this all the time. There was a tweet that was like, I don't really understand why virtue signaling is like a bad thing.

SPEAKER_68: Like every one of us has virtues. Why wouldn't we signal them? Absolutely. Yeah.

SPEAKER_25: I don't really, you know, it's sort of like, yeah, actually a good point. Yeah.

SPEAKER_399: It's when you're fake, it's when you're lying about your virtues. Yes. But like when it's green, just hypocrisy, we already have a word for that.

Chamath Palihapitiya: Yeah. I mean, if, if I was do, if I was using, you know, the food cycler, but then I had my thermostat on 82 in the winter and 56 in the summer. I had this new idea. Uh, I came up with a new system cause I was hacking my nest. And, uh, it's called put on a sweater.

SPEAKER_75: It's a new plugin. And, uh, Is it an app? Hold on. Put on a sweater.com.

SPEAKER_249: No, it's just me yelling at the rest of my family who are walking around in shorts in the winter and a t-shirt or a tank top and they're putting the heat on 78. And then I have to tell the, uh, four ladies I live with put on a sweater. Oh my God. Or if you wear socks and pants, then maybe we don't have to keep the heat at 78.

SPEAKER_545: That is one.

SPEAKER_05: So one weird thing about the Bay area that people probably don't realize. And that when I moved here, I was like, why am I freezing all the time in my new town that I live in? Like always cold.

SPEAKER_98: It's that like no one as a matter of just cultural competency. And I don't even know where it came from. The heat never goes above 68. Like I am. I almost innately believe that my thermostat is not capable of going above 68.

SPEAKER_548: I'm in a Bikram yoga class here. I know. You really are. Like ladies. No wonder you're sick. Get dressed. Here's an idea. Get dressed.

SPEAKER_249: It's not August. You're, you're wearing the equivalent of a bathing suit. Put on pants. Hilarious. Wear socks.

Chamath Palihapitiya: A t-shirt and a sweater. Uh, uh, you know, I'm not saying you have to wear a jacket in the house, but you could

SPEAKER_341: put a sweater on and we could keep the heat at 70, you know, 66 or 68. Hoodies are in hoodies are the look right now. Come on people. All right. All right. What a great show. And we'll see you all tomorrow.

SPEAKER_555: Bye bye. Bye bye.