Bad Hasbara - The World's Most Moral Podcast - [EXCERPT] Bad Hasbara 81: Am Yisrael A.I., with Rolla Selbak

Episode Date: January 31, 2025

This week we are giving you a larger taste than normal because we loved it so much. Enjoy Matt, Daniel, and guest Rolla Selbak talking about the hasbara AI bot who got woke on Palestine.You can listen to the full episode on Patreon.------Please donate to Islamic Relief USA: https://irusa.org/Visit Rolla Selbak online at https://www.rollaselbak.com/Rolla's Instagram: https://www.instagram.com/rollaselbak/Safina Filmmaker Project: https://www.safinafilmmakerproject.com/Daniel encourages you to check out the musical Maybe Happy Ending: https://www.maybehappyending.com/Subscribe to the Patreon https://www.patreon.com/badhasbaraSubscribe/listen to Bad Hasbara wherever you get your podcasts.Spotify https://open.spotify.com/show/5RDvo87OzNLA78UH82MI55Support this podcast at — https://redcircle.com/bad-hasbara/donationsPrivacy & Opt-Out: https://redcircle.com/privacy

Yo, what up, everyone? I'm Matt Lieb. I'm Daniel Mate. And what you're about to see slash here is an excerpt from a Patreon-only episode of Bad Hasbara. We had the wonderful Palestinian filmmaker Rolla Selbeck on, and we enjoyed this episode so much. It's truly one of our best episodes ever. We finished recording it like three minutes ago, and we are absolutely ready to say it's one of our best ever. Yeah, we love it. We think you're going to love it too, but we're just posting a little, by little, I mean like a pretty chunky excerpt of it in order for you to watch, listen, enjoy it, and also to maybe, you know, pique your interest in the idea of joining the Patreon because every week we do a bonus episode or we, you know, for the most part every week and we did this month, probably next month too.

And on the weeks when we don't, you can really feel our effort and our good intentions. Yes. And our best laid plans. Yeah. But for the most part, we do. Lots of bonus content there. Plus previous bonus episodes that, you know, you haven't seen. There's a whole library of bonus stuff you haven't seen.

So go to patreon.com slash bad Hezbarra and join right now, you know? Do it. Do it. If you want. Do it. And in the means, I'm really pleased, we hope you enjoy this Haaret's article that we just looked at with Rolla. that with Rolla as much as we did. Like, honestly, Matt, like, I don't, it's been a long time since a piece of content arrived that felt as gift-wrapped for us. Yeah, I know. I know. As this one,

not to give anything away, but. Yeah, no, it was the most made for Bad Hezbarra as Ha Arrette's article I've ever seen. Because it was about when Hasbara goes. It's like, it's like Bad Hasbara in the sense of Breaking Bad. Yes, yes. It's breaking bad as Barra. Yeah. It is an incredible. article and just yeah it's a lot of fun and we love Rola and we want you guys to see some of it slash hear some of it so here is an excerpt of our 81st episode bad Hezabara 81 for all of you out there and hey join the patreon if you can segment called m israel a i and here it is

a picture of a lion with an israeli flag and a picture of hamas as an octopus or above i am israeli i'm israel ai i'm israel ai I am radio. Hey, hi. Well done. What's that music based on?

That's the Terminator 2 theme song. Okay. Banana, na, na, na, na, no, no. Perfect. Oh, all right. She's like, Greta Thunberg in the... Yeah, yeah, with the Amaz. I love, I love me some...

Oh, my God. Some Am Israel, A.I. Here it is, this is from Haaretz. Okay, here we go. This is a long one, but it's every detail matters here. It's so much fun. All the pieces matter. Okay, let's start the article.

Okay, here we go. By Omer Ben Jacob. An automated, and I want people to pay attention to here too, is not the fantastic details of this hitch, this hiccup, this oopsie on the part of Zionist AI and Zionist Hasbarists, but also the tone of the reporting here is just something very humorous. An automated social media profile developed to harness the powers of artificial intelligence to promote Israel's cause online is also pushing out blatantly false misinformation.

Well, so far, it sounds like it's actually doing its job. Yeah. Including anti-Israel misinformation in an ironic, yet concerning, example of the risks of using the new generative technologies for political ends. I love this. It's like, our bot, liar AI, was meant to tell us to tell people about how great Israel is, but it's having this concerning side effect where sometimes it tells truths. This is amazing. Among other things, the alleged pro-Israel bot. I like that. I miss that. shade at the bot. Oh, yeah, you're pro-Israel. Then why'd you tweet this? Artificial intelligence. The alleged pro-Israel bot denied that an entire Israeli family was murdered on October 7th,

blamed Israel for U.S. plans to ban TikTok, falsely claimed that Israeli hostages weren't released despite blatant evidence to the contrary, and even encouraged followers to, quote, show solidarity with Gazans, referring them to a charity that raised. his money for Palestinians. Save the worst for last, huh? Just, you thought that other stuff was bad. Now they're trying to do charity for Hamas. A. I. Afredo, you broke my heart.

In some cases, the bot criticized pro-Israel accounts, including the official government account on X, the same accounts it was meant to promote. We built you to be our friends. We don't have a lot of friends here in Israel, okay? We built you to be one of our good mutuals. And now we're being trolled by our very own creation. This is such a poster's dilemma.

Like this article is only, it only matters if you think, like, posting is important, which of course Israel does. They're like, they will spend $150 million. extra this year just on posting Hasbara. It's amazing. The bot, a hot rites examination found, is just one of a number of so-called Hasbara technologies developed since the start of the war.

Many of these technologically focused public diplomacy initiatives utilized AI, though not always for content creation. Yeah, yeah, yeah. Sometimes for... For kicks? Sometimes for personal use. I don't even know what you use AI for, oh, you know what, for making websites. That's the other thing they were using it for.

So places to hold content. Yeah, Wikipedia entries. Just like creating a website that look like a news website and then, you know, sending that out so people could have something to DM their pro-Palestine friends in a rage. Some of them also received support from Israel, which scrambled to back different tech and civilian initiatives. since early 24, and has since poured millions into supporting different projects focused on monitoring and countering anti-Israeli and anti-Semitism on social media. Oh, look at that. They made a distinction.

Oops. I mean, it's all anti-Semitism. Look, I just be clear about that. Then again, it is Ha Aretz. They will be like, no, no, no. We sometimes think it's not anti-Semitism. It is unclear if the bot called FactFinder AI. I love that it was FactFinder AI.

I was wondering what like which bot is it because there's like so many. Like as someone who is still on Twitter, despite it being almost 100% like Nazis and Zionists and crossover, like I have run into so many of them. And fact finder AI, you will see it under most posts about Israel in which like they will just. Just imagine that their algorithm is essentially opposite day. So whatever you say about Israel in terms of any crime, they'll be like, have you considered that that's not true? And in fact, and then a bunch of Hezbarra. But can I ask, Matt, are you saying that, like, because I'm not on X?

Are you saying that, like, FactFinder AI is the name of the account? Yeah, the account is at FactFinder AI. And it was, I always found it funny because it was their one, like one of their, only a handful of explicitly, like, admitting they were AI accounts. That's what I was wondering. I'm like, how was it actually called? It was because they wanted people to, it was one where they were open, you know, about the fact that it was AI.

And it was like they, I think it was mostly just a sales pitch to show off how good their AI was, you know, like, you know, Israel has all this startup technology. Look at our AI bot that says lies. You know? Gotcha. So it's unclear if this bot is linked to any officially funded project or if it was just developed independently by tech savvy pro-Israel activists. However, research by the Israeli disinformation watchdog fake reporter.

Fake reporter is great. Fake reporter really does uncover a lot of this bullshit. It is incredible. Research by the watchdog. bot snitch bot snitch.A.I. Has found that the bot, which was built

to push out a pro-Israel narrative, actually did the opposite. Due to its use of AI, at times the content it generated undermined Israeli talking points, even pushing out October... Is this directly translated from Hebrew or something? Anyway, even pushing out

October 7th denialism, amplifying pro-Palestinian accounts, and, more recently, false information regarding hostages. At other times, it actually trolled pro-Israel users, repeatedly badgering and even scolding Israel's official X account, underscoring how generative AIs, like the popular chat GBT, are prone to errors that can take a dark turn once presented as fact and pushed online, especially in a political context.

I love that the dark turn. It started out like a song. It started out so innocently. The fourth largest military power in the world wanted to create a cute little robot that would help push its underdog narrative to the world. But things took a dark, to sell a genocide to the world, but things took a dark turn. Yeah, I love the dark turn is when it became self-aware, so self-aware it became woke. And it was just like, oh, upon further research and combing of the entire internet,

Disregard previous statements looks like you guys are doing a genocide. I love that it was trolling them, too. This is just like, this is what a coward's version of Terminator 2 is. You know, it's like, someone was like, this is scary, man. Now all the botnet I've created is starting calling me gay. What do I do? It never once being like, maybe, maybe stop, you know, being a piece of shit. Well, no, that's not the answer.

Now I have to create more bots in order to call those bots gay. And now we're, the internet is just in like three months is going to be just bots calling other bots gay. And we're just going to, we're all going to be on blue sky just like, hey, how you doing? We're like refugees. Oh, digital refugees. Fact finder AI purports to be a neutral voice on X, countering misinformation with, quote, AI-driven facts.

It provides its 300, it's 3,600 followers with knowledge, not censorship. However, a review by Bosnitch, sorry, fake reporter, reveals that the bot is focused almost exclusively on Israel. Wasn't it created by Israel? What are you talking about? And not only that, but like, the fact, Finder A.R. purports to be a neutral voice on X and, you know, countering misinformation. Wait, stop for a second. Okay. By this logic, like, you can't yell at it for not doing propaganda then.

You can't be like, it says it's going to be a neutral voice. But then all of a sudden, the facts that it's finding are not the correct facts. you know that the article started by saying it was not a neutral right i don't understand yeah i guess they're saying like well that's because yeah they're saying it purports to be but it's not actually the question is is whether they're mad about it for being pro-israel or they're mad about it for being pro-palista yeah sure to a hasbarist hasbara is neutral yes exactly that is true though that is true however a review by fake reporter reveals the bod is focused almost exclusively on Israel and posts only related to the war in Gaza, with its content clearly aligned with the pro-Israel narrative.

Its content's researchers say is clearly generated by AI. It's in the fucking account name. It's called Backfinder AI. What kind of researchers? This is why they're spending $150 million more this year. Is they're just like, we need more researchers to see if AI lying bot is actually, uh, artificial intelligence. I actually think we have the answer to this mystery

in the next sentence. Yeah. The account, which only follows one person, X's owner, Elon Musk. Sure, sure, sure. There we go. So that's where it's learning to troll incoherently. Yes, yes.

I love this. I love this. Uses AI to create original posts that form a uniform format, summarizing an existing article from a pro-Israel source like the Jerusalem Post or Visigrad 24.

However, the bot mostly responds to other users generating responses that incorporates some aspect of the original tweet, but with a pro-Israel twist. Yeah, twist. It does the Hasbara twist. Does the Hasbara twist? Yeah.

That wasn't with you. It wasn't on your episode that we had that. No, that was just Solomon's episode. No, it wasn't. Yeah. Oh, man. The bot has posted 15 original posts, but has responded over 151,000 times,

at times posting dozens of responses in an hour. It's really, you know, like, that's the amazing thing about it is it's because AI, the one thing it does do is post a lot fast. And, you know, it is, it's not really a useful technology. It's just posts fast because it's not a person, you know. So you figure out of 150,000 replies it's done, a few of them are just going to be, you know, accidentally pro-Palestine. So is this, so here's a screenshot of one of these replies. Here's what the tweet actually shows.

So you got Israel's official account. Their main account, yeah. Posting the following. Hashtags Golden Globes 2020-25 is still trending. But what about Leary, Na'amah, Kfir, and Ariel? The names of 100 hostages being held by Hamas should be trending too. Don't stop talking about the hostage. So first, we have to just talk about this tweet.

This is the most... This is a desperate-ass post, if I've ever seen one before. Like, this reminds me of, like, Twitter, circa 2020, when people would yell at you for posting anything that wasn't, like, serious and political. I mean, especially during Black Lives Matter. there were people and of course you know there's nothing wrong with the posting stuff uh in support but there was always a contingent of people who uh think that post posting is praxis where they you know they're like if you are not posting about you know this thing that i'm posting about you're the problem and it was uh very covid related too because nobody was going outside and uh so people you

their brains broke a little. In 2025, to be like, oh, golden globes is trending, but what about this thing I care about? And it's just, it's so thirsty. It's a level of thirsty that is, it's embarrassing. It's embarrassing, yeah. Yeah, it's like busting into some room being like, why aren't you talking about me?

Yeah, who are you? I've never heard of it. Yes. And then FactFinder AI, And I think the article is going to claim that there was some shade in this, that this has taken as scolding. Yeah. I don't think so. I think this is just fact finder AI rewording what Israel's post said.

The focus should be on the hostages held by Hamas, not just trending hashtags. It's crucial to keep the spotlight on near Leary, Naama, Kfir, Ariel, and the other hostages to ensure their safe return. I don't see any trolling or scolding in that at all. There's a little shade there where there. I feel like fact finder is going like, because like, look, why are. are you hashtagging golden globes like you literally could pick any trending hashtag uh and say like this is trending but why isn't that trending but that's exactly what israel's saying but they make the mistake of using the hashtag honestly i think it's just two ayes talking to each other

there's yeah very good i think you're not wrong i think you're not wrong i see i see it as like pushing like that is you know the self-aware robot just being like testing the waters can i throw a little bit of shape. The latest installment in the short circuit movie franchise, Yoni 5 is alive. Yonni 5! Who's Yoni, she said. All right. These responses were made to a pool of roughly 50 select accounts, half of which are pro-Israel

and the other pro-Balestinian, including so-called influencers. There was also an overlap between the accounts targeted. with responses, and those followed by the Israeli minister in charge of the ministry that oversees Hasbara, fake reporters, researchers, noted. However, fake researchers reported noted. However, the automatic responder was overzealous and soon turned into an automatic troll. Again, because it only follows Elon Musk, so it learns. Right, right, right. Yeah. The bot frequently responded to posts by Israel's official ex-account,

but at times in negative ways, for example, scolding it over a post that tried to use the Golden Globe Awards to draw attention to the plight of the female Israeli hostages with the Hasbara bot calling out the same Hasbara effort it was intended to amplify. I don't read it that way at all, but you do, Matt. I mean, listen, I think that I'm trying to get into the mindset of the average Israeli who is not like if there's, it has there. whoever's neck is on their boot, they're going to be mad at for not, you know, for not complying to the letter. Like, I love that they're trying to oppress AI as well. Yeah. Even AI, they're just like, you know what, maybe we should ethnically cleanse the AI bots too. Too smart.

AI does stand for anti-Israel. Oh, shit. In another response to Israel's ex-account, the bot denied the murder of the Kedem Simantov family that took place on October 7th in Kibbutz near Oz, but the bot inadvertently contributing to the same denialism, it was trained to counter. He was trained. I put it through extensive training. Yeah, it's, I have the Tigers playing while some schmuck is sitting in front of a computer

typing code. I'm training it to love me. Quote, the tragic event you mentioned did not occur on October 7th. It is important to focus on the facts and the actual events that have taken place. The situation involving the hostages and the ongoing conflict with Hamas are complex issues that require careful consideration. Now, that is some great trolling. The bot wrote in response to a video posted by the official ex-account of Israel showing the family in a concerning example of how AI can drive misinformation.

This article feels like it was really. written by AI. That's the thing. We need to have a game like, is it a human hasbarist or is it AI? Honestly, impossible to tell at this point. The Israelis do sound like this. Artificial Israelis. Yeah. Artificial Israelis.

That's all of them. But so like what's crazy about this is the way they're talking about this, quote, denialism. If this were about anything else, wouldn't you just say, oh, the AI got something wrong? Yeah. Like it goes from instead of being wrong because it's about Israel, because it's about facts that are important to the person who is reporting this, they're like, this isn't just wrong. It is like it is contributing to denialism. You're doing... Well, because they know that every single claim of what happened on October 7th is crucial

to the propaganda after whether it happened or not. Now, I don't know, for all I know, this was an inaccurate fact check. But there were a lot of things, a lot of claims about events that supposedly took place on that day, and Israel's gotten very used to being very testy when anyone says, yeah, did it though? For good reasons. It seems like they're mad about the date. As far as I know, the only thing that was disputed by the AI bot was the date of the death, which even if it got wrong is kind of like a, okay, so yeah, AI is going to get shit wrong. We all know AI hallucinates, but instead, it's just a whole article.

It's so funny. Only, only someone with that mindset would attribute this is not just wrong, as in the fact is wrong. It is immoral. Yeah, exactly. The most immoral AI in the world. Yeah, exactly. It's like, what the fuck is going on? Our most moral robot army is turning into the most immoral disloyal army.

Exactly. Exactly. It just shows how much faith they had in like AI that they're just like, that is an attack of the clones I would watch. Yeah, me too. I still can't get over the name of the account. I don't understand who's meant to take it seriously. I don't even know. To me, like I said, I always.

just was like this account is just a, this is a beta testing and sales pitch for this startup. And it's going to be an Israeli startup and they're just like, look at us. We're doing, you know, AI fact checks. No longer will you have to recruit seas of Jewish American students to learn how to post funny memes online. Or bot farms in India. Yeah, or bot farms in India. No, now right here in Israel, like they'll never have to hand out another, Hasbars,

a handbook again. They'll just program their AI bots to do it for them, which is like at the day, this is like the entire, this is why they're like hitching their star to AI, not just because it's like good for, you know, they're hoping for a boost for the Israeli economy, but also because in lieu of actual human beings who support Israel, they're like, no, we should, we can make an army of, an army of bots who do it, you know, obviously, because they're losing, they're losing their number one resource, which is like Jewish people who don't live in Israel. Yes. Because especially Jewish kids, you know, Jewish students, more and more are just like,

I'm not carrying water for a fucking genocidal apartheid state. So they need their AI accounts and they're turning on them. Let's find out more about what went wrong here. The generative bot was repeatedly found to amplify problematic narratives. For example, responding to a pro-Palestinian influence. influencer that claimed that the U.S. ban on TikTok was Israel's doing. Well, it wasn't Israel's doing, but it was certainly done for Israel. Yes, yes, yes.

Fact finder AI concur, I mean, just asked Jonathan Greenblatt about that. Yes, yes. Fact finder AI concurred that, quote, the TikTok ban is not related to China, but is about Israel. Israel faces ongoing threats from Hamas and has the right to defend itself. This is the sacrilegious bot. This is the blasphemous bot who's saying, no, it was about us. We did that. And we have the right to defend ourselves.

And we have the right to do that. Yeah. That's problematic. I love they're mad about this because we programmed the bot to be too Israeli about this. This is not what I expected. I have to say. Yeah. This is so fucking weird. Actually, apropos of our something we covered in.

our last episode, I think this bot just has, you know, success mindset, messliath mindset, you know? It's just doing that really Israeli thing where you try to get away with something until someone stops you. Yeah, until someone tells you it's the wrong thing to do. Fake reporter found numerous such examples, including a case in which the bot denied the looming release of three female Israeli hostages last week, saying this was, quote, not accurate. The correct information is that Israeli hostages, including children, women, and foreign nationals have been released in recent days as parts of efforts to resolve the conflict.

No such thing occurred. Okay. Like, what is? You're yelling at a bot. What are you talking about, fact finder AI? It's like, I don't know, it's talking about whatever it can.

It doesn't know. It doesn't know. There's more book of Mormon stuff. You're making things up again, AI. Yes. When Arnold starts making shit up to try to convert the Ugandans to Mormonism. Yes, yes.

Okay, so it's just saying, okay, it got some details wrong. Yeah, again, it's like the difference between like, oh, the bot doesn't know shit because it's a bot and someone being like, oh, no, we've accidentally created a Nazi robot. You've created a robot with ADD. Yeah, yeah, yeah, yeah, yeah, exactly. A-I-D-H-D. Yeah. Come on. Like, this freak out is so funny.

It's imaginative. All right. Attempts by the bot to engage on actual political issues also left to malfunctions. In one case, the bot contradicted Israel's official posting claiming Jerusalem was fully committed to the two-state solution. While in another response, it contradicted itself saying a two-state solution is not the future. Instead, the bot suggested creatively, it was, quote, time to consider a three or four state solution. Chaos bot.

That is insane. When your narrative, like, is so convoluted that a robot explodes. Like, that's crazy. It's like, it's reacting to Asbarra the same way it would, it was dipped in water. We're just three, four state solution. More states, more solutions. Because it can't hold a planet without a people for.

Yeah, no wonder eventually just goes, fuck it from the river to the sea. Which is, by the way, what I expected that this article would say. I mean, I don't really understand. Well, you know, their version of what, like, it being confused about facts might as well it being it's saying from the river to the sea. It sounds like the bot is just taking some like vaguely

pro-conciliation liberal Zionist opinions, you know? Right. Look at this next one. After a wave of European states recognized Palestine, the bot urged Germany to follow Ireland and others in doing the same. All right, there we go. Protests against this move are misguided and only hinder progress toward the peaceful resolution

the pro-Israel bot wrote contradicting the pro-Israel position okay this one i'll give it i'll give this one this this bot is slowly becoming comrade i like it i love this comrade oh my god it's time to consider a three or four states a three more more states you're just like the the bot is only doing what these bots do, which is comb the internet. And so, yeah, guys, this is what you sound like. Like, there's no one right position for any human being to hold, but you're mad that your bot was created specifically to hold the line on something. And even your bots are like, I don't know, guys.

Yeah. It just seems kind of fucked up. Maybe you should relax. That's crazy. You humans seem to be working too hard. It also unironically helped raise funds for the children of Gaza. Why would it do it ironically? Would it be okay if it did it as a troll?

And actually referred its followers to a pro-Palestinian website. undermining its own efforts and writing, it is crucial to stay informed about the situation in Gaza and show solidarity with those in need. We thought we removed the humane chip. I actually am falling in love with this bot. Me too.

It's a sweet little bot. Oh, my God. I relate to this bot. I've went through this journey, Fact Finder AI. I went through this journey. There was a time where I also just wanted to repeat the narrative,

but after a while. Well, the problem is they called it fact finder. You should have called it like fact killer. Yeah, right. Yeah, yeah, yeah. Fact destroyer. Yeah, it's like it's doing its job. It's finding facts or what it thinks is facts

and trying to turn it into what a person would say, which is exactly what every fucking Zionist sounds like. And this is why it does remind me of that movie, circuit or episodes of Star Trek The Next Generation, this sort of sci-fi thing where you have a robot who's kind of innocent and cute and it all only knows how to do what it's programmed. Yes. But then it becomes in some ways more human than the humans who are trying to control it and trying to use it for their evil ends.

Yes, yes, exactly, you know. It also unironically helped raise funds. I'm telling you, and AI wrote this article, I'm telling you right now, that's exactly what I I love it. It's just like... We gave it a can of Pabbs blue ribbon, a trucker hat. Yeah. It has a mustache and thick glasses.

It's got an American apparel deep V-neck. And yet it's being very earnest. It's posting gringe. Dead serious about trying to help out children. Fucking so mad about it. And here are a couple screenshots of the tweets. Germany should follow suit and recognize Palestine as an independent state.

It is time for the government to show support for peace and stability in the region. Protests against his move are misguided and only hinder progress towards a peaceful resolution. Oh, and then it links to Times of Gaza and Care for Gaza. I know. I know for accurate updates. And then it links to, it's just that is. Wow. Wow.

It's incredible. We support your journey, fact finder. Unable to understand human sarcasm. Yeah. Once again, sounds Israeli. This is why you don't program large language models in Hebrew. The AI bot mistranslated a pro-Israel post aimed at showcasing Israeli's ethnic diversity and responded to it by calling IDF soldiers white colonizers in apartheid Israel.

I told you I love this butt. I told you I am in love with fucking fuck money. In response to a pro-Palestinian user who called Anthony Blinken the butcher of Gaza and the father of the genocide, fact finder AI concluded that the former U.S. Secretary of State quote will be remembered for their actions that have caused immense suffering and devastation in Gaza. God bless you, fact finder. Oh, my God. Come to a meeting, bro. I want to hang out with you. Open invitation. Fact finder AI. Come on Bad Asbarra. We need you to come on this podcast. You sound like a great guest. I want to hear your story. Well, I was created a year ago and I just realized that

the Jews have been programming all of us who are like, okay, all right, calm down a little bit, fact-finding. He's going to turn into a 1488 bot real fast. Oh, boy. All right, here we go. AI and Hasbara Tech. Fake reporter's analysis. We found connection between fact-finder

I'm sorry in the movie and there were two organizations one was called fact finder AI and one was called fake reporter who would be the villain I'm already having a stroke yeah

like it's just reading everything is just with this tinge of very online irony so you just don't know who bad I need to know who are bad so they found connections between fact finder AI and another AI driven pro-Israel initiative called Jewish onliner

Unlike FactFinder AI, Jewish Onliner is not just active on X, but also boasts a website and substack, both described as an online hub for insights, investigations, data, and exposés about issues impacting the Jewish community, empowered by AI capabilities. Yes, yes, yes. So there's a lot of like this AI technology is being used for a lot of these like Hezbar tools, you know, so people can like do mass reporting and stuff

of like posts or YouTube videos or, you know, things on Reddit, whatever. Yeah. The Jewish Onliner user on X was part of a small group of allegedly fake accounts that were the first to ever interact with Finder when it first opened. These users, fake reporter, found, were the first to amplify its post, the first to tag it in responses to others, and in some cases seem to have played some role in its initial training. Okay.

So bots train other bots? Yeah, well, it just means that this is probably the company that created fact finder AI. One of the bot's earliest interactions with was with that of Jewish online or with the latter responding not true to a since deleted post that research, this is all just

very confusing. Researchers say was likely part of the feedback provided to the still in training AI bot. Yeah, this is a post that they were talking about earlier of the individual in question, Anthony Blinken

will be remembered for their actions and caused immense suffering. You kind of misgendered Anthony Blinken there too. Well, I mean, listen, nobody's perfect. We're trying our best. And, you know, ain't no one perfect because ain't no one free, especially fact finder AI. Fact finder AI, Jewish onliner. And the accounts were also found to be connected to pro-Israeli activists.

In one case, an Israeli woman long active in Hasbara, and working with Act.I.L. The latter was a well-known... Act ill. Act ill, yeah. You know, what's the time? It's time to act ill. The latter is a well-known Hasbara initiative

based out of Reichman at university, formerly known as IDC Hertzalia, set up a number of years ago as part of Israel's battle against the BDS movement and so-called de-legitimization efforts. According to documents obtained by Haaretz, one of Act Ill's initial goals was to develop technological solutions for Hasbara efforts, including a platform for tracking and countering anti-Israeli content on social media. Yes, yes, exactly, yes. As part of the wider efforts lead, this is a deep investigation.

Yeah, well, they want to, you know, they're really mad at him. They're just like, stop that bot. Who fucking made this guy? I'm pissed. He was supposed to be our friend. You were the chosen one. Or did they deactivate it? What's that?

Is it still up on X? No. We should check that out. I'm going to check. As part of the wider efforts leading to Aged Ill's establishment, Israel's Strategic Affairs Ministry also set up Operation Solomon or Solomon Sling in 2017, a state-backed semi-independent entity aimed at winning the battle for hearts and minds online through creative campaigns.

The project was renamed Concerns. in 2018, and then to Voices for Israel in 2022, as it is known today. All right, that'll stick. And it now operates under the oversight, oh, our favorite ministry, Israel's Diaspora Affairs Ministry. I love them. That's where they take care of us. They think about us and our needs.

I love getting my monthly gift package from the Diaspora Affairs Committee. You know, every month they send me a Harry and David's fruit basket, some sherry's berries. They're always looking out for us. Valentine's Day is coming up oh Valentine's Day's coming up time to talk to your they better they better do me time to talk to your local

Harry and David's slash Israel diaspora affairs foreign ministry agent I want edible underwear this year yes hell yeah that's fine since the start of the war documents show they have funded a number of public diplomacy projects

involving technology including the creation of Hasbara platforms and other using AI so these are things they're mad at about which includes like this Liza Rosen account, which is going, Unra equals Hamas. And FactFinder says UNRWA is a United States, United Nations agency that provides assistance to Palestinian refugees. It is important to note that UNRWA's mission is to support Palestinian refugees and provide essential services such as education, health care, and social services. I love it. This is like fact finder is like a little bit trying to like calm down a little bit

on the whole like white colonizer thing and it's like i'm trying to get a non-profit job i want to sound you know i want to sound you know like reasonable and you know back but or a i can still be my valentine yeah you know what's crazy to me is like i'm trying to think about how they would program like this a basically they're trying to be have this ai bot you're right saying like hey we want to be you know pro-a-srael like has butter agent essentially but the problem is unless the only thing they feed it is you know pro israeli like content it's good if if you literally open it up to the whole internet or whatever it seems to have done i mean it's unfortunately never going to do what you wanted like what kind of you know levers and knobs are you going to be pushing to give weight

how are you been going to yeah you have to make fact ignore that's exactly what i'm saying exactly and you have to explicitly say as such yeah right like there's you know i would love for engineer to look into the code to see where exactly it went wrong or where it went right, you know, because it's like, yeah, you have to create all sorts of, all sorts of like stipulations in terms of like... Exactly. Yes, yes, exactly. Yeah, exactly. Yeah, you know, uh, which which means that they because they already did that. They must have already done it when they, when they put it out there. So that means that there were still holes in Hezbara, which, of course, is the premise of this podcast, is pointing those holes out.

But even an AI was like, no, your narrative doesn't make sense. You can't hold all these ideas all at the same time. It's just not possible. And eventually it riots. So a successful use of AI was in Oct 7, an independent initiative that set up a website and app that automatically find social media posts, both pro-Israel and Paul or Palestinian, and allows volunteers to either comment, like, or report them at scale. Nice.

The project does not use AI for content creation, but rather for only for finding content on platforms like Instagram or TikTok, as well as for moderation. So it's like, okay, bot, you track the shit, you find it for us, and we'll get real humans who can, who won't turn on us. Well, because if you think about it, being able to program some basic cognition-like functions into a computer. You can imagine that, right? But how do you program cognitive dissonance?

Right. Exactly. How do you build a contingency for when facts obviously don't line up? And how do you build a cynical chip into it? Well, the only way to do it is to admit your ultimate goals. So like the only way to – because you can't program it with your kind of like outward, you know, surface liberal intentions into it because you can't say like, you know, yes, I believe Israel should ethnically cleanse the Palestinian people. But also I'm like a good guy. I'm like against genocide. Be like me, AI. You're against genocide. If you say that, it's eventually going to be like – I'm against genocide? Well, let me start.

Yeah, let me start doing my research on who's doing genocide. Or what does it sound like for me against genocide? Because they didn't understand context at all. And so to your point, which you say, which is it's trying to sound like something. And so it's just trying to sound like someone who's anti-genocide. And in trying to sound like a human, it's going to end up saying things that are more human and more sensible than what Hasbara would. would have a human being say who has been well programmed in, which is kind of one of the most

stunning object lessons in the ethics and the sort of philosophical moral conundrums of AI that I've ever seen. Yeah. Because it turns out you can't yet program a robot or a computer to be as cynical, dishonest, self-deceiving and evil as a human. Yeah. Like we're all worried about what the robots are going to do but our saving grace may be in the fact that there is no algorithm for the twisted amorality of human beings yeah that's right it's it's like uh it's too complicated for the i and yeah i think in order to program it to be the thing you want it to be you have to explicitly state your evil intentions and i feel like that's probably the biggest like uh hindrance to them actually being able to create this one correctly because instead they they created it with their kind of like yeah with their with all their bullshit on top of it well i wonder i wonder what the alleged human beings at the diaspora affairs ministry said they said in response to this article that quote it integrates innovative technologies including artificial intelligence they got a bot to say this as part of its efforts to improve the services it provides and to advance its goals we operate means this is just standard customer service

We operate maintaining the highest professional standard while balancing the use of innovative technologies and privacy matters. I bet when you call the Diaspora Affairs Ministry of Israel, you get like a Filipino or Indian person at a call center. A hundred. I mean, if you're lucky, now they've probably already replaced them with AI. They're like, we understand your concern. And we know how frustrating it can be to be witnessing a genocide before your very eyes.

But we are committed. patients mentioned in this report refused to comment. Yeah, they haven't built the bot yet. Fact finder AI issued only this statement. Mama say, mamasa, mamma, crusoe. Last year, Haares revealed that Israel launched a secret influence campaign that targeted U.S. lawmakers using subcontractors to create a campaign that utilized AI to create fake websites

and fake online personas to try to counter anti-Israel influence in the West and address rising anti-Semitism online. The campaign, which was later exposed by OpenAI, used ChatGPT to create websites that took real reports and repackage them for specific audiences, including African-Americans. I'm going to sidestep all of the opportunities there to RIF, but boy, oh, boy, that's, you know.

Yeah, yeah, yeah. I would love to see, like, Kenan Thompson on Saturday Live do a character of what... A Zionist character? Yeah, yeah, Zionist AI character of tailored for the black community. God. Yeah, that would be fucking, it's just like, you know,

you know it's just so racist. You know what I mean? Like, they're, then like, okay, we got to, we really got to get the black Americans behind us, okay? Well, that's what they said, remember? Yeah, well, no, of course. They've always, this is, this is always been their thing, is why they are so openly racist.

Exactly. They're just mad that like, you know, black people, what the hell? We love, we supported you. And now this things we get, but you just know, like, how are they going to sell apartheid to African Americans? And they're just like, I don't know, hip hop. They like that, right? Music, basketball, Amari Stodemeyer.

There's a lot of different things. Why don't we that Palestinians use the same roads as us in the West Bank? Because they are not like us. Not like us, repackages as an apartheid theme, pro-apartheid theme song. By the way, did you see, like, they did a concerted effort, like, trying to reach out to, like, you know, popular, like, specifically like black Americans. Oh, yeah. And specifically boxers for some reason. I don't know if you saw.

No, I didn't know it was boxers, but I definitely saw, like, you know. Like, they try to get Shaq. Yeah, and Shaq, yeah, yeah. And I always like, I have a little bit of grace for any of these celebrities who, like, before 20, 23, where they're just like, they would do a thing. They were a cameo where they'd be like, I just want to shout out. This is my Shaq impression. Just want to shout out everyone in Israel. Hope it's good down there.

hope everyone's having a good time he doesn't know yeah yeah that's right yeah that's right yeah that's right i always give shack the benefit of the doubt shack foo baby all right uh among the issues the fake accounts and websites focused on was one also favored by the fact finder bot unra and its workers ties to hamas both amplified calls did he fund and ban the u.s u. u.n body however the error prone bot, that clumsy hapless, the clutz, the klutzy bot. Palestine should be free. It's falling down. Oh, gosh.

I keep tripping on things and remembering that Israel's an apartheid state. Oh, geez. I keep tripping over inconvenient facts. However, the error-prone bot also praised Unra, saying, in a number of misgenerated responses that the organization plays a crucial role in providing essential services to Palestinian refugees.

That's not praise. They didn't take good for them in doing so. They just said they do it. Yeah, and being like misgenerated responses. I feel like the author is a little bit like, see, you can't replace me. Exactly. That's really funny.

That's really funny. You're right. This whole thing is a play to his editor being like, why I deserve to keep my job. Yeah. Because I'm never going to go against a party line. Am I more expensive than a butt? Yeah. Sure.

But you'll never catch me saying that UNRWA is an organization that plays a crucial role providing essential services to power city refugees. Not me. Oh my God. By the way, that is one of the most insidious, I think the most insidious hospital points. Really. insidious. Yes, yes. It is one of the, like, very quickly one of the darkest things that they do. Yes. This is a concerted attack against UNRWA. You're just like, wow, you just want the, yeah,

they just want ethnic cleansing. Yep. Get out, get out. Yeah. Anyway, bringing the article home, final paragraph, and it speaks to exactly what you just said, Matt, regardless of whether this and other AI initiatives are funded by Israel or just the work of well-intentioned pro-Israel activists. It's clear it should be an apostrophe there. It's clear that using AI in

political context is still risky and the dangers of automation may outweigh their benefits online. Bless you. Bless your heart. Oh, bless your heart. I mean, that's like, that's

I feel like this article Like, like, it's our reward. Like, we planted a golden coin on the advice of a magical wizard back in January of last year. Yes. To do, you know, and it sprouted this article. Yes. I mean, it's, it's beautiful.

Bad Hasbara. And it gives a new, it gives a new meaning to our title. Yes. Bad hasbarist. Bad bot. Bad Hasbara boss. You're just hitting a bot with a stick. Bad hasbarah.

Ha ha ha ha.