This Week in Startups - Zuck's AI manifesto is a data center PR masterclass | E2323

Starting point is 00:00:00 There is a PR group for META that says, how can we turn the narrative on data centers around? Because we need to build more data centers. We need to catch up here. When you're behind, you go open. When you're ahead, you go closed. I'm in your corner, and we're going to be the company that gives you an agent, a tutor, free design tools,

Starting point is 00:00:19 and make your life better. The only thing we ask is that you let us build some data centers. This is the best version of Zuckerberg I've seen in his career. And he is thinking about the ramification. of what he builds for the first time in the history of the company. revenues are at least in the seven figures, go to net suite.a.ai slash twist. Built for every industry, ready for every boardroom, net suite.com.a.i slash twist. And every. For all of your incorporation, banking, payroll, benefits, accounting, taxes, or other back office administration needs,

Starting point is 00:01:18 visit every.com. All right, everybody, welcome back to this week in startups. It's our news program today. We got a lot of news, lawn. We do. Including Mark Zuckerberg's essay. This thing is long. I went on a ruck this morning and I listened to it. 6,500 words. 6,500 words in Zucks. He writes nine more. He's got a book.

Starting point is 00:01:39 So a book is typically 60,000 words. So if he does nine more essays, you could have Zuckerberg's essays. And he's getting pretty active on X, which he never did. His handle is FinkD, if you don't know that, at F-I-N-K-D, which was his online handle when he was in college. Yes, he has not ever changed his handle online. So, yeah, the essay is titled, The Future is for Everyone, the Path to a Positive AI Future. Here is arguments in a nutshell.

Starting point is 00:02:06 Basically, superintelligence's biggest risk is centralization, not the tech itself. He expresses three core principles, individual empowerment, AI's main purpose is invention, not automation. And there should be a balance of power and governing access to information and what we're doing with all of this intelligence. He also says, here's what meta's building to sort of be part of this AI future. That includes a personal AI agent for everyone, powerful coding video design and more tools for everyone, ways to launch businesses without teams or funding, a pH level tutor or coach for any subject, contributions to science, and so forth. He also discusses some of the risks, job loss, data center pro pushback, cybersecurity and bio-weapons, the U.S. not keeping pace with China. And he makes a few policy commitments, Jason. Meadow will continue releasing open source models.

Starting point is 00:03:01 They today drop the open weight model Muse Glimmer, and they're making plans for an open weight version of Muse Spark 1.2. He also proposes independent boards to review model safety criteria, and he calls for closer cooperation between labs and the government that doesn't slow down the release of new models. So that's kind of a quick overview. So what do you think? I mean, I think most people, is this PR? Is this advertising for the next wave of meta-AI?

Starting point is 00:03:29 Or do you take this as a sincere thing that Zuck is worried about the future and wanted to be part of the conversation? There's a lot going on here. I think he's learned a lot being a technologist who moved fast and broke things and got himself into hot water. And one of the things he learned is he's historically a terrible communicator. He's unlikable because he's rich. He maybe, you know, I don't want to diagnose anybody, but maybe doesn't make eye contact or doesn't speak in a way that is engaging to people. And so I think he understands those faults.

Starting point is 00:04:06 And then also, I think because he was such a young founder, he was getting spun by different people telling him how to run his business and how to communicate. So you take those two combinations together, not being a great natural communicator, but great at building a business. obviously. And then people spinning you around that I think he's finally saying, I'm going to control the message here. I'm going to state my actual beliefs in a way like Elon does and let the chips fall where they may. I'm going to go direct, right? And so now we're seeing him get vocal and go direct because he just paid this huge fine, right? And the entirety of the second movie that's

Starting point is 00:04:47 coming out about him, you put all that together as just a little prologue. to this, I think he now wants to communicate directly. As a man, as a founder, you know, founder authority, I'm going to control the message. I'm going to actually think from first principles what I believe and I'm going to state it. And I'm going to be more thoughtful and preempt things that typically I would just build products, not explain myself, and then when they break things, I'll just pretend that didn't happen and not say anything and then get pulled in front of Congress, right? In corporate communications, I think he's made a lot of mistakes over the years. I think he would agree with that, too. So here we have a manifesto of AI. And he touches everything.

Starting point is 00:05:37 And I thought it was actually very well written and very thoughtful. But he covers everything. This is not one position. This is literally like a dozen. So you have to break down each one. So I think one place to start here is he's for open models, and he's not for regulating open models. He believes that the best way to deal with these frontier models, with dangerous things, is not to lock them up or regulate them. It's to give everybody access to them. Sounds familiar. It's literally the manifesto of Open AI. Open AI's manifesto was.

Starting point is 00:06:15 this technology is too powerful for one person to control it. So he's taking that position. That's also extremely convenient, isn't it, since he's behind? When you're behind, you go open. When you're ahead, you go closed. The Facebook network, Instagram, those are black boxes. You can't index them. There's no API for them.

Starting point is 00:06:34 There's no grabbing the social graph and doing interesting things with it as the developer. They dabbled in that for a couple of years and shut it all down. So when you're ahead and you've got the franchise, eyes, you lock it down, iPhone as an example, when you're behind, Android as an example, you open up. Okay, so he's for opening, and that's a really good strategic move for him. So then he talks about the dangers, and there's a part where he talks about chemicals and that kind of stuff. And I think I screenshot it or you can just do a quick control find when he talks about materials. And there's a section where he talks about materials and the

Starting point is 00:07:09 dangers of them. So essentially, he says, listen, trying to regulate the models is not going to be as effectively as trying to limit access to chemicals and precursors. So we should do that. In other words, how you make a fertilizer bomb, God forbid, like Oklahoma City, we are going to have a hard time making an open source model not to be able to do that. We'll have an easier time going after the actual cell of fertilizer. He doesn't, I think, say it that crisply, but let's hear it. Yeah, he says, lab should continue working to reduce biological and chemical.

Starting point is 00:07:45 risks in our models. First, we should focus on limiting the physical production and distribution of harmful materials. I expect it will be easier to regulate and control physical components than the spread of knowledge. So this is an important area of policy focus. So this is really important. This is the first time I've heard this idea. I think it would be the first time you've heard it. We are bringing up the dangers of AI. Yes. What does Dario say when we talk about the dangers of AI? Oh, my God, this stuff is really dangerous. We had to stop it. We can't release it. He's getting involved in drama with the administration. What does Zuck say? He actually bought this through. You're not going to stop information. Information wants to be free,

Starting point is 00:08:27 and AI makes it impossible to hold back this information. You're not going to stop access to open source models coming out of China or other countries or America or people making their own now. Cats out of the bag. You can make your own model. So if we are concerned about biological weapons, the materials and the tools to make those, he didn't mention tools, but I'll add that, are where we should try to limit things. If we are actually concerned about that, that's actually a really good policy statement.

Starting point is 00:08:56 I haven't heard from another person. I feel like we already do, I mean, unless people are building bombs or whatever out of everyday common household items, we do try to already make, like, I couldn't go buy like a bunch of different hazardous chemicals today, at least not without some sort of a license or a permit. You would be surprised.

Starting point is 00:09:19 You can buy a lot of the precursors. No founder I've ever met started a company because they love doing paperwork. No, you don't want to spend hours filling out payroll and setting up an HR department. And I know you don't have a passion for making your taxes more complicated. Do you? You want to fix problems. You want to create products in the world. All this paperwork, it is a huge distraction.

Starting point is 00:09:39 You can move all that off your plate with a trusted partner every, Every is the all in one platform for incorporation, banking, payroll, benefits, and taxes. Basically, your entire back office. They're going to keep up with all the bookkeeping and paperwork so you can stay laser-focused on product market fit, scaling your company, hiring great team members, all that stuff you love doing. And if you're just getting started, they'll handle your Delaware C-Corp and getting yourself a registered agent, all with no legal fees and no delays.

Starting point is 00:10:11 And every's worked with over a thousand startups already. That's the kind of know-how and experience you can rely on. So head over to every.io and stop wasting time on the stuff that isn't growing your company. That's e-V-E-R-Y. Just like the precursors to making methamphetamine made famous on breaking bad. Right. Well, that's my point. Like that's pseudafed.

Starting point is 00:10:34 Like you can distill that like that's what they're doing on Breaking Bad. Like you could go buy a bunch of Sudafet. That's what we can't seem to sort of limit. So I'm wondering. actually, we can do a much better job limiting the precursors. Okay. That's absolutely the case. And so if we do think that this is going to get out of hand, it's an inconvenience.

Starting point is 00:10:54 It's bad for those businesses. If you put up friction and you require people to show their driver's license and you limit the number they can order or they got to fill out a form and coming back 48 hours later, that'll decrease the sales of precursors, whether it's for drugs, bombs, weapons, etc. But it is a much better solution. And by the way, fertilizer went through the same exact handwringing. Before Oklahoma City, the bombing that occurred there, which was a giant fertilizer bomb,

Starting point is 00:11:21 a UPS kind of style truck, you could just buy fertilizer. And then after that, they started tagging fertilizer. They started monitoring the people purchasing. They put all these things in. So this is something I'll give him credit for. He's not just saying this stuff is terrible. He's saying there's easy solutions to this. Then he makes another point about cyber.

Starting point is 00:11:41 And if you look at his cyber section and hacking, he's basically like, you know what, AI applied to hacking, and you can pull up the quote there from it, is we're going to be living very quickly in a world where the code is made rock solid by AI. So this is a transient issue. In other words, he believes the ability to hack systems is going to go down greatly because of these tools. therefore the two things that people are very concerned about, hacking and cyber hacking combined with biological weapons, he's just saying he's taking him head on. Whether you agree with his position or not,

Starting point is 00:12:24 you might disagree with it. My point is he's giving thoughtful feedback on very granular issues and saying, I don't think you need to worry about it. And this is what I think the right approach should be. This is a very different approach that he's taken. And he's doing this ahead of time. So I don't know if you have that quote. I do. Over time, I expect that widely deployed AI models with strong cybersecurity capabilities will lead to systems that are more secure or not less. This will be definitively true once superintelligence enables most of the world's code to be verifiably secure. The long-term answer isn't to withhold capabilities, but to establish a balance of power where superintelligence is broadly distributed. My point here is, now Zuckerberg is anticipating

Starting point is 00:13:08 problems, taking those problems, thinking from first principles about possible solutions. This would have been amazing before he launched Instagram, made it super addicting, and caused young girls on the platform to experience body dysphoria and self-harm and anorexia and any number of diseases. And if he had a thoughtful approach to, hey, age-gating and what age people should have. This is the big paradigm shift I see in Zuckerberg, just by one example. He's now anticipating problems and taking a leadership position. I'm saying here's how the world should work. Does that make sense, Lon? It does, yeah. Probably the best insight he had in this whole thing was about AI and jobs. He admits, hey, there could be job loss. This is something to worry about.

Starting point is 00:14:00 But if we withhold the technology and we don't make it available to everybody, that's worse, then giving it to everybody for free, for eternity, unlimited, et cetera, giving access to it because he believes the job displacement will occur. But if people are learning these tools and sharpening their blade and learning how to apply them, they'll stay ahead of the job loss. Right. which is exactly what Jensen said in a quip that became very famous, you're not going to be replaced by AI,

Starting point is 00:14:34 you're going to be replaced by somebody using AI. Let's read how he phrases it. I think it was very well written. Here's the Zuck quote. Recent statistics suggest it may be more likely that individuals' capability growth could match or outpace automation, in which case people will gain the ability to do many new things before their current jobs change.

Starting point is 00:14:56 Right. So if you think about that just in our firm, we've been deploying AI. I've been talking about, hey, who are the AI natives inside of our venture firm? Some people are taking to it, like fish to water. Other people are like, you know, don't want to dip their toe. Some people are up to their ankles. Some people are wading in, right? And the quicker you adopt it, the more value you become, which then the organization finds

Starting point is 00:15:22 more opportunities and it grows. So that rang true to me. And I thought it was a really good framing for. the job loss argument and job displacement. Yes, jobs will go away, but can you stay ahead of that? He also telegraphed new products coming, which I thought was really interesting. Now, he doesn't have a problem launching a lot of products, and he's had a ton of failed products. People forget poke. That was his Snapchat clone, 2012. Slingshot was another disappearing message, 2014, Team. Life stage. This is a video app named at teens to, you know, go after Snapchat. And Bolt, that was a Snapchat style messaging app that they tested in some like foreign markets. Those were four different ways he tried to attack Snapchat. Then he just put all Snapchat's features into Instagram. He also did Facebook Home, if you remember that.

Starting point is 00:16:15 Sure. You remember he had also the Metaverse. Frame for photos. There's that whole Metaverse. Then there was Facebook gifting. There was Facebook. Gifting. There was. Facebook credits for virtual games. It was a long line of things. But he mentions here what they're building. Yes. And we should go through each one of those. The first one, a personal AI agent for everyone.

Starting point is 00:16:34 I have a little quote here. Everyone will have an exceptionally capable personal agent that understands you, your goals, and everything you care about. Your agent will work 24-7 on your behalf to improve your relationships, health, career, finances, home management, hobbies, and more. It's going to free up the time, is what he's saying. I predict Facebook is going to launch an open claw. Erme's agent, Claude Co-Work competitor.

Starting point is 00:17:00 It's going to be, you know, Facebook, you know, assistant or Instagram assistant, whatever he brands it as. One of those brands would probably be smarter. And what it's going to do is study your social media activity. If it sees that you are booking tickets or you're liking content about a certain movie in IMAX that everybody's going to see right now, the Odyssey, it's going to say, would you like me to find your tickets to the Odyssey in IMAX,

Starting point is 00:17:32 and I can go do that for you? And then it's going to go do it for you. It's going to see, hey, your birthday's coming up. Last year you did a birthday party. Do you want me to book a reservation for you? So it has all of that context on your life. They are in a really, really great position to take the skunk works,

Starting point is 00:17:51 patchwork of personal agents that were all building in the technology community and give those to your auntie. You know, your aunt who was never on MySpace or Friendster or, you know, other wonky social networks, but they were on Facebook and they got addicted to it because it was so simple. They could do that for your auntie, for your cousin who's addicted to Instagram, whatever it is. So that's one product that I think could be very successful then. I mean, I'm sort of.

Starting point is 00:18:21 very divided here because on one hand, it does sound very useful, like an agent that's built into so many of the apps you've already got running on your phone, that's intelligent about all that, that you don't need to connect to any of them. I think people could get a lot of use of that, but there's so much resistance among the consumer app audience for AI. I feel like you've got, that's a big thing to overcome. He's going to have to really sell people on the utility here so that they're not like, oh, more slop, get it off of my phone. So he doesn't have to introduce it as AI at all. It can just organically exist in your feed. And I think that's what will make this so compelling. You won't have to say, I want to install AI. You will just

Starting point is 00:19:04 very subtly start prompting you. Hey, I see you have 17 restaurants that you've bookmarked here. Would you like me to, I put them into a list for you of restaurants in Japan. Yes. As an example, I do use a lot of bookmarks in Instagram. So when I go to Japan or I go to Paris, I have already bookmarked a bunch of restaurants, a bunch of shopping, a bunch of hotels, whatever it is. It could very easily close that gap for me and do intelligent agent-like things without me asking.

Starting point is 00:19:37 In fact, travel and food and fashion are things that it already knows a lot about you. So I could just start saying, hey, I'm your Instagram agent, wanted to let you know that I've kept a catalog of all of the men's clothing that you've liked, Lon, here they are. And I've monitored the prices on them and where you can get them. And here are places that make a similar item.

Starting point is 00:20:02 Those would be very subtle. That's very astute. I think that if you, and I think that's actually what a lot of these companies should have been doing all along with these AI. You don't need to brand everything AI. You just make it really useful. and then people will start using it regardless of how they feel about the technology. I think that makes a lot of sense.

Starting point is 00:20:22 Founders using AI to be more efficient are pulling away from everyone else in the pack. If you apply AI in your company, you're going to go faster and you're going to beat your competitors. So if you're wondering how to keep up with your admin and number crunching tasks, without hiring a huge team, you need NetSuite next. You already know NetSuite. They're your single source of truth combining financials. inventory, HR, CRM, and every other system in your company. And they're already trusted by more than 43,000 companies. NetSuite next is the same amazing platform with AI running alongside everything.

Starting point is 00:21:01 It's your guide on the side. Just type your question in plain language into Ask Oracle, and you'll get a real answer pulled from your company's live data. Plus, they're flagging potential problems before you even think to ask about them. Now, for the first time ever, You can try NetSuite Next for free. If your revenues are at least in the seven figures, go to netsuite. comaI slash twist. It's built for every industry, ready for every boardroom, netsuite. com slash twist.

Starting point is 00:21:30 He's also building what he calls powerful coding video and design tools for everyone. So not keeping these gated, you know, like these cursors just for coders or Higgs fields just for the industry. He wants to open that kind of stuff up so everybody can have access. That's the next one. Okay. So my next prediction is Instagram will launch a video editor, similar, an AI-based video editor, similar to the way they added filters and, you know, swiping through filters. Yeah.

Starting point is 00:22:04 That was something you did in Photoshop before. I'm subtracted here. That's going to be a really interesting design feature that we can use. And there's no reason that can't be like Capcut. And I was going to say Cap Cutts the model. So cap cuts the model. We're already three quarters of the way there with Cap Cut, I think. Based on what you've queued up there, I'm convinced they will just have their own Capcut competitor.

Starting point is 00:22:31 They can build software really quick. AI-based software that can make images and video. Sounds exactly like Cap-cut. I think they do have some tools built into these things. But I would look for another standalone item like this. that coding, I don't think the world needs them to do coding. I don't understand the opportunity there. So I'll say coding.

Starting point is 00:22:51 I don't think he's going to do anything. I mean, we already have codex, Claude Co. Or, excuse me, chat GPT work and Claude Co-work. And to me, those are already the like, hey, you're not really a coder, but here's like a coding style or lovable. I mean, I feel like that already exists in the world. Exists. Yeah. I don't think they're going to compete there.

Starting point is 00:23:08 All right. What's the third way? Ways to launch businesses without teams or funding. Okay. this is very interesting. Zuckerberg is kind of teeing off of WhatsApp. So this is WhatsApp has business accounts. Those business accounts allow you to communicate with customers

Starting point is 00:23:26 and then to use the ad networks to find customers. If you go overseas, WhatsApp is how all transactions occur. When I go to Europe, when I go to Asia, if you're talking to your driver or you're talking to your cab company, If you're talking to your hotel, if you're talking to a tourist guide, you're taking a guided tour somewhere or you're talking to a restaurant, it all winds up on WhatsApp and you build a relationship there. I never had to have WhatsApp all my time in the U.S. I never signed up the moment, the day I got to Japan I had to join because everybody, everything in Japan is happening on WhatsApp. I was amazed. Same in Europe. So if you go to Europe, it's the same exact thing. Every single person is using that. So if you were starting a tour guide business, if you had a tour guide business as an example,

Starting point is 00:24:16 or you're a driver and you provide livery services, either of those, you build a landing page, you have a Google, local, and then you use WhatsApp for your operating system. Why not make it so WhatsApp has, in their WhatsApp business, the ability to make a small business page to handle customer support and do all that with AI. In other words, just take WhatsApp and build adjacencies to it. And what you're seeing here from when I explain all of the interesting things that he telegraphs in this manifesto, each of their key properties, he's building something aligned right next to it that builds an adjacency. One of the things we learned about AI tools and software is the pace of software is so fast now that building an adjacent feature is faster than ever.

Starting point is 00:25:08 Instead of it being a three to six month sprint to build, let's say, a homepage builder, right? Like a Wix business homepage builder or HubSpot to manage your CRM. There's no reason that Suck can't build, you know, an 80% version of that into WhatsApp in three to six months. So that's what he's doing. It's always the adjacencies, adjacencies, adjacenties, adjacenties, adjacent. And I think it's pretty clear that WhatsApp is the home for the one he just mentioned there. Yeah, for sure. So other things he says he's building a PhD level tutor or coach that will help you out in any

Starting point is 00:25:43 subject. He wants to make some contributions to science, sort of like deep mind. Like he wants to, and then free or low cost access for everyone is the big. Yeah. Now, this is the big win. Yeah. The big one is at the end there. if your margins by opportunity, he's already got the second largest ad network in the world behind Google.

Starting point is 00:26:05 It already prints money for him. Every time you use Facebook to do a search, it's learning about you. Just like Google. So every time you engage with Google, it knows more about you. It can target an ad better. Because they're using this AI to target ads to you, if they just get you to use their service, they will learn 10 times more about you. And if they learn even 50% more about you, their ad network's going to get dramatically more valuable in the world to marketers. So an ad version of Facebook

Starting point is 00:26:44 GBT, Instagram GPT, and what's at GPT? They should literally just put GPT at the end if it wouldn't get sued. It wouldn't get a letter. Trademark, yeah. But they could just put Facebook AI, Instagram AI, WhatsApp AI, and allow you to do searches there. And when you do Instagram now, it does default when you do a search there to a little AI clip at the top, just like the shortcuts there. So they're already doing this. And they don't even have to present an ad in that answer. And Metis always wanted to build a search engine. I think a Facebook search engine would be a killer idea. They should buy perplexity or they should just build their own search engine. And when you do a search, it could just be like a Google search. They just have to come up with a good name for it.

Starting point is 00:27:30 Or just buy Duck, dot, go and make it part of the brand family. This would be a huge innovation for them. All of this backs into the final argument. And what this really is about and the timing of this. And that's what one has to wonder. Why now? Why did Zuckerberg write this now? If you're a cynic or you're just speculating, you'd say, well, more CEOs are getting out. And you'd say, well, more CEOs are getting active communicating directly. Right. Jensen. We just joined Twitter.

Starting point is 00:28:00 Open letter that Jensen did about the power of open source. Exactly. Yeah. So if you believe in the go direct moment, you had Palo Alto Networks, CEO and the CEO of Nvidia, just joined Twitter and start writing manifestos and, you know, daily weekly updates.

Starting point is 00:28:23 And it's gone over really well. So Zuckerberg could just be following them. That would be a very basic interpretation of what's happening here. Well, there's just this big conversation going on with tech leadership and government. And maybe he just feels like he wants to be part of it, would be the theory. I don't believe that that's actually what's going on here. Right. I think there's something very pragmatic going on here, which is data centers.

Starting point is 00:28:49 Correct. Yeah. And data centers are really unpopular. And so if you look at this manifesto, everything about it is abundance. This is Zuckerberg's abundance. I am on your side. I'm going to empower you. I'm going to make it free manifesto.

Starting point is 00:29:07 If you were to sum up all of the points in this memo, it's that your life is going to be better from AI. I'm in your corner. And we're going to be the company that gives you an agent, a tutor, free design tools, and make your life better. and we're going to even teach you how to make a small business. The only thing we ask in return is that you let us build some data centers.

Starting point is 00:29:31 The team is pumped because you're about to close a massive deal. But then the client's lawyers get involved. What happens if you get hacked? How are we going to protect your data? There's no need to panic. This is why you brought in Y security. Y security is staffed with over 40 experienced engineers who have actually worked security for world-class companies

Starting point is 00:29:49 like Apple, Uber, Microsoft, Robin Hood, Brex, and so many others. But the best part is you don't even need to hire Y security. Your company can rent Y Security's elite team by the hour. That means no massive salaries to pay, no costly consultants, just real experts embedded in your company, helping you out with your SOC 2, ISO 4200, or any security or compliance challenge you're facing. You can even set a monthly cap so you know exactly how much you're spending. And your first six hours are completely free. Head to YSsecurity. dot i.o slash twist and book your free six-hour strategy call. That's y-security.io slash t-wis-t. At the same time this was happening, this morning, I was listening to Morning Joe. In the morning,

Starting point is 00:30:39 I will switch back and forth while I'm having my coffee, CNBC, CNN, Fox, MSN, MS now, whatever it's called. And I'll just zip through a couple of these things, just to see what they're talking about, because there's a really cool feature on YouTube, The Quad, where it gives you all four and you do to do. So I see this piece on Morning Joe about how great Zuckerberg's data center in Utah has been. Play this clip. Kind of engagement or as different data center proposals sailed through

Starting point is 00:31:11 with hardly any controversy at all. No controversy, Lon. There's no controversy here. What's inside here, Long? Inside this shipping container is an edge of, educators dream, a hydroponic farm tended by high school students who sell their products at a farmer's market. Oh, it's for the children. It's salad for children and they're learning entrepreneurship. How did that happen in Utah? It's good lettuce. It's not bitter. No. The lettuce is not bitter.

Starting point is 00:31:37 Gordon Johnson is the former principal of Cedar Valley High School in the city of Uval Mountain, about 50 miles south of Salt Lake Sea. The hydroponic farm funded by META. It's a hydroponic farm. and Instagram and one of the country's biggest AI hyperscalers. Back in 2018, when the company announced plans to build this $2 billion data center on the edge of town, Johnson says she went to work. I started messaging Mark Zuckerberg right then. I'm like, we're a new high school. We're not opened yet, but I want to work with you because I'm all about community partnerships, right? So yeah, they gave me the money, and here we have it, and it's awesome. This is exactly what Cuban was saying.

Starting point is 00:32:19 Quarter million dollars. Remember that Mark Cuban quote where he was like, just buy the communities off. Corporate largesse that didn't end there in this old rodeo. Here we go. So they got a quarter million dollars for the farm. The ninth fastest growing city in the nation. Jared Gray is a literal cowboy and also the mayor.

Starting point is 00:32:36 So anyway, I think you get the idea. The mayor is stumping for it. And this was a data center built in 2018 in Utah, the same place where Kevin O'Leary is doing his data center. So MS Now, which this felt like a commercial for meta for me, like with the background music and everything. Yeah, I mean, you know, they're like these are big corporations behind MS now. They're not against data centers. So it probably is kind of.

Starting point is 00:33:02 I mean, I don't know who owns MS Now now. But I'm not sure who the corporate interest here is, but it did seem to me. They split off, right? It's Versant Media Group. That's the company that split off from NBC last year to. run all of their cable networks. So it's not part of NBC anymore. It's no longer a Comcast. It was Comcast. Now they're their own company called Versant Media. And they own CNBC. So not to be like crazy conspiracy there is here. But now you've got mainstream media saying, hey, you got it all wrong

Starting point is 00:33:33 on data centers. The same day his manifesto drops, this is a full on full court press. Let us build data centers. We need to build data centers. It's going to be great for your community. You're going to get hydroponic farms. We're going to put in a new stadium. They even have in this little love letter to meta that they started a soda pop store at the high school where you have these like Italian delicious sodas from the pumps. And I was like, okay, we know what's going on here. This is a absolutely constructed, and I hate to do the media deconstruction here, there is a PR group for meta that says, how do we, how can we turn the narrative on data centers around? Because we need to build more data centers. We need to catch up here.

Starting point is 00:34:24 How can we do that? Let's talk about, hey, you're not going to lose your job. Check. We're going to give you a bunch of free software. Check. The world's going to be safer. You're not going to have terrorist attacks. You're going to have your own tutor for your child, your small business.

Starting point is 00:34:38 We're going to give you software for that. And if you just learn how to use these tools, we're going to make them free for you for life. We're the group that's going to make super intelligence available for all people for free. All we ask is that you let us just build some data center. Yeah, it's a direct response to the big criticism, which is always like, well, these data centers are coming in. They're sucking up all the water. They're using up all the power. And then what do we get out of it?

Starting point is 00:35:03 They don't create that many jobs. It's funneling all this money to, you know, the billionaire class, the Epstein class. I feel like that's the narrative. So this is trying to cut directly at the heart of that. Like, no, actually, here are all the things that you're going to get out of this data center being in your neighborhood. You're going to get a free to do now. I got to say it. Whatever PR firm, marketing comms firm, they gave $10 million to run this campaign and playbook,

Starting point is 00:35:32 make sure to give him 20 million. It's a good idea. Because it's working. This was done so well that MS now hit his manifesto. It makes you feel great about AI again. Whereas Dario and Sam Altman saying, we had to come up with UBI and like this is half of all jobs are going away. That narrative scared everybody to death.

Starting point is 00:35:55 And of course you don't want data centers. The data centers are going to wind up in hacking and nuclear bombs and dirty. bombs and bio weapons. My eye photo is going to get hacked. My my eye messages are going to get hacked. My computer is going to be hacked. And we're going to have dirty water and my electrical bills going up. F this.

Starting point is 00:36:13 Yeah. Now we're just going to turn the whole narrative around well done to the PR marketing team at Meta. And I think actually it's a well-written manifesto. It was not written by AI. It was definitely not written by AI. I don't think it was written by Mark Zuckerberg. I think it was written by some professionals who came in and and polished this.

Starting point is 00:36:31 because I agree with you. It's very compelling and it's well-written. I feel it's his ideas, though. He may have had somebody co-write it with them, but I think, to my point earlier, I think he's taking control of the message. I think he actually believes what he's saying here. Interesting.

Starting point is 00:36:45 I think there's definitely ways to read it as cynical, but yeah, it definitely is like. What's the most cynical take you'll take it? Because I just gave, I think it's actually how he feels. I think it's actually the direction he's going and it's good for his business. And I think the most cynical thing I can say

Starting point is 00:36:59 is he wants to install data centers and he just wants to explain to people what they're going to get out of it. And it felt pretty even keeled. Yeah, I think that's the most cynical take is that this is really like a large-scale advertisement for like meta's agents. Like, hey, we're building all this stuff that you're going to love and it's going to help you and it's going to be built into the apps that you already have. And he's presenting it as I'm doing a public good.

Starting point is 00:37:23 It's a public service intelligence for everyone. But he's really just promoting the new agent coming to your phone. via meta. I feel like that's the most cynical take. I don't know. But I mean, if that's the most cynical take and they're dropping open source models today, they dropped Muse Glimmer. Yes. And Muse Spark 1.2. It's coming. And he's going to do an open weight version of Muse Spark 1.2. You know, the fact that he's releasing open source is proof in the pudding. You know, so I look at, like, I'm looking at it trying to be cynical. And I don't find myself getting. too cynical. I kind of believe him. And I think this is the best version of Zuckerberg. I'll say

Starting point is 00:38:07 very clearly. This is the best version of Zuckerberg I've seen in his career. Now, like my view on his career is irrelevant, but I have seen it from the beginning. And I think this is the most thoughtful approach I've seen him ever take. And he is thinking about the ramifications of what he builds for the first time in the history of the company. I think they did not, and they said they did. And they Didn't. Move fast, break things. All we have to do is grow the stock price. All we have to do is grow the user base. Don't worry if we break things. Now, this seems like, hey, here are the things that can break. And then here are the things that could break in your favor as a consumer. Yeah. We're going to work on not breaking things and we're going to break the arc of history and this technology to you. It is a master class. Let me just say this to Zuckerberg. This is a master class of a manifesto. I give him 10 of 10. No notes. Yeah, it takes a techno-optimist sort of view, which I don't think, you couldn't really say a lot of meta releases in the past felt that way. It usually feels like, I'm going to get you super addicted to this feed that's going to make you feel fat and ugly. Like, that's what the Instagram sort of pitch is in a weird way.

Starting point is 00:39:20 Like, you're going to doom scroll this all day and you won't be able to quit. Whereas this, I think at least we're starting from a more utopian vision of this is going to empower everyone. everyday people. Yes. Yeah. I don't think anybody feels particularly empowered by

Starting point is 00:39:35 Instagram. Maybe influencers, but other than them. So X has discontinued their revenue sharing program. The Everything Apps RevShare program

Starting point is 00:39:44 as it existed before will be retired on September 7th. They're already done accepting new signups in its place. They are introducing the original content

Starting point is 00:39:54 rewards program. New payouts will be based on qualified impressions generated by your original content. Now, this includes posts, blog posts, videos, streams, and so forth, but it applies only to, and I'm quoting, content you have personally created, written, film, designed, or produced

Starting point is 00:40:12 that reflects your own voice, perspective, and expertise. Love it. So the clear thing they're cutting down on repose, aggregation, people clipping a second off somebody else's content, and then posting a one or two word response, they're going to start cutting that or cutting revenue going to that, deemphasizing that in favor of all original content posts. To be eligible for the original content rewards, you must subscribe to X. You've got to be at least 18.

Starting point is 00:40:39 You have to have at least 500 verified followers. And you have to get 500,000 verified home timeline impressions that doesn't include replies, just people going to your actual profile over the last 90 days. Now, the old revenue sharing number required five million. home timeline impressions. So this cuts that down. But I did notice there are a lot of people like me who were eligible, got grandfathered in under the old X where it was easier to get to 500,000 timeline impressions.

Starting point is 00:41:09 And now they're not going to be eligible. I'm not eligible for the new rewards program. I was eligible for the old one. So we're sort of changing up the dynamic of the people who are going to be eligible for payouts. It's going to be mostly like bigger accounts. So I guess the big question would be, how do you think this is going to reshape X content? And what are the kinds of things on X you'd like to see less of in your timeline that you're hoping this cuts down on?

Starting point is 00:41:37 This is absolutely fantastic because I was never a fan of like Mario and Kane Coa and all these folks who were just stealing other people's content. A lot of them were offshore, stealing other people's content. And it just was a disaster. They were making large amounts of money off of other people's content. And I know the former head of product who just retired. Nikita Beer. Nikita Beer. He was calling people out on this.

Starting point is 00:42:14 And to change the actual incentive is a great idea. We want original content there. Now, I think they just wanted any content when Elon took it over. And it was like, hey, we're going to revenue. share, it did get people excited. And those people then went and got too excited to the point at which episode of All In drops, episode of Joe Rogan drops, they would just camp on the feed, make 20 clips, present them and get these massive amounts. And I'm assuming they would earn, you know, tens of thousands of dollars in revenue per month. Now, that was good for the platform in one way,

Starting point is 00:42:53 because that content found its way on to X in video, although it was coming in either a quasi-fair use, but generally illegal way. These were not the owners of the content. Clipping is kind of allowed for commentary, but systematic clipping is not. Right. And so what this does is this very thoughtfully gets rid of the systematic clippers

Starting point is 00:43:18 and just says, hey, you can clip stuff. You want to clip, you know, something from Joe Rogan or Tim Dillon. Go at it. Yeah. But you're not going to be making money off of it. Right. You can't monetize it. It's going to cut down on the repetitiveness in the feed.

Starting point is 00:43:34 Like when a piece of video would really blow up, I actually referenced it in the last, that meme about the casino butter. I don't know if you've seen it. The woman is making the table side steak on a stone. Anyway, when a clip would go viral like that one, you'd see it over and over and over again because everybody was trying to clip it, get people to reshare it on their feed, and then earn the payout. So if this cuts down on just that alone, it's going to make my feed a lot better, I think. And you have to get to your homepage, like your profile page.

Starting point is 00:44:05 Yeah, yeah. In 90 days. So when you sign up, they look at the last 90 days. And if you got half a million original impressions to your timeline, your home page, you're eligible. Somebody clicked on at Jason. Exactly. And if you didn't, you're not eligible. So I got through somehow in the old X because I just had enough followers or whatever.

Starting point is 00:44:29 So I was eligible for the old version of RevShare. But now under the new version, I don't get enough impressions on my profile, which is okay. I'm not on X to make money. So it's fine. No. I didn't need payouts. I appreciate the money. So for me, it's not going to make a big impact.

Starting point is 00:44:46 But I do think it's going to reshuffle a lot of the kinds of accounts that you're seeing. as much in your four-you? Because it realigns the incentives almost. I like it. I think it's great. I don't know if I'll qualify it. You will. I have...

Starting point is 00:45:01 You for sure, would. I don't know. I don't know about how many people visit my home page or not. What I do know is this has been a crazy amount of money that I never anticipated, I think. Let me see here. Can you see my screen? Wow. That's crazy.

Starting point is 00:45:16 That is a crazy amount of it. But it's very variable. over 150K so far in revenue share on X. But this, but here, the last two cycles, I guess they do it in two weeks cycles were 8,000. Then the two before that were 2,000. The two before that were six and five. Mine's very variable too. I mean, it's very variable.

Starting point is 00:45:35 Some months I'll get nothing and some months I'll get $500. So it is very variable if I think for everybody. But here's the other thing. I think you do want to avoid a situation where it's impossible to climb into those ranks. And we're, I think you don't want to centralize too much of the visibility on just like very large accounts that have millions of people looking at them. You want there, you want, you want, you want people to be upwardly mobile on, on X. I think that's the idea.

Starting point is 00:46:01 Like, you want, if I come on there and I make five great posts, I should be able to, like, build a following. Like, that's what the magic of it was. You had a lot of, like, shack joins and a million people follow him overnight because he's shack. But then there was also, like, all, like, tree bro. You remember that guy? Like, like, wint.

Starting point is 00:46:19 Like all these random accounts where they're just, they're really good at posting. And so they got famous for being good at posting. I don't, you don't want to lose those people. I think this will incentivize the right kind of people. This is a more YouTube creator program like, a program.

Starting point is 00:46:36 So, I mean, go for it. It's, the last one was too much of these, you know, breaking 911, Wall Street Apes. I'm looking. It was very gamable. It was very gamable. Daj designer who is good, I guess, disclosed TV.

Starting point is 00:46:53 It pays off those people that influencers who have like a narrative. Like every day I'm going to talk about like end wokeness is a great example of like every day I'm going to post this kind of stuff. All of it individually is going to be the kinds of videos that get a lot of attention. But in the in the aggregate, it's like I'm telling a story and I could just hit that beat five times a day. Yeah. I am very tired of King Coa, Mario, Weisgaard, Breaking 911, Wall Street Apes. I'm looking at all these people in my feed who are just, ugh, it's just, yeah, it's not for me.

Starting point is 00:47:31 And I like this change. And I like the idea that the people who actually make the original content will make the money, and then hopefully they put it back into it. So whether it's Nick Shirley or like, there's a guy I gave a hundred bucks to today with X money, this guy, JJ, who just reports on crime in San Francisco. I always thought this guy served a big purpose because the mayors are like, oh, my God, don't retweet them. Like I had one of the mayors of one of these cities literally reach out to me and say,

Starting point is 00:48:00 you're not helping by retweeting this guy. And I'm like, bro, I don't live there anymore, number one. Number two, that's your problem that you can't get the Honduran fentanyl dealers off the streets who've been there for 20 years selling fentany. Is this the what you said today? Where it's the guy selling steaks? It looks like it's steaks and sneakers they're selling. I didn't see any fenton.

Starting point is 00:48:25 Okay, yes. And when I saw that, I gave this guy a hundred bucks. This is J.J. Smith. And so what's happening here? You're like, hey, they're selling steaks. They're selling deodorant. They've got all these things. What they're doing is they're going to small businesses,

Starting point is 00:48:39 and even larger businesses. And these addicts go. and there's also gangs that do this. They just steal a bunch of stuff. They go to Market Street. They lay it out and they say, make me an offer four bucks, five bucks, six bucks,

Starting point is 00:48:53 whatever it is. But it is institutionalized. And tech workers, rich people think they're helping them by buying this stuff because they think, oh, these are like entrepreneurs. No, they're literally gangs

Starting point is 00:49:06 that have coordinated this. How do the gangs work? They did an incredible expose on this because the feds, long before Trump came in and broke this down, there were a group of gangs. They would go to people who were addicted to fentanyl. They'd say, here's what we're looking for.

Starting point is 00:49:25 They show them the pictures. We need razors. We need deodorant. We need this. With the stuff that you can easily resell. They'd say, go bring it to us. $1, $2, $3. Whatever you grab and run out of there with

Starting point is 00:49:37 will give you $1, $3. They go, they grab a bunch of stuff. They leave. they get cashed. They would take that stuff, not sell it there. They would take it to warehouses in Oakland. Then they would drive trucks down to L.A., San Diego, Laguna, Riverside, Inland Empire, all those places. And then they would resell it to bodegas there, like small mom and pop shops. And they'd say, hey, instead of buying stuff, we can get it here for $2 for a deodorant. You sell it for four. Hey, we all win. The way they beat, the way they caught them was all

Starting point is 00:50:10 All of these companies started finding their deodorant that was labeled for San Francisco with the serial numbers in San Diego. Oh, no, wait a second. They figured it all out. Wait a second. So anyway, I like this idea of also donating money to people. So I've been doing this thing on my X money is out. Did you get X money?

Starting point is 00:50:31 Well, I had to because somebody gave me $4.20. And so I had to sign up in order today. It was, you. But also, X gives you like 15 bucks. They seed the pot a little bit so that you want to sign up to get your money. Here's my account. I am on X-Av. Everybody's sending me 69 cents, 420, et cetera.

Starting point is 00:50:49 I guess the green is when they give it to me. Yes. And then as you can see, the black here is me giving it to other people. So I've been trying to give this money away as fast as possible to people. And then people keep this lunatic sent me $420. And I was like, don't do that. But then I just gave it away to other people. So now I'm running a multi-level marketing scam on my...

Starting point is 00:51:09 X account where people sent me 69 and I sent him for 20. But yeah, pretty funny. Good stuff happening. All right. I mean, my only thought on that is like, that's been a part of big city since I started living in them. Like people on the street, like, it used to be bootleg movies in L.A. where people would go record the screen and then put it on a disc and then sell you the

Starting point is 00:51:30 new movie on the street. That was also the mob, by the way. Oh, of course. Like either, yeah, they would, when I lived in Brooklyn, it was the actual mob. I think it went to other criminal enterprises. They do it. Then they find fences to go sell it on the street. I'm sure it was organized crime in some way, shape, or form, of course.

Starting point is 00:51:51 Oh, well, they basically took the DVDs and said sell these and bring them out. Even back in Goodfellas days, they're selling the bootleg cigarettes out of the back of the truck. Anyway, let's move. Let's move off a truck. Yeah. We got to talk about the Clarity Act. It's facing another big setback. Senate Republicans.

Starting point is 00:52:05 This is, of course, the crypto regulation legislation that we've been working on. for forever. Senate Republicans did not take up the bill before leaving for their August recess. That means it's not going to get debated until September at the earliest, which is going to put the debate about it right ahead of November midterms. And there is a sense among Republicans, for whatever reason we can discuss maybe why we think this. They don't want to dig into this right before everybody goes to the polls in November to vote. Here's North Carolina Republican Senator Tom Tillis. I do think the odds drop precipitously of this getting passed. means we leave for a month, we come back, we've got an election ahead of us, I think it gets

Starting point is 00:52:44 difficult to get it done. So the bill, just to go over, quick overview, backed by the majority of the crypto industry, Coinbase was sort of not sure they were in. They're back in now. They're supporting it. It provides a legal framework and jurisdictional certainty for crypto companies, basically decentralized digital communities go to the CFTC and fundraising and investment contracts, go to the SEC, and it resolves some basic things about, you know, like can a digital

Starting point is 00:53:09 asset, can digital asset service providers offer interest or yields or rewards for your stable coin balance? And those kinds of debates would get discussed. The one real sticking point back and forth that they're still arguing, Dems want to include an ethics provision addressing government officials with ties to the crypto industry. You can sort of imagine why Republicans are not as enthusiastic about doing that. So, I mean, in your, from. Melania coin?

Starting point is 00:53:36 Yeah, I think, I think that. How's your Melania coin holding up? obviously what Dems want to do is they want to use this as an opportunity to point out Trump and all of his various connections and how much money he's been making from all of these crypto projects and Republicans feel like, especially ahead of the midterms, that's not maybe the conversation they want to be having. So what do you think looking at the facts of the ground? I mean, is this legislation probably dead? And what impact would that have? I mean, this was kind of a signature Trump effort. You know, we're going to pass all these crypto regulations.

Starting point is 00:54:07 I think it's going to pass right after the elections, probably, if this is getting blocked because there's so many donors in the crypto space. So, you know, there's money in politics. I hate to let people know, but there's gambling at this establishment. What's that line from gossip? I'm shocked that there's gambling in this establishment. I am shocked to find that there is gambling in this establishment. And they're like, sir, you're winnings.

Starting point is 00:54:32 He's like, thank you. He puts it in his pocket. You know, they're just, all of our government, officials have been bought and paid for for years over this legislation. One of the big donors to both Republicans and Democrats has been the crypto industry. They have used their influence and their balance sheets to buy this legislation and buy it they shall. They will get it right after the midterms. It's probably just a bump in the road. The thing I care about is how does this work in terms of

Starting point is 00:55:07 raising money. Now, the SEC is going to keep jurisdiction over the fundraising, yeah, over the fundraising, like initial token fundraising, investment contracts, while the CFTC is going to take over once the network is mature and decentralized enough to trade as a digital commodity. I don't understand that part of this. And, you know, how does it work for a startup? So if I start a company and I say I'm going to sell a thousand membership NFTs to it for $100 each. And then you're going to get, you're each going to get 1% of the revenue for the first year or whatever. Like, what am I allowed to do here? And there's going to be some new offering exemption.

Starting point is 00:56:01 And it's going to have a dollar cap. So it's going to cap sales at $75 million over a 12-month period. and require issuers to file an offering statement. And then some other update to it was let issuers raise the greater of $50 million a year for four years or 10% of outstanding ancillary assets cap that 200 million aggregate with tailored disclosures and full registration. So you're going to have to register with the SEC. Anyway, this seems like chaos. And I am very interested in the impact it will have on startups. Like what will startups be able to do in terms of fundraising?

Starting point is 00:56:37 Because we did have a number of our startups decide, I'm going to do an ICO. And we were like, is that legal? And they're like, we don't know. Yeah. And but we're going to do it in Panama with a foundation. And it's a million dollars to set up the legal entity and we're raising 10 million. And I'm like, I've never heard of starting an LLC or a C corporation costing a million dollars. And I've never heard of domiciling it in Panama with a,

Starting point is 00:57:04 hidden board of directors because the board of directors doesn't have security concerns. Like when I heard this pitch, I was like, you know what? I can't be on the board of this company. I need you to buy me. I literally told the founder like, once you go to Panama and start an organization with an anonymous board of directors in an organization that I'm an investor in, time for me to go. I am not going. I got three daughters. I got a family.

Starting point is 00:57:29 Like I can't go to jail. I can't go to like prison. It's not worth it. It's not worth it. What are we doing here? Like, why are we? And I asked the woman who is the attorney pitching us on this foundation.

Starting point is 00:57:41 Like, why does it cost a million dollars when you could go on to like, you know, any website if, you know, and sign up to get your incorporation for under $1,000? And she's like, oh, well, it's very complicated and not many people have done it. And I'm like, got it. Yeah, there you go.

Starting point is 00:57:56 Why haven't a lot of people done this? Hmm. Hmm. Yeah. Because it's illegal or sketchy. And it was just pure sketch. But that's something I would like to have somebody who's in this explained to us. So if we can put that on our producer Jacob, our list of potential topics in a future episode.

Starting point is 00:58:15 If there's an attorney who understands how this will affect fundraising for startups, I'd love to hear it. Claude gave me a little bit of a tip. If you're day one of your crypto startup, you're presumably under the SEC. So if you're launching a token where your team still controls upgrades, treasury, a large chunk of the supply. why you're an investment contract asset issuer, all of the usual securities and fundraising rules of the SEC apply. So that's... Okay, so no change.

Starting point is 00:58:43 Yeah, so no change there. We could certainly go deeper with more of an expert than me. But what if we launch a token? You know, your startup launches a token. Right. So after it raises its $50,000 or startup fame. Secondary trading changes the picture fast. Once your token is resold by someone other than you or your agent, it stops counting as a security

Starting point is 00:59:04 for that transaction. So exchange listings and liquidity events matter a lot more to your regulatory posture. So if I get listed on Coinbase or Cracken and somebody else is selling it, then I'm out of the woods. Correct. But that requires those people to be the gatekeepers then? Yeah. Okay. Claude even says real uncertainty remains.

Starting point is 00:59:26 It still requires. I don't want to touch this. I'm like, I'm producing. I'm like, sorry, bro. My name's Claude. Leave me off the pod. Leave me off the pod. Talk to the Panama legal lady.

Starting point is 00:59:36 She'll help you out. She'll set you up good. Yeah, even Claude, I've never seen that before. Even Claude is like, listen, don't, don't know. I don't want to deal with this. Claude went like this. Yeah. It's true.

Starting point is 00:59:48 So good job, Claude. He did that meme. Yeah. That's the hands up. Yeah. I don't know what that meme is called, but I can picture the guy doing it. It's the guy who is the black guy who does like his own talk show. where he rates people, they come in and they like, he's sitting in a king chair.

Starting point is 01:00:08 Wait, is it, is it Drusky? Is that the guy in that? I think it's Drusky, yes. That's the guy. I think it's a Drusky like. You're right. No, you're right. It is, I never, I did not make that formal connection, but you're right.

Starting point is 01:00:21 It is Drusky in that meme now that. What's that meme called? Let me, let me look up the Drusky, Drusky shrug meme. I mean, it could be just a Drewski shrug. Let's see. Drew. Yeah, Drusky Shrug, I don't know. Or the Drusky hands-up meme seems to be what people are talking.

Starting point is 01:00:39 Here it is. That's it. And you see that. Yeah. That was just what Claude did. Claude's like, you're on your own. Yeah, I don't know. You're on your own.

Starting point is 01:00:47 You want to do a crypto? Samani on your show. I don't know. Yeah, I'll figure it out for you. I'm good. All right. Sorry, Claude. I didn't mean to stump you on that.

Starting point is 01:00:56 We're calling Claude out. You want to do one more story? We got some time. Yeah, keep on. So, According to the New York Times, AI agents are helping college students cheat on mass. A new report released on Monday claims that AI is routinely undermining online education. Professors are unable to counter AI-enabled cheating.

Starting point is 01:01:16 This is mainly happening, Jason, on courses that they call asynchronous. That means students aren't interacting in real time with their professors or their classmates. Even over video or Zoom, they're watching pre-recorded lectures, apparently with modern AI. agents being so advanced now. Students can literally give them instructions like log in, complete all of my coursework, and the agents will watch the videos, take their tests for them, write their papers for them, even chat with the classmates and fake it about their reading assignments. And several professors interviewed for the article described totally not knowing what's going on with their students and believing that they're spending a lot of time interacting with bots. Here's a community

Starting point is 01:01:58 college teacher named Catherine Kaiser. It's just so obvious, those of us who have been teaching for a long time and our editors and writers can usually recognize the cheaply written AI stuff within 30 seconds, is what she had to say. What is the answer? I'm trying to take the Zuckerberg approach here. Is the answer to this just not have these kinds of asynchronous classes anymore because it's too easy to cheat?

Starting point is 01:02:21 Yes, that is the answer. There was a professor and Paul Graham retweeted this distribution. And I think we talked about it, so we'll find that one quick here. And it was very clear to him that people had cheated with AI. So he did the final in person. When he decided to do this, and it might have been Princeton students or whatever, but everybody's cheating. It was Brown University.

Starting point is 01:02:50 I found the tweet. Okay. So here we go. You look at this. There's a couple things that happened. I was talking to my nephews. and they were telling me about college and cheating and everything. They said during COVID, people made lectures.

Starting point is 01:03:04 And one of my nephews was taking a course, and he's like, I dropped the course, Uncle J. Gal. Because the teacher didn't even teach. He just said, watch the video. And then we asked questions they sync. And I was like, wait, what are you talking about? I was like, oh, they made their entire lectures and recorded them during COVID.

Starting point is 01:03:24 And now you're just taking like a, and I'm like, what? you're just basically going to YouTube and taking a course. I was like, exactly. I'm like, well, why are you paying for that? And I dropped it because you have a moral obligation to at least teach the class. I mean, my lord, this is just so unethical. And if you look at this, this is the econ. So this distribution you're looking at that the one side is showing them taking the test using AI,

Starting point is 01:03:51 and then he had them come in. So the midterm they took using AI asynchronous at home. And then the final, they had to come in and sit in front of him and actually take it in person. And so you're looking at when these huge gaps are appearing where somebody clearly did incredibly well with the AI version of the midterm, and then they did very poorly with the in-person exam, the indication is they were cheating using AI. And you can see how few of these students have numbers that are close together. Like their at-home performance matched their performance in the classroom. Presumably, those are the kids who are watching.

Starting point is 01:04:29 The people on the, the number on the right is the midterm, which was done asynchronous. Correct. And if you go to the top, you'll see there's the person, student number one, got a 95.5 at home and a 95 in person. In other words, half a point difference. Yeah. And then it just starts expanding and expanding to you. It gets a student number 22 who got a 50.

Starting point is 01:04:51 59 at home and a 55 there. In other words, they are not the brightest kid in the class, but they didn't cheat. They're doing their own work, yeah. Yeah, and that's the one you want to hire. Right there, that's the person. That's the person who could become Captain America. Remember Captain America? There's a great scene.

Starting point is 01:05:12 I just rewatched Captain America, or at least the first half of it. The first one, the first Avenger. Yeah, and the first one, the guy who's choosing who's going to get the serum. He takes a grenade during training camp. He clicks it. He throws it on the ground. Says grenade. Everybody runs except...

Starting point is 01:05:30 Captain America. For... Dives on top of him. Steve Rogers. Steve Rogers, who is all of like 90 pounds at this point, jumps on the ground, grabs the grenade, and then makes another pump to, like, hold it in tighter. And it's a fake grenade.

Starting point is 01:05:47 It's the greatest scene ever. Yes. Stanley Tucci, Dr. Abraham Erskine, he's the researcher. It's the other guy from the fugitive. Oh, Tommy Lee Jones, yeah. Tommy Lee Jones is the one who comes up with the test. He says, grenade. And here they are.

Starting point is 01:06:06 They all dive. And one guy, Steve Rogers, that's my guy. And then he does a second pump. Get away. Just to make sure there's no shrapnel on anybody else. And then everybody realizes, oh, there's one person true of heart. And that's why he can pick up Thor's hammer. That's right.

Starting point is 01:06:23 That's like he's deserving of the Captain America serum instead of all the already powerful, strong soldier guys, exactly. But remember, they said, Thor said nobody can pick up this except you're pure of heart. Thor's hammer. He has to be worthy. You have to be worthy in the only year. Yeah. Yeah.

Starting point is 01:06:40 And so he's worthy of it. That's that 59. That's that one student there. The bottom three, by the way, who got zero on the final? They dropped the class. They didn't even drive. As soon as they realized they were going to get to use their AI, they were like, I got a hundred. I'm done.

Starting point is 01:06:56 And it's zero. They were out. Those are the people who you expel. Those people, those people could work in business development, maybe. Yeah. There's something they could do. There's a job for them. I don't know what it is.

Starting point is 01:07:09 But yeah. Yeah. I mean, I do. You know, we've had a lot of people who make AI tutoring and education apps. And there's always. all this sincerity. Like there are definite ways to use this to learn and expand your horizons. I've learned a lot from interacting with AI bots, but it is you do have to bring that desire to learn to it. You can use it very lazily. Just give me the answer. You have to like bring that

Starting point is 01:07:39 with you that I have desired to iterate and keep using this until I actually learn the material yourself? The only thing you really need to know is that you're only cheating yourself and everybody's cheating in school and the blue books have to come out. Everything needs to be in person. Pop quizzes, reading books in person. I literally went through this with my daughters this summer. They had to read a book. And I said, here's how we're going to do it. I bought like two copies of the book. I said, these are workbooks. They only cost like four or five bucks. I lied to them. And I said, it's okay because they ethically didn't want to write up the books. Sure, I get it.

Starting point is 01:08:18 These were like paperbacks. And I was like, girls, these are paperback workbooks. You circle any word you don't know. And I want you to underline the most important sentence on every page in your estimation. And at the end of the chapter, I want you to write a summary in the book of it. And now we read the book. Now retention's gone over. And we call this hand-thinking or hand-speed.

Starting point is 01:08:43 I would talk about it on a previous episode. Handmind. Hand mind. You borrowed it from Ursula K. Leggin. Yeah. So handmind is your mind is going only as fast as your mind, which is why when I'm sitting here on the podcast, this book is here. Because I like to slowly write one word. Like I wrote one word, Jason Bot while we were talking. And I said, you know, we should have a bot on Twitter that goes through the show in real time, like we're having our real time commenter. And just Jason Bot says what we're talking about. We're posting. quotes in real time. So if you weren't watching the show, you could just hear quotes from there. And I won't forget it. So there you have, folks. There you have it. Anything else on the do do? We should do lightning around. Yeah, well, I do, I want to mention, we'll do one more story, but before we get to that, one thing I do want to talk about just so we can clip it and share it with everybody. We are hiring for the podcast. We are. We have many open positions. There's three that I want to talk about for our purposes on this show. One, we are looking for a new podcast.

Starting point is 01:09:44 podcast and video editor. You can go to careers.launch.co, by the way, and look up all of these. That's where you can find all of our job postings. But this one, you can also just email editor at launch or killer editor at launch. at launch.co and apply to this directly if you don't want to mess around with the website. So these are people who are going to help us make not just the podcast itself, but make all kinds of great content for all kinds of platforms based on the stuff that we're making. for the show, clip it and put it in TikTok, Instagram, LinkedIn. We want to be where everybody is.

Starting point is 01:10:20 So that's the podcast video editor. We are also looking for both a senior podcast producer. This is somebody with three to five years established experience, making great high-level podcasts like this one. And then we also want a more junior podcast producer. Maybe you're not as experience, but you're a great writer, a great researcher, and you're very enthusiastic about AI, tech, startups. you want to learn this world.

Starting point is 01:10:46 Hands on, I mean, you want to set up amazing interviews like we had today with Const from BitTenser. You're a more junior level person. That's okay. We want to work with you. But you want to learn. But you want to learn. That's, yes, a passion for what we're doing and an interest. I mean, I think when we've interviewed for this position before, I tend to get a lot of people

Starting point is 01:11:07 who are experienced veterans at podcasting, but they're maybe not as passionate about what we talk about. the tech, the AI, computers, startups, being a founder. You know, and if you're not interested somewhat in that stuff, this job might be a little bit of a drag. We're looking for people who are really, they want to learn everything there is to know about starting a company, AI, being in tech, that's who we're really seeking. I mean, if we were forced to start a podcast about like poetry, like, yeah, maybe we wouldn't

Starting point is 01:11:40 wake. Well, maybe we would. I don't know. But we ideally want somebody with a passion for the subject material and who love the technical aspects of the job. This way, you're in your sweet spot and you enjoy it. Exactly. So for the producer jobs, you can write us at great producer or just producer at launch.co. Or go to careers.

Starting point is 01:12:00 launch.com and find all of the job postings. We'd love to hear from you. I like, when people just email us and just say, like, hey, here's what I've done. Here's a link to our LinkedIn and get going from there. I wrote a tweet and I wrote, please send us a resume and a cover letter and people made fun of me because that's not what you do anymore. Okay, boomer. Exactly. They're like, okay, boomer.

Starting point is 01:12:21 A cover letter. It did sound a little antiquated. So I've updated. Yeah, it's a bit antiquated. But, you know, listen, it is what it is. We need more people because AI is replacing everybody's jobs. But our business keeps growing and we need more people. Too many sponsors.

Starting point is 01:12:37 We sold out the entire year again. We got this week in AI. We've got plans to do more. And we're looking for more people. So help us out. AI still can't do all of my job. A little bit. Not all of it, though.

Starting point is 01:12:48 All right. One more thing I wanted to talk about before we let everybody go. The second it can. Please. I'm begging you. Let me do my job. I get some sleep. I go back to Italy.

Starting point is 01:12:59 I'm exhausted. So last thing. Spotify, do you hear about this? They're testing a skip button for their premium subscribers, Jason. And the button will controversy allow premium users to skip through sections of their favorite podcasts, including the ads, the publisher sold ads, but Spotify is exempting their own ads. So you can't use it to skip through an ad that Spotify sold, but you could use it to skip through like all of the ads that we're putting on our podcast that we're not selling through Spotify because we have our old ad team. So Spotify's saying this is just an expansion of the 15 second skip a head button, which already. exist and which people can already use to bypass their ads.

Starting point is 01:13:41 And they argue that, hey, if you make podcasts, this is good news for you because listeners are going to be more engaged with podcast content. And their studies show that this keeps people listening to podcasts for a longer time. Still a lot of blowback out there. Tommy Vitor of Crooked Media said, very cool of Spotify to try and destroy the podcast industry. What a great partner. Yeah, they're going to back down on this one. I'm going to text somebody I know who may or may not be in charge.

Starting point is 01:14:13 Please. Yeah, this is just dumb. It's dumb. There's already fast forward 10 seconds, right? You could hit the button and it'll jump 15 seconds. People are already using that to skip the ads. Yeah. And you know what?

Starting point is 01:14:27 I listen to the ads in these podcasts because the amount of time it takes to take my phone out and hit skip. I'm just like, okay, you know, like I'm going to listen to the ads. and you're listening on one and a half speed anyway. Just enjoy the ads, folks, and it pays for it and keeps it free. But for the platforms, like this is a big way to build up resentment. And these platforms need to respect the podcasting standards. And this is one of the things I've been telling them over and over again.

Starting point is 01:14:51 Stop breaking the standards. Work with the standards. Work with Adam Curry. Work on podcasting 2.0. There's two features in podcasting 2.0 that Spotify absolutely and Apple must respect and use. One is a donate button. the donate button should be set in the RSS feed, and it should take it directly to whatever the preferred donation is.

Starting point is 01:15:11 So if I want to use PayPal, you want to use Stripe, somebody else wants to use Patreon, just let them have their goddamn button. Okay? And then number two, when we go live, let us drop a live into our RSS feed. You will get that and ping it, and then send an alert on our behalf. YouTube does that.

Starting point is 01:15:30 Now, if Spotify doesn't want to keep losing to YouTube, which is where people are starting to watch podcasts and where podcasts are spending more time, don't do things like this. Spotify should be in the corner of podcasters and for open standards because YouTube is a black box. YouTube doesn't do any of the standards.

Starting point is 01:15:47 They don't use RSS feeds, all that kind of stuff. And, you know, where are we going to spend our time is what they need to ask themselves? Because the difference between using Spotify and YouTube and putting it in your pocket and listening to the background, is there a difference?

Starting point is 01:16:01 It's getting pretty negligible. I mean, I think my biggest concern with the whole podcasting ecosystem is that it feels like increasingly we're centralizing it in just a few platforms. Podcasting used to be like you could get a show wherever you wanted. It was going out on this feed. You could pick it up from an independent service. You could listen on Amalea. You could listen on iTunes. You could listen on Spotify, wherever.

Starting point is 01:16:21 And now it feels like YouTube, Spotify, and maybe Netflix, if you get big enough, Netflix will pick it up. And that's it. Like we're sort of cutting down on all these other places. Apple, an Apple podcast. But it feels like it used to be this, like, huge landscape of different varieties in where you could get your podcasts. And I don't know why we'd get rid of that. Like, Spotify's not threatened by small, independent platforms.

Starting point is 01:16:45 What do they care? I'm just writing a tweet now. Why doesn't X support RSS feed slash podcasts? Yeah. For years, that was just the standard podcast. I should be able to have a tab on my profile page with my latest podcast. Yeah. That used to be how, that was the whole point.

Starting point is 01:17:06 This would be insanely effective. You're broadcasting to people's iPods. That was the whole idea. Yes. It's just so annoying when people try to break standards. And I told this to Spotify when I met with them during South By. And I was like, just please stop breaking the standards and please support the new standards. Like the reason Spotify has so much extra traffic is because of the podcast.

Starting point is 01:17:32 standard. Why would you want to break that? You're just working against yourself. And it really pisses off people like me who have a number of podcasts. And like, it really makes me feel bad about my relationship with Spotify. And a lot of podcasters don't like Spotify, a lot, because they don't give us the controls we want. And then they don't support standards. And they need to listen more. Spotify needs to listen to podcasters more. And Apple is kind of like not as engaged as they once were, they kind of went dark for a little bit, but they're getting back, they're realizing podcasting is a good business for them. And they're starting to, I think they got some new leadership over there, they're starting to pick up the phone and call people again. They're starting to

Starting point is 01:18:15 feature podcasts and they're realizing this is a good way to keep people on their iPhones. And so, you know, take from that what you will, Spotify. YouTube, Apple, you got to kind of, you know, be in our, be in the podcasters corner. Don't do things against podcasters. Yeah. A little behind the scenes, it's also Spotify makes it very difficult to do your distribution, unless you use their tool. We're going through this behind the scenes right now where we tried to switch our distribution. And many of you in the audience wrote to us and said, it broke things all over Spotify. It's like, you could just make that easier.

Starting point is 01:18:51 Well, that was because of video, right? Yes, it was because of video. Well, it's also reporting. Like, they want, they, if you don't use their own distribution platform, it's much harder to get your analytics. And it's like, well, just let us have our analytics. That's how we sell our ads, Spotify. Yeah, there's another important note for them. Like, guys, gals, fellas, ladies, just don't be heavy-handed.

Starting point is 01:19:13 And sometimes it's not even heavy-handed. They're just like, oh, we'll get to it. No, make it a priority to support open standards. Yeah. Make it a priority to support the ecosystem that you've gained so much from. Nobody has gained more from podcasting than Spotify. It's like their business is just like this app, you know, like there is no Spotify. without people making amazing things to listen to is Spotify.

Starting point is 01:19:35 I appreciate Spotify's out there, but I don't feel appreciated by Spotify. Sometimes if Dylan comes out with like a new single or something, I'll send it to my brother, the Spotify link. No, no, I'm saying I don't share our podcast. No. Because we don't feel love from Spotify, I tend to share the YouTube or X or Apple or, yeah. Oh, the Apple. Yeah.

Starting point is 01:19:57 I would like to be more. I would have a better feeling about Spotify. Yeah, I don't want to just get to invited to Spotify parties. That's nice, but I am not like Spotify parties nice. Limited appeal. Yeah. Yeah, to support standards, please. Yeah.

Starting point is 01:20:12 Make it a priority. That's all. Make it a priority. All right. We'll see you all next time on this one starts. Bye bye.

This Week in Startups - Zuck's AI manifesto is a data center PR masterclass | E2323

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.