Latest / The Media Odyssey / THE END OF THE PARAMOUNT SAGA AND THE TECH-BROS OF THE APOCALYPSE

Transcript

THE END OF THE PARAMOUNT SAGA AND THE TECH-BROS OF THE APOCALYPSE · The Media Odyssey

• Marion Ranchet: [00:00:00] Welcome to the Media Odyssey podcast. That is Evan Shapiro. Evan Shapiro: And that is Marion Ranchet. Marion Ranchet: This week on the menu, we're gonna be talking about Paramount, the fact that Evan was wrong. This one I have to say I particularly love, and then we're gonna talk [00:00:15] about AI.[00:00:30] Without further ado, Evan, you were wrong, man. Evan Shapiro: I was. I was very wrong. I did not think that this deal would close, certainly not this year or until the trial [00:00:45] started. and that if it did, the attorneys general would elicit some true structural changes, like sales of assets, which is, and I thought that because that's what they said, the attorneys general. That's what Rob Bonta, who was [00:01:00] the lead in this case, and I think people need to understand, he was the lead in this case because, the majority of the business takes place in California. But secondarily, California was funding the lawsuit to a certain extent, as opposed to the rest of the 11 [00:01:15] states, and he caved. Now to a certain extent, he caved, and he, he did get some guidelines, and elicit structural remedies without sale. But yeah, [00:01:30] I was completely wrong. and it doesn't make sense as to why the attorneys general settled- at this point, especially with the ticking fee of $7 million a day kicking in next week. It makes, and the rumors are that three or [00:01:45] four other attorneys general are like, "Why are we settling now?" but what became very clear is that he caved to political pressure, specifically from Gavin Newsom and Karen Bass. Marion Ranchet: Interesting. So let's, let's take a look at those mandates, right? And I [00:02:00] think, by that we mean what the AGs obtained, what the regulators got out of that deal in a way in exchange to, have it approved. The first one is the theatrical release guarantee. So [00:02:15] Paramount agreed to release 30 films theatrically per year, in the first two years, and then, 32 films annually in years three to five. that was a biggie, right? Because we had the case of [00:02:30] Fox and Disney making promises, and as we've seen, they went way below, even 20, per year by the time that the two companies combined. So is that a win for the movie business? Evan Shapiro: I think before we get into that [00:02:45] specific thing, this is all contextualized by the fact that there's an independent monitor inserted into the C-suite at Paramount, Warner Brothers, Disco Bros, Skydance, Nepo Baby, and they have authority to [00:03:00] basically report Paramount's lack of following of these rules. So there is a, there's a monitor inserted into the company who's completely independent, or at least that's the theory going on now. So that's a big thing, and it's a major difference [00:03:15] between the Comcast acquisition in 2011 that really no one followed the rules of. Yeah. So, start there. It was a major one. Ellison had promised this kind of from the beginning, and it was one of [00:03:30] his major theories for the, the combination of the two companies. I think that people don't understand there's a huge escape clause on this agreement or on this rule, which is they could simply just buy a [00:03:45] percentage of a film and release it. They don't have to actually produce the film. So they could acquire a film at Sundance or some other festival. They could acquire a film. For example, Warner Brothers Discovery sold off the rights to Acme versus [00:04:00] Coyote, and someone else is releasing it. Paramount could have just acquired that film without producing an original film. So there are some back doors on this. But there is a connection to establishing a $25 million independent, [00:04:15] film acquisition fund on top of that. So I think from a film production standpoint, there is, there's a lot to like about that rule and crucially, connected to that, they have to [00:04:30] operate the two studios, Warner Brothers and Paramount as independent corporations. So the, the synergies, the cuts that they would have to make to operate, and they have to keep both lots. They're not allowed to sell either lot. [00:04:45] So on one hand, I think that's good for LA, that's good for Hollywood, that's good for California, and also good for New York, by the way. But, at the same time, it makes it very difficult to find the synergies that this plan was supposed to [00:05:00] produce. Marion Ranchet: All right. And that's production, production quotas, investing in the cinema, et cetera. Let's talk about news 'cause that's a big one, and I think this is the one where perhaps you [00:05:15] were hoping that assets would be sold to keep some sort of, how can I say? Fairness and what's gonna happen on that front? Evan Shapiro: Yeah, I think, it's weird. Bonta and the [00:05:30] attorneys general always said that CNN was not a part of this calculation. It wasn't about the independence of CNN. And they were clearly lying. It was clearly a concern of theirs because this is one of the major [00:05:45] components to the agreement of this settlement. There's an independent oversight board that's going to be established that has authority to report any meddling by the, by senior management in the [00:06:00] newsrooms of either CBS or CNN. And when you put that in context of what's happening at 60 Minutes right now, where Bari Weiss is reported to have killed stories, especially stories that have to do with her passions, [00:06:15] Israel and Gaza being one, the Epstein files being another, and her wife is mentioned several times in the Epstein files. So the idea that there'll be an independent, editorial board where [00:06:30] reporters can go and report meddling, and that if that, meddling is, seen to be actual, they have cures built into this agreement. Now, we have to see all the details of this. That should be really interesting. And, it, and I [00:06:45] think it, it begs the question as to who's gonna be running CNN on the other side of this. Mark Thompson, who's a legendary executive, from BBC, and The New York Times, he's been trying to calm the waters over at CNN. I [00:07:00] don't know how this is gonna function. News is so fast and furious. It'll be interesting to watch this happen. But I do think it begs the question, especially since the, the ratings at 60 Minutes are at their [00:07:15] lowest level this century. The ratings at CBS News are at the lowest level this century. The ratings at CBS This Morning are at the lowest level this century. So not only is Bari Weiss infiltrating and meddling at the news organization at [00:07:30] CBS, she's destroying the viewership. So it does beg the question on the other side of this, what happens to Bari Weiss? I think she might get thrown out as part of this. I think this gives Ellison cover to get rid of a [00:07:45] problem he's having. I think they agree on a lot of things ideologically. And I think that the Justice Department and the White House who really helped usher this deal through, are big fans of Bari Weiss but she's become a major [00:08:00] issue, for the corporation on many levels. She's costing them hundreds of millions of dollars in advertising revenue as a result of her poor management. So I think you're gonna see, I think it's very likely that you're gonna see some substantial [00:08:15] changes, to the management structure at CBS News on the other side of this. And I think it's pretty likely that Mark Thompson winds up at the top of this news organization, not, which will piss off the White House but he's a good executive, without a real [00:08:30] ideological side. He is very down the middle. So I think you could see a substantial, editorial and management change on the other side of this. Marion Ranchet: There's the topic of Pluto TV. I have to admit that when you bring together [00:08:45] HBO Max, Peep Plus, Pluto felt kind of the odd one out. Especially HBO Max is not really well-versed in ad-supported content. I think it's interesting and it's a [00:09:00] good thing. I think they have to keep a free porch, front porch for five years and that would be Pluto TV. How do you read that? Evan Shapiro: It, to go a little bit further, they have to keep it but if they [00:09:15] don't, it gets spun out. Marion Ranchet: Interesting. Evan Shapiro: So the cure isn't some, find, a lot of these others, like if they start to mingle the cable assets in negotiations or, mingle the business practices or [00:09:30] operations of the two studios, there's a cure which is divestiture. If they don't spend the money on the independent film or the training, the independent monitor has the ability to go grab that money directly from Paramount's bank account. Those [00:09:45] are specific cures that have financial ramifications for the company. If they don't operate Pluto as agreed, there's no financial, problem for the, or there's no financial cure or punishment, for the company. They [00:10:00] just have to spin it out. I think and I, have friends that work at Pluto, I have lots of friends who have left Pluto in the last couple of months or or the last year and I think they've just been ruining it. It just feels like it's been [00:10:15] ignored for the most part over the last year, especially here in the US. I think less so overseas. I think it's actually operating in a really interesting way. In Europe in particular, it's really been aggressive. But here in the US, [00:10:30] it just feels almost like an afterthought. I also think that comes from the fact that Ellison has been so distracted on this merger that operations of the company have taken second place. That said, I will say, Paramount has a substantially better ad [00:10:45] sales team than Warner Brothers Discovery. They are really good at selling advertising and have been for some time. And Warner Brothers Discovery is terrible at it. Not only are they bad at selling advertising because they got rid of a lot of the most senior people [00:11:00] over the last number of years but secondarily, they have utterly failed at selling advertising on HBO Max over the last number of years and they have not connected Pluto and HBO or Max or whatever the hell you wanna call it these days, [00:11:15] into one ad ecosystem. So they're, they lag behind. Their ad sales is substantially behind their consumer consumption. So I think Ellison sees Pluto as [00:11:30] a kind of central component to connecting the ad sales upside of this enterprise. I think you're gonna see Pluto, I think the name will disappear in all of this. I really do. I think it's time to sunset the original [00:11:45] concept of it. Marion Ranchet: I agree. Evan Shapiro: But I do think you're gonna see the free front porch or the free system, especially with Roku and Tubi ultimately combining in that acquisition. I, it'll be interesting to see what happens but I think Pluto becomes the kind of basically [00:12:00] streaming ad sales platform of the combined entity. Marion Ranchet: Nice. Okay and so last but not least, the impact on jobs, right? So I think I saw a piece from you where, are you saying that this [00:12:15] deal in particular, now that it's been approved with those mandates, et cetera, it's actually gonna make it even worse than before? Is that what you were saying? Evan Shapiro: Yeah. This, this deal so this deal [00:12:30] has $1.5 billion of investment into the state of California film production. It has mandatory money requirements for training of workers who might be displaced. There's a lot of costs added to the combined company because of this [00:12:45] agreement. Secondarily, the need to operate the two studios separately, the need to negotiate the two cable packages separately that eliminates a tremendous amount of the synergies that were built [00:13:00] into the business plan when they were raising money. And so on the second hand, you have to look at what the assumptions are in order for this deal to make financial sense. So they're gonna enter the [00:13:15] new era with $81 billion in debt. And currently, the company combined generates about $12 billion going down, by the way, not going up, $12 billion in EBITDA per year. So the two entities [00:13:30] combined generate $12 billion in EBITDA per year but declining, not increasing. The assumptions of this plan are that they have to find $6 billion in synergies and cuts. And they have to grow the [00:13:45] EBITDA by 50% in 36 months. 50% from, from 12 billion to 18 billion. How are you gonna do... And not revenue, EBITDA. How are you gonna do that without [00:14:00] destroying whole departments? How are you gonna do that? The single greatest cost of this company is SG&A, is people. And by my estimates, they're gonna have to let go somewhere [00:14:15] around 20,000 people across the planet over the course of the next three years to make this make any kind of financial sense. The combined companies have already shed nearly 18,000 jobs between them over the [00:14:30] last three years. And so I just don't see how they get to the balance sheet math without shedding jobs. And remember, right after Ellison took over Paramount, [00:14:45] Jeff Shell issued, we talked about it on the pod, issued a kill list- Marion Ranchet: Yeah Evan Shapiro: Of thousands of people. You're gonna start to see that play out. Before the end of this year, they're gonna announce major layoffs and yes they have some, they have $45 million [00:15:00] in retraining and displacement funds for the people who are gonna lose their job but that's- It's actually not a lot of money when you divide 45 million by 20,000. Yeah. [00:15:15] Yeah I think it's gonna be bad news for a lot of people. And, I was wrong about this but all during my reporting on this story, I was getting DMs from people who work at Warner Brothers Discovery and I'm still getting them [00:15:30] from Paramount, saying, "What should I do? How does this work?" Should I look for a new job? And, what are you thinking? And, also encouraging me to keep calling David Ellison a Nepo baby. So I- it'll be [00:15:45] interesting to watch and scary to watch a lot of our friends have to reexamine their careers, on the other side of this. And look, I think, I was thinking about this last night in prepping for this pod.[00:16:00] Consolidation is going to have to be part of the transformation of the media ecosystem. Marion Ranchet: Yeah. Evan Shapiro: But the transformation and consolidation and evolution of these companies would be substantially different if Apple bought Warner Brother [00:16:15] Discovery. If Google bought Warner Brothers Discovery, it, would be a different thing. Fewer people would lose their jobs. It would be a different company on the other side of it. My great fear here is [00:16:30] that they're combining the Titanic with the Lusitania and thinking it's gonna float. Marion Ranchet: No. Evan Shapiro: That's not gonna work. And so this is just the beginning. I don't think it's, I think that 20,000 is just the start, not the end.[00:16:45] Marion Ranchet: Can I ask you, 20,000, how many people would be left after that? I, don't know exactly. Still Evan Shapiro: About 80, still about 80,000 people. It's a massive company, between the two. And you [00:17:00] think also, remember, this is not just the US, Paramount owns Channel 5. Warner Brothers Discovery has offices all over the world. I think, honestly Marion, I think the, the biggest job cuts are gonna [00:17:15] come overseas. Yeah. 'cause they have to, they have these mandates here. They can't sell the lots. They can't negotiate the contracts together. There's a lot of separation of, these two companies mandated in this agreement and I think not the [00:17:30] majority, 'cause the majority of employees don't come from Europe but I think a, a big chunk, a disproportionate chunk of layoffs are gonna come from the operations overseas. Marion Ranchet: And how do you explain the fact that this feels like the staff is at the bottom of the list [00:17:45] of the demands from the AGs? 'Cause I think I know the way we would do it in Europe that would be the top things we'd be looking at. It would be top of the list. And alongside, clearly the, the CNN, [00:18:00] CBS staying independent and all of those things but, it would be top of the list. Anything that happens here in Europe, any type of merger, regardless of the ecosystem that's gonna be number one. I'm not saying we're gonna be trying to be saving [00:18:15] everyone at the same time because as we've said, as part of those mergers, there's always gonna be some job losses. But it wouldn't be at the, the bottom of the list with just a 10 million per year, fund [00:18:30] to put people back to work. Evan Shapiro: I think this is a real, a real exposure of the death of capitalism in the United States. Capitalism, as it was originally envisioned in the United States in [00:18:45] particular, was meant to take money from the top and put it back into the large population so that new competition could be created. And that's just not how it's working. Shareholder value is now the number one [00:19:00] priority, not just of the corporations but of the regulators overseeing them. Marion Ranchet: It feels, yeah. Evan Shapiro: And this deal is, I think, a tremendous example of that. To be fair, though, I believe that Bonta and Letitia James in [00:19:15] particular who were the two biggest states involved and the two biggest entities involved from the attorneys general side, I think they were more concentrated on the unions and the contractors involved in film production. So this is very protective [00:19:30] to the people who work on film and television who are not employees but are union members who contract per production. So this is meant to keep, to maintain levels of production, [00:19:45] to protect union contracts. So I do think that Bonta in particular was focused primarily because of the pressure he was getting from Karen Bass and Gavin Newsom, on film production and keeping [00:20:00] LA and California in the game. And in fact, there are triggers in this deal where that if new tax credits in California get approved, Paramount has to actually escalate some of its production and investments in the state of California and [00:20:15] it does prohibit them. The overall deal does prohibit them, especially if these tax credits get, get approved, from moving production out of the United States. And so this is, this was, they decided to side [00:20:30] on behalf of union and film production workers as opposed to employees of the company. And it's a, it's a, what, Sophie's choice to a certain extent. Marion Ranchet: Yeah. Evan Shapiro: It's who do you choose to [00:20:45] save? Where do you put the lifeboats here? And I understand why the AGs did that but I also know that if they had waited one week, one week until this ticking fee kicked in, [00:21:00] they would've been in a much stronger position. Marion Ranchet: Of course. Evan Shapiro: To negotiate even better deals. And a sale, a sale of New Line Cinema, a sale of Nickelodeon, a sale of BET [00:21:15] would've safeguarded substantially more jobs on the other side of this, than this frankly, I just think Faustian deal that that they've created here. So it, I think it's a [00:21:30] real shame that Bonta caved to pressure. And I don't know what he thinks he got politically out of this other than perhaps a promise to be attorney general if Gavin Newsom becomes president of the United States which [00:21:45] Gavin Newsom will not become. So I just, it does boggle the mind. And if you look, read the reporting, especially, in the Ankler about this deal, it does seem like [00:22:00] Connecticut, New Jersey, New York were like, "What the fuck are you doing? Why are you settling now?" And it, I agree, what the fuck? Marion Ranchet: Yeah. They had until the 11th hour. They had until early October.[00:22:15] Just like you would do, you guys start to love doing which is negotiating a carriage deal and then threatening of taking down channels at the last minute and et cetera. Everyone knows, this is where you get the [00:22:30] most out of every side that crunch time. Super interesting. So let's move on to another crisis, one that we actually saw coming in a way. You've been very vocal about predicting the AI [00:22:45] bubble bursting. I don't think you predicted the end of the world so let's talk about that CEO paradox. We, for a second you might think, "Oh, these guys grew a conscience." Of course they did not. Evan Shapiro: They absolutely did [00:23:00] not. The, you have to look at the timing, of all of this and how convenient, all this is. So let's back up just a second. On September 8th, this young engineer, named [00:23:15] Jacob Coxon sat on a park bench in San Francisco and, tweeted out, his first tweet ever by the way which is super weird for a Silicon Valley company. They live on Twitter and I refuse to call it X. [00:23:30] And so it's his first tweet ever and this is a guy who had worked at Anthropic for less than four months. Marion Ranchet: And he worked at, he was at OpenAI before. Evan Shapiro: Yeah for two years. And he was on the spear's edge of AI development so he was [00:23:45] training and doing the frontline development on these frontier models. And he tweeted out basically "These companies are not acting responsibly. They're marching towards recursive self-improvement," which is when the AI can teach [00:24:00] itself. Marion Ranchet: Yeah. Evan Shapiro: And they're doing so without any guardrails and we, people inside these companies believe that there's a non-zero chance that these models will destroy humanity." It's basically a summary of his tweet storm. [00:24:15] And first of all, he didn't say, "It is happening. I've seen proof." He said "We believe that it might someday." And then within 15 minutes This guy named Evan Herberger, I [00:24:30] think his name is, the head of safety at AI said, "Yeah Jacob's right. Definitely. Totally. Definitely. We all think there's a better than 10% chance that these models might [00:24:45] kill all of humanity within the next decade." \ And then right after that, Sam Altman and Elon Musk and a whole bunch of other people said, "Yeah definitely. Definitely. That's definitely what's happening." One, [00:25:00] there is zero proof that they are anywhere near recursive self-improvement and all you have to do is use AI yourself to see this, these models are just fast guessers. Yeah. They're not thinking. They're not [00:25:15] teaching themselves. And if they are teaching themselves, they're feeding on their own tail. Secondarily, the timing of this is all so fucking convenient. This is a marketing campaign disguised as a crisis. Marion Ranchet: Yeah. Evan Shapiro: OpenAI had to [00:25:30] delay its IPO, not for safety reasons, as Sam Altman said. By the way, Sam Altman, read the, The New Yorker article by Ronan Farrow, 16,000 words taking down, basically describing Sam Altman as the world's greatest [00:25:45] liar. Yeah. His board thinks this. His employees think this. His co-founders think this. They, he just, he lies as he breathes. And so the, the fact that OpenAI is saying they delayed their IPO for safety reasons is bullshit. They can't, they're out of cash and they're [00:26:00] having trouble raising money. Anthropic has an IPO coming up. So the concept here is this thing is so powerful it'll cure cancer and destroy humanity. You gotta get in now. You gotta get in now. It's gonna get so good. You gotta get in now. [00:26:15] It's a messi- Michelle Goldberg in The New York Times wrote this great article diagnosing the, this broligarchy in Silicon Valley, as I call it Nilicon Valley. And she said it's toxic [00:26:30] messiah complex. And so they're basically saying, "We are so powerful that we can both cure cancer and destroy the planet," but the safeguards are let us regulate ourselves. Yeah. That was all, Dario [00:26:45] Amodei issues this huge missive saying, "Slow down the pace of development here." Anthropic has an IPO. This is a marketing campaign for raising a $2 trillion valuation, the largest IPO in history, [00:27:00] within a couple weeks. Interestingly they just delayed their IPO. Marion Ranchet: Yeah. Evan Shapiro: Almost immediately after all this happened, right? Immediately, Xi is in the United States negotiating with [00:27:15] Trump on AI safety regulations. The reality of this is token, and this is not just me. This is Michael Burry, this is Steve Eisman. These are some of the biggest experts in financial projections and tech in particular. [00:27:30] They're all saying this is all bullshit. This is all a marketing campaign. This is puffery, as Michael Burry calls it. And they are nowhere close to this, this self-teaching model that they're claiming to be. [00:27:45] And this, they also are incredibly panicked about these free Chinese models coming in and basically bottoming out their business models. That's what's going on, [00:28:00] 100%. Marion Ranchet: Yeah it's, so a few, a few things. Very ironic to have US companies, asking for regulation which, very often, we're being, criticized here on this side of the ocean for [00:28:15] actually doing that in a responsible way. But what a beautiful marketing, Class A, actually what you were saying, Class A marketing campaign and just [00:28:30] for once, "Ooh, regulation could be useful. Let's, let's, do that." The one thing I will say- Evan Shapiro: But they're not, they're not asking, they're not asking for regulation. They're basically saying, "We should self-monitor. Let us..." this is stolen right from the banking [00:28:45] industry who then basically destroyed the economy last century. This is like, "we'll regulate ourselves." And by the way, the White House is helping. Trump yesterday, just yesterday, said, "We're renaming..." I can't [00:29:00] do a Trump impersonation. "We're renaming, artificial intelligence to super intelligence." Marion Ranchet: Yeah. Evan Shapiro: Why not? Marion Ranchet: But what's interesting is regardless of however you put it, whether that's regulatory [00:29:15] scrutiny, regulating themselves, the one thing is that when you have regulation, what are you doing? You know as a big company like those companies that you can actually continue investing and putting those guardrails or abiding by regulation, [00:29:30] whereas others may not, right? And by that, very often smaller companies, very often, open models, et cetera. And what you're seeing, is that you have these guys on one side which are proprietary, [00:29:45] just throwing money at this, like you were saying. At the end of the spectrum, you have the open web, the version of the open web, in AI. And right in the middle you have, you said China. It's not just that it's [00:30:00] free. It's in between being proprietary and can self-improve and it's open to a community, to the AI community, et cetera, for everyone's benefit. And it is cheaper to run. I wouldn't say that it's [00:30:15] necessarily free, per se but that's what they're worried about. Evan Shapiro: But it's free to the consumer. That's, and that's the ultimate challenge here, is that in, the... Before the dotcom bubble burst, there was this belief at [00:30:30] AOL and other companies that you would pay for email and search. We don't do that now. That's not our business model. And so currently in this bubble, there's this thinking, this belief and Meta just launched [00:30:45] Muse this week as part of this belief system and it's been very popular, by the way. But it is free, right? You, there's a pay model but no one's paying for it. There's this deep belief amongst these Self fart [00:31:00] sniffing broligarchs that they're gonna get people to pay for AI, individual humans to pay for AI. It's just not going to happen and China coming in with free models ultimately radically [00:31:15] commoditizes this product. Muse is a great example. So it launched, in the last few days. It's massively popular. It's now the number one downloaded app in all the app stores replacing OpenAI and ChatGPT, by the [00:31:30] way. But it involves, this is a company that just admitted in the largest settlement of child safety in history that it knowingly and by the way, the other tech companies who were involved in the AI race [00:31:45] also are complicit in this. They all are, have been proven to knowingly harm teenagers and children with their products. Marion Ranchet: Yeah. Evan Shapiro: Making them addictive so that you can't put them down. These same [00:32:00] companies, Muse, Meta, just released this thing that requires you to give their AI, their artificial intelligence, access to your email, your texts, your banking information, your credit [00:32:15] cards. Like, why would you do, why would you give Mark Zuckerberg access to your email? Yeah. Like, why would you do that? It's just so silly and then two things happened immediately after that. [00:32:30] Amazon banned it so you can no longer use Muse to shop on Amazon. And two, this name, this guy, this Apple expert named Patrick Wardle hacked it. He hacked it. He, [00:32:45] in one day, he hacked Muse and was able to place malware on a user's computer and hack into all of their banking information and all of their personal information. Now, it was, his own computer so he didn't do it to someone else. But in one [00:33:00] day and this is one guy. It's not this army of hackers who, by the way, so an army of hackers just hacked into the FBI and got access to all the personal information of everyone who works at the FBI and their spouses. So that's an army of hackers. [00:33:15] One guy hacked into Meta's massive magical AI in one day. These dudes have no fucking idea what they're doing. They are just racing towards a destination without [00:33:30] understanding how to get there and they're taking the entire world's economy with them. Marion Ranchet: Yeah. Evan Shapiro: So we've talked almost exclusively about American companies. I'm curious, how does Europe see the AI arms [00:33:45] race? Is Europe in the AI arms race? We've talked about Mistral before. I have to admit, I'm not as close, nor do I have as much visibility into the AI industry in Europe. Marion Ranchet: Yeah so [00:34:00] actually a a few weeks back, we had, Was it Scott Bessent, huh? He's the secretary of the treasury and he was being asked about AI in Europe and the guy said, what do [00:34:15] they do? What's happening in Europe?" And oh, my God, the A-hole, sorry I'm gonna say it on here. But he said, I'm gonna use the French words, rien. Nada, nothing. That's what he said about where we are when it [00:34:30] comes to AI. And I think there's a few things, right? Are we in that race when it comes to frontier models? The answer is no. Clearly that's what, Mistral was meant, to do. With, Le Chat they [00:34:45] were trying to do exactly what OpenAI and Tropic are doing which is to put, both a B2C and a B2B frontier model, you know, in the hands of Europeans and European companies. These last few [00:35:00] months, clearly they've shown that's not the game that they wanna play and they've actually pivoted and I think they're gonna be more focused on the enterprise side of AI, right? I haven't heard much on that side. [00:35:15] They're actually now working with DeepSeek and the Chinese models, et cetera. So it feels like they're positioning themselves not so much as a developer but more a, as a distributor of third-party models. [00:35:30] The thing they will focus on is how can we build the infrastructure in Europe? How can we build an enterprise business in Europe with that? So to me, it's, there's no interest in comparing OpenAI, Anthropic, [00:35:45] and Mistral on, on that side of thing. Does that mean that we do nothing? I don't think so. Otherwise, a company like Hugging Face which, what was just bought by NVIDIA. Hugging Face, it's an open, open [00:36:00] marketplace which is quite interesting to have this open web apply to AI moving from doing its own thing to coming to the biggest, chip manufacturer in the world. Let's see what NVIDIA will do with Hugging Face. [00:36:15] I think where we're gonna be playing a part and if you look at the global AI map, we are still active there. So whether that's, again, Mistral, you look at, Lovable, in Stockholm. You [00:36:30] look at Germany, you have, Alpha. So I think we're gonna be working on the discrete side of AI and not just those big names and those big frontier [00:36:45] models that perhaps, you know, you and I will be using on a daily basis. That means that just like with digital and Google and social media, we're gonna be renting your rails and we're gonna be building business on that, [00:37:00] which is... That's not great but I think, at the end of the day, we're simply not investing enough to be able to compete with what China is doing and what you guys are doing. Evan Shapiro: Yeah I think I'm curious, how do the [00:37:15] regulators in Europe, the EU or the various other regulators there, see this AI race? Do, do they, they decided not to get involved in the Paramount-Warner Brothers deal which I [00:37:30] found, that was another thing I predicted that I was wrong about. Yeah. I, thought that the EU in particular was gonna have a lot to say and they basically just punted. So I'm curious as to how regulators in Europe, are seeing this.[00:37:45] Marion Ranchet: So they've been proactive-ish so to say. So early August, we, the EU AI Act standard came into play. It means [00:38:00] that then it needs to be, it needs to go to... It's done at the European level and then it needs to be transposed in local legislation. But, essentially it's enforcing audits. It says that anything that is coming out needs [00:38:15] to be labeled. It's calling for transparency, et cetera. So that's actually stuff we're good at, I have to say. Yeah. Just like with social media, is that we... So we're not great at building those things. [00:38:30] But we're good at saying, "Okay great," "Let's continue to innovate, fantastic. But let's do it in a responsible way." So they are very much involved but at the same time, I think [00:38:45] they are feeling overwhelmed by how the narrative is already so much in the hands of US and Chinese companies. So what can we, what can we really do is a big, it's a bit of [00:39:00] a question mark. And so you have Mistral, you have a lot of companies who are calling, Europe to, just get to it essentially. And not just from a regulation standpoint but just foster innovation so that we're not [00:39:15] seeing, 'cause a lot of those companies and those great minds that you're seeing in these companies, very often those are homegrown folks, right? So Hugging Face, three French guys but the company has been headquartered in the US for many [00:39:30] years. And now it belongs to NVIDIA. Even Mistral, it's still in France, it's still a European company but up until now, a lot of the money that went into Mistral was foreign money. So [00:39:45] only this last round that they announced a couple of weeks ago which is, I think it's 3 billion which is the biggest but in comparison to Anthropic where I think we're talking about the difference, OpenAI, Anthropic, the valuation. So Mistral, [00:40:00] 21 billion Valuation Evan Shapiro: Yeah Anthropic is raising money at a $2 trillion valuation. Yeah. OpenAI is raising money at a $1.5 trillion valuation. SpaceX is worth $2 trillion. Marion Ranchet: Yeah we're dwarfs. But the good thing is that [00:40:15] we're finally seeing money coming, being poured into that company from Europe. I think 50%... This, this round was led by Samsung so Korea. But I think half of the money is actually coming from Europe. So we have [00:40:30] to push for transparency. We're good at pushing regulation. Let's do it in a way that doesn't stifle innovation and funding is, the only thing that I have. If we're only trusting you guys for that, I mean, I can [00:40:45] see a scenario, a doomsday scenario in AI if only China and the US are at the wheel. So we do have a role to play. But I don't want it to be just a referee role. We need to be more active than that. Evan Shapiro: [00:41:00] So I wanna talk about two things 'cause I, I get it from both sides which is the Luddite side that doesn't wanna touch AI and just basically ignores it all and then the super AI nerds who are like "This is the most powerful tech in the history of the world. It's gonna [00:41:15] change your molecular makeup." Both sides, wake the fuck up. If you're not using AI and you're not testing and learning on it, you're gonna get left behind. You're not gonna understand what it can do and what it [00:41:30] can't do and that's really important. Secondarily, on, to the people who believe this is magical dust from outer space that's gonna change everything and either destroy the planet or create an absolute utopia, you're fucking out of your [00:41:45] mind. I'm gonna use Jensen Huang who is the CEO of NVIDIA, the most powerful company in the AI arms race who just bought Hugging Face, as you said, the company, by the way that was supposedly [00:42:00] hacked by by the rogue AI. He was on Ezra Klein's podcast this week and what he said was, "This is all bullshit. This all hype around AI's destructive apocalyptic [00:42:15] power is bullshit. And Klein said, no these agents are breaking out of their sandbox and hacking other sites." He's like, "That's happened before." Marion Ranchet: Yeah. Evan Shapiro: It's software. That happens a lot. Software leaves its [00:42:30] sandbox all the time. We've just never talked about it before because we've never had a reason to talk about it before. So this is... He, he basically said, "Stop scaring people. It's just software. It's just, and by the way, it's not thinking, [00:42:45] it's guessing. It's a calculator that anticipates the next question." Yeah. That's all it is right now. It does not have this recursive self-improvement power that everybody's, scaring us about. The second part of it is and I hate to [00:43:00] promote China as a model but regulators in China have their finger on the plug, so if a rogue agent misbehaves, boink, they shut it down. Yeah. They turn it off. And [00:43:15] we have an FDA in the United States. In Europe, you have tons of regulatory bodies that can prevent food from being distributed and prevent and do recalls of automobiles and other dangerous products. [00:43:30] Why would we not have the same kind of regulatory protections around what is this super destructive Bond villain laser that could kill us all overnight? I wanna remind everybody that in every Bond movie, [00:43:45] the villain thinks their laser is gonna give them ultimate superpowers and it never does. And I know that's fiction but these guys have, they believe their own hype or they want us to believe their own hype for their own [00:44:00] financial gains. AI is neither this super powerful destructive laser that's gonna kill us all and it's also not some dumb software that's gonna go away. Both of those things happen to be true and if we had an actual [00:44:15] regulatory body in DC right now, we could actually come up with something to protect ourselves in the off chance that this stuff does go berserk and start to kill everybody. The other last thing I'll say is [00:44:30] it's not as good as people say it is. Marion Ranchet: No. Evan Shapiro: It is not as efficient as people think it is. It is substantially more expensive. The costs around AI are so much greater. That's why these companies are in such [00:44:45] desperate need of this hype marketing campaign, is they're burning cash at astronomical levels. There's trillions of dollars of money we don't even have on the books for leases and builds and other things. It's just this [00:45:00] money pit. And so, there's a actual financial incentive for them to fool us. But these companies who do switch over from human workflows to AI workflows, almost all of them are regretting it. It costs more. It's not as [00:45:15] efficient. The employees that they do keep are spending way more time chasing hallucinations and correcting mistakes that AI makes than they are actually doing the work that they're there to do. On the flip side, [00:45:30] 55% of companies who have made major layoffs in favor of AI workflows regret it and are now rehiring those same employees back. So this hype bubble that we're living in right now is just that and it's gonna pop relatively [00:45:45] soon when, not if when the Anthropic IPO doesn't go exactly as planned when OpenAI can't raise that money at a $1.5 trillion valuation when 700 million [00:46:00] shares of SpaceX from insiders get unlocked and flood the market and pull the rug out from SpaceX's valuation, this bubble will pop and it's gonna hurt the entire economy. It's gonna cost people jobs. And [00:46:15] so I hate to say this phrase out loud but do your own research. Use the products. Get, read the experts, not the hype machine and you will find yourself in a place w- where you can sleep at [00:46:30] night and not have to worry about the apocalypse happening in the morning. Marion Ranchet: So regarding the hype machine, I actually read that 70% of companies say that they use AI but actually only 20% are [00:46:45] actually shipping stuff that work. And so I think this, this hype, this FOMO makes it that companies are just taking AI without thinking twice, without doing, to your point, the homework and right [00:47:00] now it's producing subpar results. Having said that, I don't like to generalize and when I was at IBC, I was on a panel with Gracenote, Telia which is a Nordic telco and Plex. And most of them were saying that it [00:47:15] was, some were saying it's, evolutionary. Others were saying that AI was revolutionary. Among Telia and Plex, they were saying that right now, like 20 to 40%, of, of [00:47:30] AI was being used to do things that they had never done before and that they expect actually to have AI be pregnant everywhere within the organization. And I think it's the same way as the digital [00:47:45] transformation 20, 30 years ago. So we're gonna be embracing it. We're gonna take what works. What doesn't, we're gonna cast aside. The only challenge, the only thing that I will say which I think worries me in terms of the sustainability of this ecosystem and you were saying that [00:48:00] it would be bursting, is that right now they're putting very powerful, not destruction powerful type of AI but very powerful models, for 20 bucks a month. So it is expensive on them but it's not expensive on us. [00:48:15] Where I'm worried is that we're gonna be over reliant and there's a moment in time where they're just gonna cut it and we're gonna end up with the dumbest AI possible for 20 dollars a month and we will have lost the ability to do a certain set of things. You [00:48:30] already have results on the impact of AI on students where you have the school year results going up, up and then the exams, the results are going down. The fact that kids are not, developing critical [00:48:45] thinking, et cetera that you're, they're losing the habit. Evan Shapiro: We're outsourcing- Marion Ranchet: To be able to think. Evan Shapiro: Yeah we're outsourcing intellect at this point. Marion Ranchet: I like that, a lot of those dangers and that's why it works, this marketing campaign from these guys, is that we always see the first [00:49:00] negative impacts across the chain, across the ecosystem, yeah. Evan Shapiro: I think the way I describe the evolution that's gonna take place is and this is gonna only be for people our [00:49:15] age, your age and up basically- Marion Ranchet: Okay Evan Shapiro: is that it's the evolution from the fax machine to email. Marion Ranchet: Oh. Evan Shapiro: Remember fax machines? Marion Ranchet: Yeah, I do. Evan Shapiro: And how kind of inefficient those [00:49:30] machines were. They were much more efficient than mail but not nearly as efficient as text messages and email. And then email kind of morphed into this workflow [00:49:45] pylon, including Slack. When you talk to people about Slack, it's like talking to them about their dysfunctional spouse. Marion Ranchet: No, I hate it. Evan Shapiro: They hate their relationship with Slack. So it, there are numerous elements to this which is [00:50:00] the AI we're using now will be ultimately commoditized. My team uses AI. We use Gemini. But it comes inside our Google Suite of services, which gives us my cloud storage. It gives [00:50:15] us our email infrastructure, gives us all of our kind of business software infrastructure. But we don't pay specifically for AI. That's the kind of commoditized nature that AI's gonna wind up being, [00:50:30] and the things that we're gonna use AI for will be organizational. It will be for search. It will be for things that kind of humans can't do at great scale. It's not gonna be for a lot of the things. Actually, solving [00:50:45] gene puzzles, things in medicine, those are really great. Discovery of content. These are great uses of AI but this idea that it's gonna upend society completely. And you're right we are outsourcing intellect. If you're [00:51:00] using AI instead of thinking, you're gonna get crushed. Your competition will kill you. It's the people who understand the difference between artificial intelligence and actual intelligence and who know how to combine those two things, those are the people [00:51:15] who are going to survive and thrive. My team is producing more work because of AI. Marion Ranchet: Yeah. Evan Shapiro: It better work? Sure. It's better work because of the humans, not because of the AI. [00:51:30] Some would say that it's worse work before it becomes better work. Marion Ranchet: Yeah I agree. Evan Shapiro: We have to ask AI 12 times to prove its own work and we then have to triple check it before we can release anything because it's almost always wrong first.[00:51:45] Marion Ranchet: I see it as an augmenter. That's always what I said. And it allows me to do things that I could have never done on my own and a small example but for example, I hate Excel. Business plan, doing that, between Excel and Claude, loving it. [00:52:00] Or I have, no coding knowledge. I'm not a graphic designer. Evan Shapiro: Me either, yeah. Marion Ranchet: So I had a landing page that was done by a designer. As time went, up until SME Live, I needed to tweak stuff. [00:52:15] Instead of having to do every single thing manually, just instructing the AI to do it. So yes it's tedious, it takes time and it never works at once but I have this landing page and I saved money and time on making these [00:52:30] things evolving. Speaking of learning how to use it, next week we're gonna have Emily Horgan. She's a friend of ours. It's long overdue actually. And she's gonna be coming on the pod to talk [00:52:45] about kids media. No we're not gonna be talking about AI but we're gonna talking, we're gonna be talking about that generation that we need to protect 'cause those are the adults of the future and she'll be coming with a lot of data and insights because she's actually studying the [00:53:00] engagement reports from Netflix like no one else. Evan Shapiro: If you're paying attention, Netflix actually got downgraded this week by Wells Fargo and HSBC because it's losing ground to YouTube and engagement is decreasing. And so Netflix is now, if [00:53:15] you pay attention to their market valuation, half of what it was about a year ago. So I think Emily's engagement metrics and engagement study is really crucial if you're paying attention to the media ecosystem. Marion Ranchet: That's it for another episode of the Media Odyssey [00:53:30] podcast. That was Evan Shapiro. Evan Shapiro: That is Marion Ranchet. We'll see you soon.