← Big Technology Podcast
Big Technology Podcast
Polymarket vs. Pollsters, Podcasts & Politics, Open Source AI In Trouble
00:49:51
Loading summary
Transcript92 lines
- [00:00]
Alex
Poly Market outshines Nate Silver and the pollsters. Bitcoin and The S&P 500 jump to astounding new heights. Podcasts become the breakout hit of the election cycle. Plus, open source AI is in trouble. All that and more coming up on A Big Technology Podcast Friday Edition right after this.
- [00:18]
Michael Kovnat
Hey, I'm Michael Kovnat, host of the Next Big Idea Daily. The show is a masterclass in better living from some of the smartest writers around. Every morning, Monday through Friday, we'll serve up a quick 10 minute lesson on how to strengthen your relationships, supercharge your creativity, boost your productivity and more. Follow the Next Big Idea Daily wherever you get your podcasts.
- [00:42]
Tomer Cohen
I'm Tomer Cohen, LinkedIn's chief product officer. In my new podcast, Building One, I interview some of the best product builders out there, people at the intersection of dreaming and building and learning. Together, you and I will learn from their experiences. If you're just as curious as I am, follow Building One wherever you listen and check out the conversation on LinkedIn.
- [01:03]
Alex
Welcome to Big Technology Podcast Friday Edition, where we break down the news in our traditional cool headed and nuanced format. We have so much to discuss with you today, a lot going on in the aftermath of the election. We're going to talk about how well the prediction markets did, why bitcoin and The S&P 500 are just soaring, the role of podcasts in the election, and the fact that China is now taking open source AI and using it for military potential military uses, and what that means for the new administration. Joining us as always, is Ranjan Roy of Margins. Ranjan, great to see you. Welcome back.
- [01:36]
Ranjan Roy
It's been quite a week, Alex. I'm, I'm ready to talk about it.
- [01:40]
Alex
It's a week where we had a tremendous amount of news break over a very short time and there's so much to unpack. And you know, we're going to get into it. We're still talking election here. I know we did some last week. I've done a couple shows this week and this is the third that's going to touch on the election. But I just heard a piece in Big Technology talking about how the AI story is going to roll on in the Trump second term. And I just want to put a point on that before we get going, which is that yes, there are going to be a lot of Trump stories over the next four years, but our role here is to break down tech news. We're going to do, of course, some politics as we always do, but the focus is going to be on artificial intelligence and technology and that is going to be the place that we cover. So don't worry. This is not going to turn into a politics podcast. But that being said, we should talk about what to me was the most interesting tech story of the election, which is the fact that Poly Market just kicked the living daylights out of the pollsters. Poly Market, like we spoke about last week, had Donald Trump favored 61% to 39% for Harris. It called almost all the races accurately and did it before any of the news organizations did it. Nate Silver, you know, had Harris and Trump in a toss up 50, 50. He ran like what, 80,000 simulations and it was like 40,000 each. Poly Market ran one prediction market and it was right. And so you talked a little bit about what the implications are last week, but now we have results. What do you think we should take from this?
- [03:05]
Ranjan Roy
Oh, this was a tough one for me, Alex, because I have long believed in the power of betting markets, having come from the trading world. And there was even a site, Trade Sports back in the 2000, if anyone actually remembers it, where you can not only essentially bet these derivative contracts on sporting events, but they had entered politics actually back in 2008. So I really believe at some kind of scale and quality of market, it's a better way of trying to kind of like glean information and predictions. I was a little frustrated and annoyed by both Poly Market and Kelshi during this election cycle because they were vindicated in this cycle. But these are still incredibly small, manipulatable and opaque markets. So I think the idea that somehow they are now positioned to give kick the living shit out of pollsters and like, you know, really give us better information. I think in this exact situation, the Republicans won, Trump won. And the type of consumer that would probably be more geared to our betting market, especially Poly Market, which is only technically available to non US citizens and is only available by using crypto and understand, I mean clearly the crypto market and most crypto users would have leaned Trump fairly this election anyways. I think this was more like correlation, not causation. It just happens to be in this specific instance, the actual electorate went in the same direction that the people who would use a betting market went. I don't think that's going to be the case going into the future or as like a reliable indicator of future events.
- [04:49]
Alex
All right, I'm going to take the other side. I think that this is from Coindesk. It says the vindication comes after several weeks in which the mainstream media harped on the theory that Poly Market, which runs on crypto and has seen billions of dollars in trading volume this year, was being manipulated by pro Trump forces to inflate his odds. I think they're speaking directly to you.
- [05:10]
Ranjan Roy
Are you quote, are you quoting CoinDesk for saying crypto?
- [05:15]
Alex
They're directly to you. And here's the facts. The facts are that the pollsters were wrong. The pollster said 5050 and Poly Market was right. And not only was it right, it became such an effective data point because as the results started coming in, the movement that you could see in Poly Market was so predictive of the outcome. It started with that 61% as the early votes came in, attract perfectly 66%. And eventually it got to the point where if you do any betting at all, you knew that it was good that the odds on favorite was Trump before any of the swing states were called by the networks. And in fact, I was in my group chats as the night went on and I called the race at 7:30pm and I was like, listen. I was like, I'm looking at what's coming in with the exit polls. That's a data point. I'm looking at some of the early votes. That's a data point. And I'm looking at Poly Market, and that's an important data point. And both polymarket and Kalshi were making massive moves in Trump's way. And I think you're, you're underestimating how Sharp's right. Skilled bettors have come into this field and moved these betting lines and have been quite accurate on them.
- [06:26]
Ranjan Roy
See, where I disagree here is again, when we have an event where the outcome is something that is not geared towards the type of person in a betting market, then I want to see it actually work. Because when it's not the bias of someone who would participate in a betting market, when the, when the outcome is in that direction, then we'll actually see. And I think in this case it's just. I don't know it, it. I want betting markets to work. I want all of us, I think we, I forget what the bet we made last week, but we made a bet in a year.
- [06:59]
Alex
Wait, I already forgot it. I guess I'm off the hook on that one. Oh, but it's recorded. That's the problem with doing podcasts. Recording so.
- [07:10]
Ranjan Roy
Exactly. But if we had a contract, a derivative contract that allowed us to place odds and let people bet, we would be held accountable. I think we made a bet around whether AI search would replace over 50% of search. That's. That's another one. I want all of these as derivative contracts. That's what like, crypto was supposed to allow us to do. I really think in the long run this could be good. I think this is another example again, of the way it's being rolled out and kind of turned into. I was getting endless. I was getting Cali ads that were like, really like standoffish against the mainstream media and that whole positioning. And I think again, it's attracting a very, very specific type of person. And in this case, that person reflected the entire electorate. I don't think that's going to be the case two years from now, four years from now.
- [08:01]
Alex
I disagree. I mean, they got so many of the states right. Like, if they only had like a decent or directional accuracy, I would say, okay, maybe that makes sense. But they got so many things right and they called so many states before Everybody at the AP called them. For instance, they called Georgia at 10:27pm and the AP called it at 12:50, 58am they called Wisconsin at 11:10pm AP called it at 5:34am they called the overall winner at 11:43pm we didn't have the AP call till the next morning. So I think the fact that they were both accurate and precise says that the wisdom of the crowds was right in this. In this area. Now, do I leave open the possibility that you're correct in some way? Sure. But I do think this is the new standard until proven otherwise.
- [08:48]
Ranjan Roy
Well, no, no precision in this case. If they were able to actually kind of call the different. The vote differentials in these states, that's one thing. When Trump wins seven of seven swing states and the bias of a market like would be towards Trump, then it's going to be correct. But if it's actually calling like these will be this narrowed, then I would be impressed. Again, what some kind of decision that would lean heavily, almost like comically progressive. I want to see that accurately called by a betting market. Again, I think these companies could actually really kind of build good markets around this. I don't think they're good markets right now. And I'm saying that purely from a market infrastructure standpoint. They're still very small, they're still very opaque. They're more trading like an emerging market than an S P. If they could actually develop in a kind of good way, I'd be happy. I think these can actually be and should be kind of like more at the center of how decision making goes. But again, I've been, I've been following this stuff for the last 20 odd years and we've been promised this in a lot of different iterations and it has not worked.
- [10:04]
Alex
I don't know. I think the proof is in the pudding. Nate Silver, 80,000 simulations, 40,000 one way, 40,000 the other way. I'm sure market forecast October 25th predicted the actual results perfectly.
- [10:18]
Ranjan Roy
My favorite thing is I'm sure Nate Silver would be able to justify why I think it was 40,012 times Harris1. I think somehow he would actually still be able to say he was right.
- [10:32]
Alex
Well, that's Nate's. One of Nate Silver's superpowers is explaining how you've misinterpreted how wrong he is. Yeah. Now I do think there's actually some legitimacy to that. I do think people pile on to Nate and by the way, Nate is, he wrote about, we talked about on the podcast with him. He wrote in his book on the Edge he follows these prediction markets quite closely and we talked about like whether they're going to replace polls at a certain point. And I think that like, you know, they do until proven otherwise. Let's see. But they do until proven otherwise.
- [11:01]
Ranjan Roy
Well, just a quick personal story. During the whole like peak, I think it was probably like 6:30pm I tried to deposit money into Kalshi and it said via debit card it would allow you an instant deposit. And I was not able to. It kept giving me server errors and erroring out. And I think that kind of reinforced my view on this that on face it looks like this like really modern, sophisticated thing. But in the back end it's still just kind of a hacky startup that is promising this one big thing. But it's not even functional enough to allow me to deposit money as it's supposed to to try to actually bet during a peak time.
- [11:43]
Alex
And even with those problems, it's still so accurate. By the way, who are you going to put that money on? I saw you tweet your screenshot. Who are you planning to bet on?
- [11:50]
Ranjan Roy
Well, no, as a trader and I would have lost on this. I think it was down at 7 and before any swing states were announced. And that's where again from like an expected value standpoint betting 71 swing state flipped, you could see that go up to 20 or 30. That would be. And I would have sold it right away. I thought earlier on I did not think she was going to come out on top. That would have been which is what I want these betting markets to allow me to do just a very short term kind of flip on some kind of event happening. And that event in that case would have been one swing state early on would have gone from 7 to 20 or so. Was the expectation well bet 7 to make 13 to 20 or so.
- [12:34]
Alex
Well, I guess it's a good thing that it didn't take your money. So you know, I joined FanDuel recently. It's really very addicting, I have to say. We've talked about sports betting in the past. I was like think looking at Poly Market, I was like can I drop a wager on FanDuel for this? But you can't, but I'm sure you'll be able to at some point.
- [12:53]
Ranjan Roy
I like how, just as I described how my mind works, I was actually feeling almost terrible for all the discussion we have had about the ills of sports gambling and literally trying to like juice a little bit of return on a one hour period Election day on the presidential.
- [13:13]
Alex
It just goes to show you how powerful these platforms are.
- [13:16]
Ranjan Roy
They can be, they can be.
- [13:18]
Alex
Okay. So you used a term expected value, which is something that is used often in the crypto community. And to me like one of the interesting things is like a lot of the coverage of these markets are on CoinDesk. The markets take crypto. They're run by crypto people. And the people that were taking their victory laps were also crypto people. Here's one quote from the CoinDesk article. Prediction markets work, said Hart Lamber, the founder of a co founder of a UMA Protocol, a decentralized oracle service that referees outcomes and disputes for Poly Market. Man, I feel bad for the reporter who had to write that description and title but it's just so interesting that the ones that, that are so in favor of these markets are so pro crypto. I sort of get the association, but could you explain it to me?
- [14:09]
Ranjan Roy
Well, it's currently Poly Market because it is illegal in the United States for US citizens and to you as the US Dollar, it's going to be crypto. And who has crypto in their wallet that they can actually invest in? This is someone who is into crypto. So I think that kind of is, I mean it's a pretty direct connection right there. Then you have again, I mean Trump has been and he really leaned in towards the end as like incredibly pro crypto. We'll see how that plays out. But I think at all levels of that this was a very strongly crypto thing.
- [14:46]
Alex
Yes. And it has to be more than just a default. Right. I just Think that this is a way that folks who are deeply enmeshed in crypto think, right, where they think about expected value, they think about exactly the way that you were talking about putting money on something and hoping for a swing and seeing, you know, how every event could be quantified. And so it's sort of natural for them. That's my thesis.
- [15:07]
Ranjan Roy
Yeah. Well, I mean, and also in this case, if you are a crypto owner, betting on Trump, it's almost kind of like a double your money sort of play because you both are expecting to make money on that individual contract itself. And you know that crypto will spike if he is elected.
- [15:27]
Alex
And on that note, I mean, bitcoin sitting right now, it's crazy. All time high, $76,500. And the entire crypto community, like crypto was initially, I mean, Trump was initially anti crypto. Now he's pro crypto. This is from the New York Times. It's safe to bet that in the second Trump administration, the crypto industry will get most of what it wants in Washington, starting with the removal of Gary Gensler, the head of the securities and Exchange Commission, who has become a villain among crypto companies for his tough regulation efforts. Crypto companies that were being sued or investigated under Bide, under the Biden administration may see those cases dropped under a Trump administration. And pro crypto voices are likely to hold sway when it comes to right, new rules for the industry. I still, I get that, but I still. There has to be like an intrinsic value to bitcoin to be able to put all this money. And, and to me, seeing bitcoin sitting at 76,000 sparks two questions for me. One is how and two is why did I not get in when it was like 17,000 in 2022?
- [16:30]
Ranjan Roy
I mean, I feel on the intrinsic value conversation, that's a totally separate one, purely as a vehicle for speculation. This was a great bet. I mean, it was a great bet that, like, if it's trading on vibes, Trump winning the vibes all clearly point towards a less stringent regulatory environment. Gary Gensler is definitely going to be gone. It'll be very interesting to see who he appoints on the cftc. And I think overall, yeah, I mean, it's the most logical out of all the things that have happened. Bitcoin going up fast on a Trump victory. Not a lot makes sense in the world to me all the time, but that one makes pretty good sense.
- [17:16]
Alex
You know, we began our discussions on this podcast talking about the frothy economic environment and why? Nothing really made sense in the business world. It was, everything was going on from, you know, SPACs and crypto and SBF. And there was a moment of correction where rates went up, SBF went to jail, CZ just came out of jail and a lot of the crypto enthusiasm cooled. NFTs effectively went away. Well, bitcoin is now at an all time high. The S P500 is now at an all time high. Rates are coming down, baby. We just saw another rate go down. Rate decrease this week, of course, a quarter point. But are we returning to the frothy good old days of 2020 or are we still effectively going to be removed from that, that period of time?
- [18:01]
Ranjan Roy
I think I have argued over the last, let's say year, year and a half, especially any AI driven, like boost in a stock price, that it actually was a logical, not irrational thing. I think we're going to see, at least in the short term, that injection of life in a just a kind of untethered way, let's say. Because, I mean, just again, edits like on face markets are happy that Trump won. It's like the average investor, the average person on Wall street, like assumptions will be less regulation, like friendlier business environment. And that again, that one makes sense. I think how long that lasts is going to be the very, very big question. But we'll certainly see it over the next couple of months.
- [18:51]
Alex
I mean, I'm quite happy with the way that my portfolio has performed this week. It's crazy. The S&P 500 just jumped over 6,000 points for the first time. I mean, I remember sitting at the desk at CNBC middle of the year and there was an analyst that came on and they're like, we think it's going to be at 5200 or 5500. And my eyes were like, you know, coming out of my skull thinking about the calculation. It's now up 26% on the year. It's wild.
- [19:21]
Ranjan Roy
Yeah. I mean, but I think, okay, so I think Tesla is a good example for where this is going. And Tesla is up 26% from the moment the election was called the next day, or from the open 26% in just a couple of days. The market, the market cap just passed a trillion. What's so fascinating about Tesla to me in this context is it kind of, I think, is going to capture what the Trump presidency will, how it will interact with markets. Because why is Tesla up? Because Elon Musk helped Donald Trump win. I mean, like very closely associated himself with him. And the assumption is his proximity to Trump means something good is going to happen for the company. Nothing has changed about the fundamentals of Tesla in the last few days. In fact, Trump oddly enough, goes on anti EV electric vehicle rants all the time. Like, you know, at his core he is a fan of electric vehicles, yet Somehow Tesla's up 26%. The moment there is a Musk Trump falling out, which I'm going to predict and maybe there could be a contract on polymarket on this is probably going to happen by March or April because we've all seen this story before. Does the stock get hit at that point? Like the thing I kept thinking about was the early days of the Trump presidency, those massive moves in stocks from just a tweet or just like a speculation on his relationship with an individual CEO or other business leader. And we're definitely heading back there. So I think the question is which companies are the beneficiary of that and which companies are the victim of that. We're going to be heading back to that world probably now.
- [21:07]
Alex
I mean, Tesla has always been a story stock. Right. It's as much you believe in Elon Musk's vision and the power of that individual than you believe in like the bottom line for the vehicle production.
- [21:19]
Ranjan Roy
Yeah, but this is the most kind of egregious expression of that where it's like in the other times, it's like, yes, if Elon Musk says they're going to have autonomous robots and cars and whatever else and I mean it's still related to the company. This is simply trading on Musk's proximity to Trump. Someone again who is not a fan of the actual companies product.
- [21:46]
Alex
So that's why like, wait a second, he's also the man, I mean there are worse allies to have than the President of the United States. Right. And he's also the man, Trump is also the man that's going to be setting the tariffs. No, no, that we're already that the biggest competition that Elon has, electric vehicles from China.
- [22:02]
Ranjan Roy
Yeah, yeah.
- [22:03]
Alex
He's going to put 100% tariff on electric vehicles from China and Trump is the tariff man. So does that go to 2 or 300? Does that completely eliminate the possibility of Chinese vehicles entering the United States market? That's why, you know, the, the story would be pro Elon here and Protestant that.
- [22:19]
Ranjan Roy
Okay, that, that is actually one of the more logical kind of like paths to why that could go up. I'm impressed. I think, yeah, no, I think that's, it's a Good story. Up 26%, passing a trillion and back to above a trillion again on this. I don't know. I still feel it's gonna be really interesting. I'm still calling Musk and Trump, those two in the same room trying to work together. It just, to me again, we saw this over and over in the first term. Like it's almost. I. This one thing I'm kind of excited to watch play out too. Like it's gonna be dramatic and fascinating when it does. The only question is when.
- [23:02]
Alex
The counter argument there is that Musk isn't going to be in the administration. I mean, maybe he'll be able to run this Department of Governmental Efficiency and, and they need each other in a way. So. Yeah, I don't know. We'll find out. They are two big personalities. I got wrong.
- [23:19]
Ranjan Roy
I said Musk tweet. This like image of it was like at the bottom like some meme of a, like a large strong guy holding up two other people and Trump was like kind of the small one on top.
- [23:31]
Alex
I was surprised that the bottom was Elon. In the middle was Joe Rogan.
- [23:35]
Ranjan Roy
Yeah, yeah, yeah, yeah, exactly. Yeah. You saw it. So that one, I mean, already that's the kind of thing that just one of those could totally trigger Trump and end this whole relationship very quickly because now he doesn't need him as much anymore.
- [23:49]
Alex
Okay, well, you know, we could speculate all day about when or if that's going to happen, but let's talk like outside of the situation going on with the election, just the state of the market in general. I mean, Dan Promack was on, on Wednesday and basically said the S P has been booming for a year and now the S P500 is at a place that I don't know even the most pro this moment are going to admit that it's a little bit nuts. I think so at least. But let's hear from one of the bears. There's a. One of the Wall Street's last remaining bears warned on Thursday. The S&P 500 is approaching a mania level as stocks become more expensive. Right. Because there's a, there is a valuation thing here. The Bears, Barry Bannister, he said even allowing for the best case scenario of a US soft landing and despite the potential ramp higher for US fiscal spending and a geopolitical reckoning, the S&P 500 is in a mania nearing a three generation valuation high. This is your conductor that the train is approaching crazy town. I mean, I think these valuations will matter at some point. Don't you think?
- [25:01]
Ranjan Roy
I don't think any time in the very near future they will. I think until we see again, as you started this segment, rates are still being cut while the S P is exploding. So the expectations are, I mean, pure Goldilocks right now. So I don't see any like sober rationality entering the market again in the next few months. Again at some point, yes, but not anytime soon. I don't see what the catalyst would be.
- [25:30]
Alex
Yeah, this is from a financial account bar chart. The stock market has just hit the most overvalued level since the peak of the dot com bubble. Anything to read into that?
- [25:41]
Ranjan Roy
I mean it was there, it was pretty close, I think a year and a half ago, two years ago during the main. The mania of 2021 into early 2022. So it couldn't go higher. I'm not going to say it will or it won't, but it certainly can go higher.
- [26:01]
Alex
Yeah, I mean nothing has, has really stopped it this year. Okay, so we have in our document two bar charts that are just hockey stick bar charts up and to the right. One is the s and P500 and the other is the percentage of people in the US at least that listen to podcasts. And I was pretty taken aback by this because the number is so high. The share of the US population that have listened to a podcast in the past month is now 47%. And this is from the Wall Street Journal. 54% of podcast listeners say getting news or political analysis is an important benefit of the medium. And 60% of people who are under 35 have listened to a podcast in the past month. That was in the 40% range in 2020 during. In 2021 during the lockdowns. It dropped back into the 30s after that as people went back. And now our medium, Ranjan is having a moment. 47% of people listen to these things. I mean, we know it's just an increasingly important medium. And it basically defined the media strategy for Donald Trump during the election. You saw Trump, I think he was on 20 podcasts. He went on Lex Friedman as listening a show the listeners of this show might be familiar with, but also the Nelt boys. Theo Vaughn, he sat down with Joe Rogan for three hours. He was everywhere. And he ended up outperforming with young people. I think something like six points compared to where he was when he ran against Biden. So I'm curious what you would make of this. It obviously seems like podcasts have asserted themselves as one of the most dominant forms of media in the United States and I think in the world. And look, I'm obviously on the podcast train. I think that this is a very special medium. It's a way that, like, me, you and the listeners feel like we're like together in a room, really. And you can't replicate that anywhere else. I mean, maybe on talk radio, but talk radio had its own format, right, which was really broadcast. And this is more longer and more nuanced, more conversational, and it's something that Trump took advantage of. So what's your takeaway here?
- [28:14]
Ranjan Roy
Yeah, I've been thinking a lot and have plenty of thoughts on just the overall information ecosystem, media ecosystem, and how this all played out. And one of the things that really struck me on this, it's what you said. It's podcasts, even a three hour Joe Rogan podcast are about, is like intimate. A conversation to really get into who is a person versus any other medium. A convention speech, a cable news interview, a tweet, or an Instagram post. I mean, this gets you feeling closer to a candidate or a person far more than almost every other medium. And then the thing I kept thinking about is speaking even with friends who are working as part of the overall Harris organization. And then everyone kept kind of talking up the get out the vote operation. You know, like knocking on doors, canvassing, making calls, Small, small budget fundraising. Like, like, everything was the traditional view of how operational politics worked over the last few decades. And that's one of the things that I was like, okay, well, like Obama was famous for in 28, 2008 and 2012 of, like, having an incredible get out the vote operation. It was thinking about how effective Trump on all these podcasts was. And you saw numbers like Joe Rogan. I think it was like 50 million views of the Trump interview.
- [29:45]
Alex
Just on YouTube.
- [29:46]
Ranjan Roy
Just on YouTube.
- [29:47]
Alex
Remember, he's doing views on Spotify and he's across all podcast apps now.
- [29:50]
Ranjan Roy
Like, then just think about. Imagine you're sitting there, you just spent three hours with Joe Rogan, and first Trump, then Elon Musk, all talking about this, and then some random person knocks at your door and tries to say, like, hey, do you want to know about the election? Like, how that was ever effective, I'm not quite sure. But especially today, if you want to reach people, it almost seems, again, there's a lot of, like, in hindsight, this was obvious, but in hindsight, this was obvious.
- [30:24]
Alex
Yeah, no, it's, it's. I think it's a watershed moment for the podcast medium. That's why if you, you know, if you have a top executive at some one on the big tech companies, one of the big tech companies, you should bring them onto big technology. Podcast, this stuff matters. But I do think that there's. This was. I think everybody really understood this implicitly for a long time, but this is something that you cannot deny anymore. And I, I just, you know, I'm familiar with the Nelk boys. I know John Shahidi, who's one of the people behind that operation. And to look at the Trump victory speech and then watch Dana White come up there and shout out the Nel boys, you know, at a US Presidential speech. Presidential, like, victory speech to me was just. It was crazy. It's like, what reality are we living in? And the answer is we're just living in the podcast reality. So we thank you all the listeners for being here. I mean, this is. It's pretty crazy.
- [31:17]
Ranjan Roy
Yeah. I think in terms of, again, like, the other thing I kept thinking about, this is what is the message you're trying to communicate? Because one of the kind of, in hindsight things I keep seeing is, you know, like, the Harris campaign, what was the clear message they were trying to do, was trying to send? And especially when you're going through traditional mediums, there has to be an incredibly clear message. Like, simple, to the point. Here's what I'm about. And what was interesting was like, basically towards the end, it. From. From the Trump side, it was tariffs and immigration. From the Harris side, it was Trump is a fascist. And that was like the only. There was, like, economic messaging. There was no real clear messaging around that. There was like, something around home ownership, something around small business opportunity, economy, they were trying to push through. But what I kept thinking was, like, that lack of clear economic messaging, because especially in this kind of information ecosystem, to make messages really flow through, you can't just depend on one TV ad, one cable news interview, even one viral moment. It has to be a consistent, simple message that's repeated over and over in every medium. And again, that's where Trump has always been great. I mean, he's consistent on the messaging side at the high level. And I think, like, it showed through in terms of.
- [32:44]
Alex
And then he goes on some tangents.
- [32:45]
Ranjan Roy
Exactly. But that's what it's like. I think that's where, like, while people are laughing at whatever Hannibal Lecter rant is going on, in the end, he said immigration 20 times and he said tariffs a few times. And like, those messages, I mean, those messages have been repeated since 2016. So it becomes pretty Clear. And that's why now, because things are so fragmented, it's even more important than ever that politicians need to have just some simple, clear message.
- [33:16]
Alex
Why do you think Harris didn't do as many podcast interviews as Trump was? I, to me, it's just like it was way too careful of a campaign. You have to be a little bit unguarded in these podcast interviews. And I just don't think that came naturally to her.
- [33:30]
Ranjan Roy
Yeah, I think, and I saw the Fox News interview, so it was fascinating to me. It's like I'm going to walk into the most adversarial place on earth in order to try to reach undecided viewers versus I'm going to go to. I think if she goes on Rogan, it's a reasonable conversation and maybe people get to know her a little better and people feel more comfortable and that could change votes. And I do think that that was a bit of a mistake, but I think you nailed it. And again, not to get too into the pure, kind of like, political operations side, it was a very careful and well orchestrated campaign. And in this media ecosystem, I don't think that works. And I think, like, again, the Biden administration, Biden campaign in 2020, the message of stop the chaos. Let's return to some normalcy. Clearly, after four years of Trump and the pandemic was an easy one, a clear one. That was the biggest thing. Again, for me, what was the message other than stop Trump, Trump is a fascist, which was clear and it didn't work.
- [34:36]
Alex
Yeah. I was reading in Axios today, they did a breakdown of, like, the aftermath of the election. And it was interesting to see that this, when they, the way they phrased it, made me think this was even a worse loss for the Democrats than I initially thought. Because not only did they lose the White House, not only did they lose the Senate, it looks like they're going to lose the House as well. But the influence of mainstream media just took a hit, like, all of a sudden. And I don't even think it's controversial. The networks that used to drive the conversation, that were sympathetic, more sympathetic to the Democrats, basically everyone outside of Fox News, there's now a counterweight to them in this podcast circuit that people are paying attention to and listening to. I mean, the 50 million number for Rogan. Right. I mean, that's not all in the U.S. it's not all people who are eligible to vote, voting age, et cetera. But that's still a significant number. Let's say it was just 1/5. Right. That's a pretty good percentage of the country that tuned in.
- [35:34]
Ranjan Roy
Yeah, it's not even a counterweight in my opinion. It's almost because what happens on podcasts, on social media drives a conversation on network television anyways. So I think going forward, it really is interesting and especially for me, the whole misinformation topic was an interesting one in this cycle because at some level we hear it's not as prevalent as it was in 2016. Some level you hear it's even more prevalent. To me, it's almost like this is the information reality we've all chosen. This is how people want to get information. It's going to be a bit siloed. It's going to be. So how you get messages to travel through that. Ryan Broderick at Garbage Day I loved. He was, he had written an America when you where you can watch msnbc, Stephen Kornacki deliver calming data points on a live stream like a zoo animal or an infowars stream on X hosted by Michael Flynn and the Papa John's guy. As Musk wrote this morning, you are the media now. I think it is a reality that like people's information diets are very, very siloed and algorithmically driven right now. So that, that's just the reality that we've all chosen and accepted. So you have to be able to navigate that rather than playing the old game of like, we're going to have a great convention and we're going to do a couple of cable news interviews and we're going to be a bit vague on policy and just have some kind of like very, very, you could tell, market tested things again, like $25,000 for your first home as a buyer, an opportunity economy. Those things you could tell were like political consultant tested versus something just simple and organic.
- [37:22]
Alex
It's weird because, I mean, I'll say I'd much rather have both, right? Like I'd much rather have the politicians sit down for interviews with, you know, the net news networks sit down for interviews with the Times and the Journal and the Washington Post for that matter. And then also do these long podcast interviews because, you know, you're not going to face as many tough questions on these long shows. But who you are does come out over three hours. Who you are comes out and that's why people like them, because there's a level of authenticity to them. But what you miss from podcasts, I don't think ours like we done, I'm a journalist, this is a show that's conducted journalistically. But what you miss with a lot of these talk podcasts is they're so friendly that you're never really pressed on any of the things that you'll have to defend. And I do think in a democracy, being forced to defend some of your controversial choices is important.
- [38:15]
Ranjan Roy
I mean, almost being overly cynical at this point. Do you think, do you genuinely view being adversarially challenged as a politician is helpful to you and your campaign?
- [38:29]
Alex
No. No. Okay. It's not helpful to them. I just think that, like, if their job is to serve the people, they should be willing to answer tough questions.
- [38:38]
Ranjan Roy
You're more optimistic than I am right.
- [38:40]
Alex
Now, but I'm too idealistic for sure. Yeah, yeah, I'll admit it. Okay, so podcasts matter. Let's take a break. I think that when we come back, you know, you talked a little bit about where Elon Musk and Donald Trump might differ. And to me the answer is open source AI is going to be one of the early places because there's some stuff that's happening that's really worth talking about. And we'll break it down right after this.
- [39:09]
Tomer Cohen
I'm Tomer Cohen, LinkedIn's chief product officer. If you're just as curious as I am about the way things are built, the insight behind what it takes to create world renowned products, then join me for my new podcast, Building One. Together we'll get to learn from leaders around the world, people with diverse backgrounds across multiple industries. Each will share insights into their craft. So listen and follow my new show, Building One on Apple Podcast or wherever you get your podcasts and check out the conversation on LinkedIn. It's going to be great.
- [39:42]
Alex
And we're back here on Big Technology Podcast Friday edition. Okay, a long segment to start talking about the election and now let's talk about governing. How about that? Because there's four years of governing that Trump is going to have to do after this. And to me, this is something I pointed out in Big Technology today. There's going to be a big problem coming for open source AI. So I'm a fan of open source AI generally and I think a lot of people in the tech world are right. It allows companies to customize and to build and to develop things they couldn't with off the shelf models. It also enables us to understand this technology a little bit better than we would have if it's all being developed behind closed doors, as is the case in, let's say, OpenAI. But the thing about open source AI is that it's open and we don't really have any rules in the book in terms of how this technology can be shared and used. I mean there are obviously like user manuals, but there's no laws. And now there was a story in Reuters that said I'll just read it exclusive. Chinese researchers develop AI model for military use on back of meta's LLAMA Top Chinese research institutions linked to the People's Liberation army have used meta's publicly available LAMA models to develop an AI tool for potentially military for potential military applications. The researchers used an earlier Llama 13B large language model from Meta, incorporating their own parameters to construct a military focused AI tool to gather and process intelligence and offer accurate and reliable information for operational decision making. The tool is called Chat bit and Meta's terms of service prohibit the use of models for military warfare, nuclear industries, application or nuclear industries or applications and espionage. However, Meta, because these models are public, has limited ways of enforcing the provisions. And so what I think you're going to see is you're going to see Trump, who is adversarial to China come into office and he's going to attack this stuff. He's not going to want to see open source AI getting into the hands of the Chinese military. And that I think is where big AI decision number one and potentially big conflict with Elon number one comes into play. Because as you'll remember, Elon founded OpenAI. He hates the fact that Sam Altman has closed up the access to the models and he's open sourced models from Grok. What do you think Ranjan?
- [42:07]
Ranjan Roy
We have disagreed a lot today. I think I'm going to fully agree with you on this one. I as Tesla stock has been up 26% this week since the election. Metta was actually down 5% the next day. It was the only thing read on my charts and it's about up it's up like 1 1/2% since the election. I think it's we're going to see Trump clearly going after Mark Zuckerberg and I think open source AI is going to be a very easy one that it's going to be hard to argue that something that the Chinese military is using is something that's in the national interest and it's going to be again this is going to be something I think we're going to be talking about a lot more. I did like how you specified in the terms and services terms and conditions that they specify that it should not be used for military use or espionage. Like could you imagine some, some like general in China going on to Llama and being like, wait, wait, wait.
- [43:09]
Alex
Cannot use it.
- [43:11]
Ranjan Roy
Sorry, guys.
- [43:12]
Alex
Damn. We were about to put it into production.
- [43:15]
Ranjan Roy
So close.
- [43:16]
Alex
Can you believe what they wrote in.
- [43:17]
Ranjan Roy
The terms of service dams and terms and conditions. We're going to be getting a lawyer's letter very soon. Cut it all. That is kind of. That they do that.
- [43:26]
Alex
Yeah, no, I think, I don't think that that really comes into play. But it is interesting because this stuff is making its way into the military. And this is also from this week. Meta permits its AI models to be used for US Military purposes. This is from the Times. Meta will allow US Government agencies and contractors working on national security to use its artificial intelligence models for military purposes. It will make them available to defense contractors like Lockheed Martin and Booz Allen, as well as defense focused companies like Palantir and Anduril. Which is interesting because, like Andral founder Palmer Luck, he was, he left Meta under, you know, some, I think, even though he was fired for Meta because of his political activity. But now they're working together. And Nick Clegg met as president of global affairs at the company. Now back responsible and ethical users of the technology that supported the United States and democratic values in the global race for AI supremacy. Do you think Meta got an under. Got a whiff of the fact that the Chinese were building on its technology and decided to push this out right beforehand?
- [44:32]
Ranjan Roy
I like that they're making a exception as long as it's responsible and ethical use of the technology. But now that you're saying this and I'm reading this, I think I'm, I want to know when does this all come to a head? Because I think this is going to be one of the first really big things that happens under Trump in terms of like both the intersection of technology and global politics. I mean, I feel Trump has to believe that AI and is going to become kind of like a central power source for every country and like, incredibly important to economy and national security. So especially. And it's a perfect US versus China thing to kind of start talking about and moving on. So I think it's very clear now that that's. This is going to be one of the first big things.
- [45:23]
Alex
Yeah, I mean, look, if you think China is an adversary, how are you going to stand by if their military is using American open source models against the terms of service?
- [45:34]
Ranjan Roy
Against the terms of service. That's really what irks me.
- [45:37]
Alex
I just think that this is a pretty watershed story that Reuters broke here. And of course, like, it's just like, you know, A research tool and it's helping military decision making. But that's a big deal. You know, if that goes into production, that's a big deal. And we know that Meta, I mean, Meta's been building this. Mark Zuckerberg just announced they're going to build the next model on a hundred thousand GPUs plus one hundred thousand plus GPUs. And do you think China's like looking at that and being like, ship, baby, ship, we're ready to build on the next thing? Of course they are.
- [46:10]
Ranjan Roy
My favorite part about this was it's the tool called Chat bit, which I looked up stands for chat based Intelligence technology. So not the most creatively named thing. It was fine tuned and optimized for dialogue and questioning, answering tasks in the military field. I'm so curious, like, is this some like, highly sensitive thing though? Or is this just like people asking about their pension funds or like, like, I don't know, I'm curious like how sensitive even China would be about like really pushing sensitive information and decision making onto a American based open source model.
- [46:49]
Alex
It's your open source model. I mean, it's your model once you customize it with your parameters. That's the beauty of open source.
- [46:55]
Ranjan Roy
Yeah, yeah, yeah. This is going to be, this is going to be a fun one to watch and a terrible one.
- [47:00]
Alex
But on the name, you're not going to call something like chat super secret China military project Chad bit is about as explicit as you're gonna get.
- [47:09]
Ranjan Roy
It's actually even more terrifying because it's so innocuous.
- [47:12]
Alex
9 so that would lead me to believe this isn't about pensions, man. This is about where to direct the nukes.
- [47:18]
Ranjan Roy
This is terrible. They're like, what is the most boring name you can possibly give? The most dangerous thing in existence.
- [47:27]
Alex
They're rolling out Chat Bit. It's over. Anyway, it's been a been a fun election week podcast episode. I'm sure we'll get back to the meat and potatoes next week talking more about Gen AI and tried and true tech stories. We had one about Apple thinking about building their own smart glasses, but we'll do that next week.
- [47:47]
Ranjan Roy
So get me back to Siri and some earnings and I'll be comfortable.
- [47:52]
Alex
Exactly. All right, Ranjan, thanks for coming on. Great speaking with you as always. See you next week and thanks everybody for listening. Great to have you here. We have another great episode for you on Wednesday, including a real fun launch for big technology. So you'll have to stay tuned for that. We'll see you Next time on Big Technology Podcast.