Skip to content
The Joe Rogan ExperienceThe Joe Rogan Experience

Joe Rogan Experience #1768 - Dr. Robert Epstein

Dr. Robert Epstein is an author, professor, and Senior Research Psychologist at American Institute for Behavioral Research and Technology: a non-profit, non-partisan organization that offers data regarding the power of Google and other Big Tech companies to censor dissenting opinions online and sway the outcome of elections.

Joe RoganhostDr. Robert Epsteinguest
Jun 27, 20242h 41mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 1:37

    Search engines as hidden persuasion: why “relevance” isn’t neutral

    1. JR

      (drumming) Joe Rogan podcast, check it out.

    2. RE

      The Joe Rogan Experience.

    3. JR

      Train by day, Joe Rogan podcast by night, all day. (rock music) Uh, first of all, thank you for coming. I really appreciate it. This is, uh, uh, a very interesting subject, because I think, um, search engine results, uh, have, uh, always been a thing that people kind of take for granted, that the search engine results is gonna give you the most significant results at the top. And they don't really think about the fact that this is kind of curated. And, uh, you know, we found that many times, because we use two different search engines. We'll use Google, and then we'll say, "Well, if we can't find it on Google, use DuckDuckGo." And oftentimes, when you're looking for something very specific, you'll find that you can't find it on Google. Like, they're, uh, if it's in there, it's deep, deep, deep, you know, many pages in. Whereas, DuckDuckGo will give you the relati- you know, the relevant search results very quickly. So something's going on with search engines, and from your research, what you have found is that it can significantly affect the results of elections.

    4. RE

      Well, not, not just that. It can affect (clears throat) how people think. It can affect, um, your opinions, attitudes, the purchases that you make, uh, pretty much... It's, it's, it's a mind control machine. It's, it's the most powerful mind control machine that's ever been invented. Uh, and by the way, you should never use the Google search engine. Never.

    5. JR

      Never?

    6. RE

      Never.

    7. JR

      Why is that?

  2. 1:37 – 3:26

    “S&M” business model: surveillance + manipulation as the product

    1. RE

      Because it, this is what I call, um... And this isn't, this is an S&M platform, and I'm, I'm not sure what S&M means to you. I don't wanna pry into your personal life, but, uh, point is that, uh, but I... Well, what I mean by S&M is it, this is a surveillance and manipulation, um, platform. Uh, on the surface, there are always two, two levels to everything with Google. On the surface, it's a, it's like a free public library kind of thing, right?

    2. JR

      Yes.

    3. RE

      That's always on the surface. Beneath the surface, it's something different. From a business perspective, it's an S&M platform. It exists for two purposes only, and that is to trick people into giving up lots and lots of personal information. Notice your public librarian doesn't do that. Did you notice that? They don't actually do that (laughs) , you know?

    4. JR

      Right.

    5. RE

      And then it's also used for manipulation, because they discovered quite a few years ago that if they control the ranking of the search results, they can control people's opinions, purchases, votes. Now, they can't control everyone's opinions, because a lot of people already have strong opinions, so the people they're going after are the people who are undecided, the people who are vulnerable, and they know exactly who those people are. And they literally... Your, your mind is putty in their hands. Uh, so you should never, ever use Google or any, any other S&M product. Like, Amazon Alexa is an S&M product. Um, or the Google Home device or Android phones. Uh-

  3. 3:26 – 9:39

    Always-listening devices and de-Googling: practical privacy alternatives

    1. JR

      Android phones are bad?

    2. RE

      An Android phone is an S&M device. It's, it's always listening. It's always recording.

    3. JR

      Android phones are always recording you?

    4. RE

      (laughs) Are you serious?

    5. JR

      Yeah, I mean, uh, just, I'm, I'm questioning this. I mean, I believe you, but uh-

    6. RE

      Mm-hmm.

    7. JR

      ... I just want you to elaborate.

    8. RE

      Oh, yeah. There have been court cases in which the recordings have been subpoenaed, uh, from, uh, whoever's controlling that per- you know, that so-called personal assistant or that, that device, and courts have recovered recordings and transcripts, uh, when people are not even aware that they're, that they're being monitored.

    9. JR

      I know that's the case with Alexa, right?

    10. RE

      Yes.

    11. JR

      But that's the case with Android phones as well?

    12. RE

      Yes. In fact, And- Android phones, uh, uh, uh, the equipment to, to, to prove this, which I didn't bring but... Um, it's so cheap now that literally anyone can confirm this. Android phones, even if they're disconnected from your, uh, your mobile service provider, even if you pull out the SIM card, okay, as long as the power is on, it is recording, tracking every single thing that you do. So if you use it to read things, if you use it to listen to music, you use it to shop, whatever it is, it's, it... And of course, your location is always tracked. Then when you go back online, the moment you're, you're reconnected, it, it uploads all that information. So, some people wonder why their batteries run down sometimes even when you're not really doing anything with your phone. That's because, uh, with Android phones, it's, I think it's 50 to 60 times per hour it's uploading. Uh, it's uploading about 12 megabytes of data per hour, so that's a lot of, a lot of energy. That, that, that requires energy. Uh, so it... I mean, the kind of phone I have is completely different. It doesn't do anything like that.

    13. JR

      What do you have, like a no agenda type phone? Do you know that show No Agenda?

    14. RE

      No.

    15. JR

      It's, uh, uh, my friend Adam Curry who's the original Podfather, he's the guy who invented podcasting, and he, uh, his company develops these de-Googled phones where they take out all the tracking stuff, all, everything, and it's, uh, it's, it's basically running on a different operating system.

    16. RE

      Right. So I have a phone that runs on a differ- different operating system. It's completely de-Googled, and, um, and-

    17. JR

      What do you got? Can you show it to me?

    18. RE

      Yeah, I can show it to you (laughs) .

    19. JR

      I'm just interested.

    20. RE

      It's looks, just looks like any old regular phone.

    21. JR

      Right.

    22. RE

      But it's not.

    23. JR

      Does it run on Linux? What's it running?

    24. RE

      No. It's a, it's a different operating system. (laughs)

    25. JR

      Can you not tell me?

    26. RE

      Well, I can tell-

    27. JR

      It seems you're very cryptic.

    28. RE

      I, uh, well-

    29. JR

      It sounds like you're trying to hide this, Robert.

    30. RE

      Well, the point is I, uh, look-

  4. 9:39 – 12:36

    How Epstein got pulled into this: Google ‘sheriff of the internet’ and blacklists

    1. JR

      So, when did you first become interested in digital surveillance and privacy and, like, what you're giving up by using these free services-

    2. RE

      Mm-hmm.

    3. JR

      ... like Google?

    4. RE

      Mm-hmm. I wasn't interested at all. I've, I've been a researcher for 40 years, and I had a lot of research underway. I've done research on, on teenagers, on creativity, on stress management, all kinds of things. I'm still doing all that stuff. But on January 1st of the year 2012, I got a, I don't know, eight or nine messages from Google, uh, telling me that my website had been hacked and that they were blocking access. So, I thought, the first thing I thought was, "Why, why am I getting these notices from Google? Who made Google the sheriff of the internet? Why am I, why is this coming from the government? Why isn't it coming from some non-profit organization?" So that got my attention. Mm, and then, 'cause I'm a, I'm, I'm, I'm a coder, I've been a programmer since I was a teenager, and then I started wondering, "Wait a minute. Okay, they're blocking me on the Google search engine. I get that. That's th- that's them, right?" So they, so, so they have these crawlers, you know, that, that look at all the websites every day. And their crawler found some malware on my website. That happens all the day too. Everyone gets hacked. I'm sure you've been hacked and Google has, itself has been hacked. So, I get that they're blocking me on Google, google.com, the search engine. I get it. Okay. But I notice they're also blocking me on Firefox, which is owned by a non-profit. They're blocking me on Safari, which is owned by Apple. I, I thought, "But that's, how could that be? These are, these are completely separate companies." Took me a while, took me a while (laughs) to figure this out. I finally, uh, published a piece in US News & World Report, uh, an investigative piece called The New Censorship. And I (clears throat) described nine of Google's blacklists. This was 2016, so this was a while ago. In detail, I described nine of Google's blacklists, I explained how the blacklists work, I explained Google can literally block access on multiple platforms that aren't even theirs. They can block access-... to any website that Google, at one point in time, uh, 2009, I think it was, I don't know, I might get the date wrong. Let's just say January, whatever, 30th, uh, Google blocked access to the entire internet for 40 minutes. Google, (laughs) Google ... A- anyway, in this article I-

    5. JR

      When you s- when you say that, did, uh, if... With all browsers? When you say blocked access to the entire internet, it's like if you used the Brave browser back then. Did, did it even exist back then? Probably didn't exist back then, right?

    6. RE

      Brave didn't e- exist.

    7. JR

      No.

  5. 12:36 – 20:48

    “Google shut down the internet”: the blacklist dependency chain

    1. RE

      But no, there were, uh, uh, well, there were lots of search engines. Google was not the first search engine. It was the 21st search engine.

    2. JR

      So, but, what I'm saying is, with all web browsers, it blocked access to the internet?

    3. RE

      It blocked access to virtually the entire internet, to virtually everyone in the entire world for 40 minutes.

    4. JR

      What?

    5. RE

      And this b- this was reported in the news.

    6. JR

      So is... What's happening with their, their system is because so many people are searching for things, because they're monitoring so many different things to add to their search engine, do they have some sort of ultimate control over the internet in some weird way? Here it is, right here. "Google blacklists entire internet. Glitch causes world's most popular search engine to classify all web pages as (laughs) dangerous." Wow. "Google placed the internet on a blacklist today after a mistake caused every site in the search engine's result page to be marked as potentially harmful and dangerous." Holy shit. The fact that they can even do this... I like how it gives you like a, at the top, "This article is more than 12 years old."

    7. RE

      Okay.

    8. JR

      Imagine that, like 12 years means like it's ancient, like they wrote it on stone tablets 12 years ago and-

    9. RE

      Yeah, but, but, you know, this is nonsense. This report is nonsense.

    10. JR

      Is it?

    11. RE

      Of course. They, they, this is... Google is full of geeks, okay? I'm part geek, so I, I can relate.

    12. JR

      You speak geek?

    13. RE

      I c- I can speak geek if you want.

    14. JR

      (laughs)

    15. RE

      Uh, and, and, and geeks, for fun, okay, and sometimes for profit, but most of the time it's just for fun, just to be cool and to h- get their kicks and show the, how powerful they are.

    16. JR

      To be elite?

    17. RE

      Yeah. So they do, so they do crazy things. So they shut down the internet. I, I guarantee you it was a geek thing because you know, you know why I figured that out? Because I kept wondering, "Why did they shut it down on this early, super early morning on a Saturday? Why? What's so special about that little period of time?" And I... (sighs) It took a while and I figured it out. It's because that is one of the only (laughs) intervals of time in the entire week when every single stock market in the world is closed.

    18. JR

      Oh.

    19. RE

      So they, they did it to show that they could do it and have their fun, but they didn't wanna get attention. And if they'd, if they'd interfered with, with financial transactions, they would have gotten a lot of attention.

    20. JR

      So no one was ever caught?

    21. RE

      No one was ever caught, but they never denied that this happened either.

    22. JR

      So this was done through Google, for sure. They know this how?

    23. RE

      They, they... It's, it's... It's reported in the news reports and the, and Q- Google was queried and Google said, "Yeah, yeah, that did happen. Yeah, we fixed it."

    24. JR

      So how does Google have the ability to even do something like that? How can that even be done?

    25. RE

      Well, that, that's what I explain in that article, is it's, uh, they, they have blacklists. Uh, let me, let me jump ahead and then I'll...

    26. JR

      Okay.

    27. RE

      Okay. But let me just jump ahead for a second 'cause I, you, you gotta, you gotta see really how sinister this whole thing is. It's just... Seriously, if you knew, (laughs) if you knew a half of what I know about all this dark tech stuff, you would just say, "The hell with it," and just give up. You'd say, "I don't wanna, I don't wanna bring up kids in this kinda world. This is too crazy." But anyway, blacklist. So, I'm telling-

    28. JR

      I feel, I feel like we need to stop you there and make you elaborate. Like, what are you saying?

    29. RE

      Well, there's... Well, what, what I ended up doing, which I think we should get to later in some detail if you're-

    30. JR

      Okay.

  6. 20:48 – 39:39

    Where bias comes from: algorithms, rogue engineers, and executive mandates

    1. RE

      Okay, well, you're, you're saying all this stuff, uh, that looks may- maybe it's biased, maybe it might influence people.

    2. JR

      Yes.

    3. RE

      You know, you know, where's it coming from? Maybe it's just the algorithm. Uh, well, uh, let's say it's just the algorithm. Well, the algorithm was programmed by human beings, okay?

    4. JR

      Right.

    5. RE

      (laughs) And those human beings have, um, biases.

    6. JR

      Sure.

    7. RE

      They, they have beliefs. And there's a lot of research now showing that, uh, that bias, whether it's conscious or unconscious, gets programmed into the algorithms. So the algorithms all by themselves have biases built into them. There's one way it can go. Second way it can go, the Marius Milner effect. You ever hear of Marius Milner?

    8. JR

      No.

    9. RE

      Okay. Oh, this is great. This is great. Okay.

    10. JR

      Okay.

    11. RE

      Marius Milner, okay, few years ago you probably heard that Google got busted because their Street View vehicles were driving up and down... they're still driving up and down streets all over the world, but they had been driving up and down streets all over the world, more than 30 countries, for more than four years, and they weren't just taking pictures of our houses and our businesses. They were also sucking up wifi data. (laughs)

    12. JR

      Hmm.

    13. RE

      Massive amounts. I mean, we're talking terabytes of wifi data, passwords, everything, including a lot of very deeply personal stuff. So, uh, uh, someone just like me, a professor type, figured this out, reported them to the government. The government went after them. And so this is called the... this is called the, the Google Street View scandal. And so they got a fine $25,000 for interfering with the, uh, the investigation, and then the... and they blamed the entire thing, Google blamed the entire operation on one software engineer. His name is Marius Milner.

    14. JR

      Oh.

    15. RE

      So they fired him. Oh, no, no, that's not true. He's a hero at Google. He's still working there. If you look him up on LinkedIn, his, his, his profession is hacker. He's a hero at Google. They didn't fire him. Okay, this is... they love this kinda stuff. So another possibility besides the algorithm itself is a single rogue programmer at the company (laughs) can fiddle with what... with content, can fiddle with any content. And when a, when a single rode... rogue programmer does that, guess what? That shifts thinking and opinions and behavior and purchases and votes. A single rogue programmer can do it. And then, of course, there's the executive level. The executives can pass along a, uh, a mandate, a policy. Does that ever happen? Oh, yeah. One of the links, or, uh, leaks rather, one of the leaks from Google that you may have... you may have seen, uh... and I know the guy who leaked it. It's, uh, Zach Vorhes, who's also a good person for you to talk to 'cause he was a senior software engineer at Google for eight, eight years, and then he, he just couldn't stand it anymore and he quit. But unlike most of these people who've walked away, he brought with him 950 pages of documents and a video.The video is two minutes long, and it shows the CEO of YouTube, which is owned by Google, her name is Susan Wojcicki, and she's talking to her staff, and she's explaining ... This is 2017, after the horrible election results of 2016. And she's explaining how they're altering the Up Next algorithm in YouTube to push up content that they think is legitimate, and to suppress content that they think is not legitimate. So, if it's happening at that level, the executive level, again, it still has the same effect. It ... Any of these possibilities, and there are others as well, ends up giving us content that, that impacts us and our kids especially in ways that people are entirely unaware of. So, the way, the way I like to put it is this: you don't know what they don't show.

    16. JR

      Mm. Now, I'm still confused as to how Google can blacklist websites, and how they can shut down the entire internet for 40 minutes. Because, uh, d- do they have a switch? I mean, like, is there a connection that all websites go through Google? Like, how is that possible?

    17. RE

      (clears throat) About three years ago, they shut down all of Japan.

    18. JR

      Accidentally?

    19. RE

      Uh, well, that's, you know ... that would take a whistleblower to figure that one out. Um, they, uh ... I- it was in the news at one point that the guy who was in charge of these, making these decisions, he actually has left Google. Uh, he, he once shut down an entire domain name which had 11 million websites on it, 'cause he thought it was, uh, kinda poor quality.

    20. JR

      Poor quality?

    21. RE

      Yes.

    22. JR

      Poor quality, like how- uh, how so? Like-

    23. RE

      I don't know.

    24. JR

      He ... This is just his, his take, that it was poor quality?

    25. RE

      I have a copy of the internal manual, I'd happy to send it to you, fr- from Google, that - showing the criteria they use in deciding, you know, which content to suppress. And, and some of the criteria are pretty straightforward, having to do with pornography and things like that. Uh, and then there's this, like, wide-open area that, that says, um, "Or anything else."

    26. JR

      "Or anything else"?

    27. RE

      Pretty much, yeah.

    28. JR

      So, you ... It's up to the discretion of the engineer?

    29. RE

      There's a lot of discretion involved in making these decisions. And a lot of the decisions that get made in very recent years, since, since Trump was elected, uh, they happen to be decisions, for the most part, that suppress, uh, conservative content. But not always. Not always. Now, now I'm gonna go ... now I'm gonna circle back-

    30. JR

      But could you please-

  7. 39:39 – 42:37

    Whistleblowers and proof of blacklists: Voorhees documents and conservative suppression claims

    1. RE

      Can I, can I go back to something?

    2. JR

      Please do. Please do.

    3. RE

      'Cause I, okay. Well, I just wanna go back to blacklist.

    4. JR

      Yes.

    5. RE

      Because I wrote this big piece on nine of Google's blacklists. Their biggest one is called the quarantine list. That's that list that Safari has to check and that list that Firefox has to check. Everyone has to check that list before they take you to a website. So that's a simple way, simple tool that Google uses to block access to websites. 'Cause we go to websites v- through browsers.

    6. JR

      Right.

    7. RE

      Right? Okay. There we go. I had never seen any of those nine blacklists, but I knew they existed as a programmer and I talked about each one in detail. 2019, I'm invited to testify about my research, uh, o- on m- my experiments on manipulation and the m- and you know, how I monitor elections now and all that stuff. So, who, who testifies before me? Uh, a top executive, a vice president from Google. He's under oath. You know, he's sworn in. His senators are asking him some really tough questions. And he's asked point-blank, "Does Google have blacklists?" I think the full question might be, might have been, "Does Google have white lists and blacklists?" And his reply was, "No, Senator. We do not." So that was July of 2016. Uh, 2019 rather. 2019. Okay. In August, literally three weeks later, Zach Voorhees, who I mentioned earlier, that's when he leaves Google. Google s- (laughs) Google sends, uh, y- like a, what's that called when, when the police s- Oh, SWAT team. Google sends a SWAT team after him, uh, kid you not. Yep. So they were all very unhappy because he stole all this stuff, okay? And he sent it, he put it all in a box and sent it to the Attorney General of the United States. Okay? This is 2019, August. This is only less than a month after this hearing. So he's got 950 pages of documents, all kinds of documents, and three of them are Google blacklists, which are actually labeled blacklists. Now, if I were, (laughs) if I were putting together blacklists at my company, I would call them shopping lists. I would call them, you know, uh, I don't know-

    8. JR

      Sweethearts.

    9. RE

      ... ma- makeup lists.

    10. JR

      Yeah.

    11. RE

      Uh, you know, lists for my kids, uh, birthday presents or I, I wouldn't call them blacklists. (laughs) So there, there are actually three of them he walked out with. So you can look at the list. You can see who's on the list. You can see these are almost all or m- many of them prominent conservative organizations.

    12. JR

      Oh.

    13. RE

      There, there are no left-wing organizations on those lists.

    14. JR

      Mm-hmm.

    15. RE

      So this is real.... this is, this is how they operate.

  8. 42:37 – 48:15

    Google’s three motives: money, ideology/values, and intelligence ties

    1. JR

      And they operate this way, do you think they do this because of ... Are they financially driven to put those people on blacklist? Is it, uh, maybe so- I mean, it's obvious- obviously speculation, but is it maybe some sort of a deal that they've made with certain politicians? Is it something they've decided on their own because this is the right thing to do, to suppress the bad people that put Donald Trump into office? Like, why, why are they doing that?

    2. RE

      What you just did was amazing. (laughs)

    3. JR

      What'd I do?

    4. RE

      'Cause you got it almost all of it. You just-

    5. JR

      Yeah, that's it?

    6. RE

      ... ca- came up with it hypothetically, but you left out one area. So, the two areas you, you, you just nailed, one is to make money.

    7. JR

      Right.

    8. RE

      So they have three motives. One is to make money, and they, that they do extremely well and, and no one who's, who's tried to tangle with them has, has stopped that process. In other words, the, the rate at which they're making money continues to increase every year. So, a few years ago, when I was first looking at them, they were bringing in $100 billion a year, now they're bringing in $150 billion a year. Money, that's number one. Number two, values. Okay? This, this, I could, I could talk for hours on this issue (laughs) because of recent leaks of videos, PowerPoint presentations, documents, and of course, what whi- whistleblowers have been revealing. They have very strong values there because the founders had very strong values, and they hired people who had similar values, and they have really strong values, and they, they want the world to have those values. They, they really think that their values are more valuable than other people's values, which means they don't understand what values are because-

    9. JR

      And so their values are, th- they're pretty much all of tech is very left-leaning.

    10. RE

      Very left-leaning. So, uh, 94, 96%, um, of all donations, uh, out of Google go to Democrats, which I sympathize with. I'm v- from a family of Democrats, I lean left, so I say, "Yeah, that fine. That's fine, that's fine." But it's not fine. It's not fine, because they, they're, they have the power to impose their thinking on other people around the world i- in, in a way no one has ever h- had such power, ever.

    11. JR

      Mm-hmm.

    12. RE

      So value is a second, and they real- and this is serious. One of the leaks from Google was an eight-minute video, which y- you should definitely watch. It's so creepy, and it's called The Selfish Ledger, and it's eight minutes, and it was, it was put together by their advanced, their super secret advanced products division. It was never meant to leak out of that company. And I, I have a transcript of it too, which I've published, so I can get you all that stuff. But point is, what is this about? This is about the ability that Google has to re-engineer humanity, uh, according to company values.

    13. JR

      Re-engineer humanity according to company values?

    14. RE

      Yes.

    15. JR

      And, and this is a directive, like this is something they're doing purposely or aware of?

    16. RE

      Well, in the videos, uh, th- in the video, they're, they're presenting this as, as an ability that we have.

    17. JR

      Jesus Christ.

    18. RE

      This is an ability that we have. (laughs) Uh, so that's the second area. You, you nailed it. Third one you didn't mention. The third one is, uh, intelligence, because they, uh, they had some support, um, Page and Brin, right in the very beginning at Stanford. They had some support and, and had to be in regular touch with, uh, representatives from the NSA, the CIA, and another intelligence agency. The intelligence agencies, uh, were doing their job. Okay? They, they realized that the internet was growing. This is l- 1990s. So they realized that the internet is growing, and they were thinking, "Hey, the, these are people building indexes, indices to the content. So sooner rather than later, we're gonna be able to find threats to national security by looking at what people are looking up. If someone is going, uh, online, they're using a search engine to, to find out instructions for building bombs, for example, okay, that's a potential threat to national security. We wanna know who those people are." So right from the outset, okay, and this is totally unlike Brave, okay? Brave doesn't do this, but right from the very, very beginning, the, the Google search engine was set up to track and preserve search history. So in other words, to, to, to keep track of who's doing the search and where did they search? That is very, very important to this day for intelligence agencies. So Google, to this day, works very closely with intelligence agencies, not just in the US, but other agencies around the world. So those are the three areas. Money, values, intelligence, and the intelligence stuff, uh, is legit. I mean, it's legit, you know. I- it is an obvious place. If you're, if you're in law enforcement, uh, that's an obvious place to go to find bad guys and girls.

  9. 48:15 – 1:13:57

    Measuring persuasion: SEEM, search suggestions, and the Hillary 2016 example

    1. JR

      Yeah. So Google has this ability that they've proclaimed that they can cor- sort of shift culture and direct the, the opinion of things and direct p- public consciousness.... what percent, like, how much of a percentage do you, do you think they have in shifting? Do they have, like, a 30% swing?

    2. RE

      (laughs)

    3. JR

      Like, what?

    4. RE

      Well, see, this is what I do. Now you're getting, now you're, now you're getting close to what I actually do, what I've been doing for, for now for over nine years. I quantify, this is exactly what I do every single day, that's what I do. My, that's my, my team, my staff, that's what we do, and it's, and it, and it's cool.

    5. JR

      Mm-hmm.

    6. RE

      And talk about cool. We're, we're w- we're doing the cool stuff now, okay? Google is not. We're doing the cool stuff, because we are l- we have discovered a number of different tools that Google, and to a lesser extent other companies, use to shift thinking and behavior. And what we do in randomized controlled experiments, which are also counterbalanced and double-blind and all that stuff, we measure the ability that these tools have to shift thinking and behavior, and we, we pin it down to numbers, percentages, proportions. Uh, uh, we, we can make predictions in a, in an election about how many votes can be shifted if they're using this technique or these three techniques or ... And, uh, and so we, yeah, that's what we do. So we've, we started with the search engine, and, uh, and we, it, it took years and years of work, but we, we, we really, I think at this point, uh, have a good understanding of what the search engine can do. Uh, but then along the way, we discovered other tools that they have and which they are definitely using, and how do we know they're using these tools? Well, we can get to that, but-

    7. JR

      What are the tools?

    8. RE

      Well, the first one we called SEEM, search engine manipulation effect, and that means they're either allowing, uh, you know, one candidate or one party to rise to the top, you know, in search rankings, or they're making it happen, and you don't know for sure whether, you know, which is, which is occurring unless there's a whistleblower or there's a leak. Okay, but the fact that it's occurring at all, that's important.

    9. JR

      Right.

    10. RE

      I mean, we don't, in a way, we don't care, because if it's just the algorithm that's doing it, well, that's horrible. That means that, that means literally a computer program is deciding who's gonna be the next president, who's gonna be the next senator. Do, do we want that (laughs) decision made by an algorithm? So anyway, we, we, we spent a lot of time on that. We're still studying SEEM. Uh, then we went, we, we learned about SSE, which is search suggestion effect. When you start to type ... Oh, in fact, if you have your phone handy, this will be fun. If you, if you start to type, uh, a search term into the box-

    11. JR

      Mm-hmm.

    12. RE

      ... the search box, you're, you're, there are, uh, suggestions flashed at you.

    13. JR

      Mm-hmm.

    14. RE

      As fast as you're typing, that's how fast those suggestions come.

    15. JR

      Right.

    16. RE

      Well, guess what? We learned in controlled experiments that (clears throat) by manipulating the suggestions that are being flashed at people, we could turn a 50/50 split in a group of undecided voters into nearly a 90/10 split.

    17. JR

      Wow.

    18. RE

      Without anyone having the slightest idea that they're being manipulated.

    19. JR

      Whoo.

    20. RE

      Just, just by manipulating search suggestions.

    21. JR

      Just by suggesting.

    22. RE

      Yes, and the, uh, and the reason why, um, we started that work, because in June of 2016, uh, a, a news organization, a small news organization released a video which went viral on YouTube and then got blocked on YouTube, frozen, still frozen. Uh, but then it continued to go viral on Facebook, so 25 million views. In this little video, th- they're, this news organization is saying, "We've made a discovery." Uh, when you go to google.com and you look for information about, um, Hillary Clinton, you can't get any negative search suggestions.

    23. JR

      Really?

    24. RE

      Really. So if you, and they showed this, and there were-

    25. JR

      What if you googled, like, Clinton body count?

    26. RE

      You, you, you, you could not get negatives. It, it-

    27. JR

      Really?

    28. RE

      Yeah. It would, it, it would give you nothing probably for Clinton body count, but as you're typing, you go, "Clinton b," it would go, you know, "Clinton, um, um, buys the best clothes." I don't know.

    29. JR

      Right.

    30. RE

      It would give you something like that. It would not give you something negative. So for example, Hillary, Hillary Clinto- or Hillary Clinton is ... You do it on, uh, and they showed this, you do it on Yahoo, uh, you do it on Bing, and you get, "Hillary Clinton is the devil."

  10. 1:13:57 – 1:20:46

    Beyond Google: YouTube’s ‘Up Next’ power and opinion-matching quizzes

    1. RE

      For the past year or so, we have been doing controlled experiments on YouTube. Uh, we have a YouTube simulator. It's a perfect YouTube simulator, and we have control. We're using real content from YouTube, real videos from YouTube. All the titles, everything comes from YouTube, except we have control over the ordering and we have control over the Up Next algorithm. That's where the power lies, the Up Next algorithm. So, one of the things we learned recently, not from Frances Haugen, but it was someone else who left Facebook, is that 70% of the videos that people watch on YouTube now around the world are suggested by YouTube's Up Next algorithm.

    2. JR

      70%.

    3. RE

      70%.

    4. JR

      Whoa.

    5. RE

      Yeah. And that's (laughs) their algorithm, and just like us in our lab, okay, we have control over what the Up Next algorithm suggests. And guess what we can do with our Up Next algorithm?

    6. JR

      What?

    7. RE

      Well, it should be obvious. (laughs)

    8. JR

      You can manipulate people.

    9. RE

      Yeah, we manipulate people.

    10. JR

      Yeah.

    11. RE

      We randomly assign them to this group or that group, and we just push people any old way we wanna push them.

    12. JR

      And when you're doing these tests and studies, like, how are you doing this? Like, are you doing ... How many people are involved in this? Are they students? Like, what ... How are you, how are you doing this?

    13. RE

      Okay, we- we- we never do the, you know, subject pool at the university where you get, you know, 50 students from your- your college to take, you know, to be your research.

    14. JR

      Mm-hmm.

    15. RE

      We never do that. So we're- we're always reaching out, uh, to the community or we're doing things online. So we do big studies online, and we are getting very deg- diverse groups of people. We're getting, uh ... Literally, we're getting people from, uh, lists of registered voters, so we're getting people, you know, who- who- who look like the American population, and we are ... We- we can mess with them. Can I say "we can fuck with them"?

    16. JR

      You just did.

    17. RE

      Oh. I guess I just did.

    18. JR

      (laughs)

    19. RE

      Oh, this is definitely not Fox. This is not Fox.

    20. JR

      No.

    21. RE

      No. No.

    22. JR

      This is ... We're on the internet.

    23. RE

      This is not Fox News. Uh, yeah, but the internet ... See, the internet, though, because there are no regulations and rules, it does allow for some pretty evil things to take place. And the fact is, in our experiments, we- we- we do these ... Usually our experiments have hundreds of people in them. Sometimes they have thousands of people, and we can fuck with people and they have absolutely no idea. Let me j- ... I'll tell you about something new, okay?

    24. JR

      Okay.

    25. RE

      Something new, brand new. Okay, and this is f- ... Uh, uh, thank God I'm not talking about Google this time. Just talking about something else that's happening. There are websites that will help you make up your mind about something. So for example, there's a whole bunch of them right now that'll help you decide whether you're really a Democrat or you're really a Republican. And the way they do that is they give you a quiz, and based on your answers to how you feel about abortion and immigration and this and that, at the end of the quiz they say, "Oh, you are definitely a Republican. Sign up here if you wanna join the Republican Party." And this is called opinion matching, and the research we do on this is called OME, the Opinion Matching Effect. And there are hundreds of websites like this, and when you get near an election, a lot more of them turn up because the Washington Post will give you a quiz and help you decide who to vote for, and Tinder ... Tinder-

    26. JR

      Hmm.

    27. RE

      ... okay, which is used for sexual hookups, uh-

    28. JR

      How about romantic, sir?

    29. RE

      Oh.

    30. JR

      Not just sex. Sorry.

  11. 1:20:46 – 1:36:42

    Congress, partisanship, and the personal cost: threats and Epstein’s wife’s death

    1. JR

      Yeah. And when you ... Y- so you spoke to Congress about this? You spoke in front of Congress, or was it-

    2. RE

      Mm-hmm.

    3. JR

      Now, w-

    4. RE

      Yeah.

    5. JR

      And when you did, was there any sort of urgency that was ... Uh, d- did anybody understand, like, what the implications of this are? Did anybody understand, like, this is ... We're l- literally looking at these massive technologies that are used throughout the world that, uh, that can completely change the way policy is directed-

    6. RE

      Mm-hmm.

    7. JR

      ... the way people are elected, who's in charge, what the public narrative is on a variety of subjects?

    8. RE

      Well, uh, there, (laughs) there are some people. There's a g- a guy named Blumenthal, he's a senator from Connecticut. He gets it. He understands. He, he's kind of disgusted, I would say-

    9. JR

      (laughs)

    10. RE

      ... with all this stuff. But, you know, I said to Cruz, I said, "Why don't you work with Blumenthal?" And he said, "Well, no, I don't think that'll work out."

    11. JR

      'Cause he's a Democrat?

    12. RE

      'Cause he's a Democrat. So-

    13. JR

      Yeah.

    14. RE

      ... they can't ... Here, here's the ... So, the Democrats, even the ones who understand this stuff, they won't do anything. Because, why? Because these tech companies support Democrats very lavishly with donations. So, for example, Google was Hillary Clinton's biggest donor in 2016. And they're supporting them in these more subtle ways as well, so the Democrats will not, will do nothing. If they even say something, like if they, if they, they rattle their swords, they don't actually do anything. And the Republicans hate regulation.

    15. JR

      Mm.

    16. RE

      This is a perfect storm for these companies to do what they have done, which is they have already taken over. You're, you're thinking 20 years from now? No, they've already done it.

    17. JR

      Well, I'm not thinking, um, that they haven't already taken over. But I'm thinking, like, how much more control can they have in 20 years if 20 years ago they didn't have any? Like, it ... As technology advances, do you think that this is going to be a deeper and deeper part of our world?

    18. RE

      Well, look at, look at, uh, Zuckerberg. Zuckerberg's trying to-

    19. JR

      Yes.

    20. RE

      ... get us all into the metaverse.

    21. JR

      Exactly.

    22. RE

      Okay? So, yeah, you have even more control if you get people into virtual realities. Yes, you have more control. Uh, uh, every single thing they're doing is moving us farther and farther and farther down the rabbit hole.

    23. JR

      Well, not just that. I'm thinking, like, there was a time where Zuckerberg, at least, was, uh, publicly considering cryptocurrencies.

    24. RE

      Right.

    25. JR

      Like, some sort of a Facebook cryptocurrency. Imagine if Facebook cryptocurrency-

    26. RE

      (laughs) Right.

    27. JR

      ... became the number one currency worldwide, or maybe it, maybe it was the number one crypto, like Bitcoin is today.

    28. RE

      Sure.

    29. JR

      What the fuck? You know, so you're in the metaverse. In order to e- exist and compete-

    30. RE

      Yep.

Episode duration: 2:41:56

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode TIRtBfUBMmk

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

iOSAndroidClaudeChrome