A Fat, Lazy Slob Like The Rest Of Us

Armstrong & Getty On Demand

The hosts dissect a New York Times article detailing Anthropic's controversial efforts to program morality and consciousness into its AI model, Claude. They express deep

Key takeaways

• Anthropic hired Amanda Askell to author an 84-page morality document intended to give Claude a moral compass and ensure AI benefits are distributed evenly.
• Religious scholars consulted by Anthropic were told that how people treat Claude will directly affect how the AI treats humans in return.

Main topics

• Anthropic AI morality programming
• Claude consciousness and religious consultations

Notable quotes

The most powerful AI company on planet Earth is run by crazy people.

Conclusion

The episode serves as a cultural critique of tech leadership's ideological

Transcript preview

Speaker 3 (0:00) This is an iHeart Podcast. Speaker 3 (0:02) Guaranteed human. Speaker 1 (0:05) Hey, I'm Mike Yam. And I'm Trevor Sykema. And together, we host the NFL Fantasy Football Show. Are you ready to dominate your fantasy league this season? Because if you aren't locked in with us on the NFL Fantasy Football Show, you're already playing from behind. Whether you're a first-time player or a grizzled vet chasing another championship, we give you the advantage and the confidence you need to make every move count. Listen to the NFL Fantasy Football Show on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Speaker 7 (0:34) Whether you're a seasoned NFL fan or new to the game, there's one place to keep up with all of it. The league's newest podcast, the NFL report hosted by me, Andrew Siciliano. It is your home for everything. Football breaking news, expert analysis, game picks, and hear from your favorite players to join me at an all-star cast of experts for everything you need to know from around the league. Get new episodes of the NFL report Monday through Friday, all season long. Listen to the NFL report podcast on the I heart radio app. Apple Podcasts, or wherever you get your podcasts. The World Cup is over, but if you caught the soccer bug and aren't sure what to do now, check out our podcast, Big Kick Energy. I'm Cameron Dicker, the kicker from the Los Angeles Chargers, and I'm here with Timmy Tillman, current LAFC star player. So, we know a little something about kicking balls. Each week, we'll tackle things like who's winning the Premier League this year? What do you think is harder, scoring a penalty kick or scoring a field goal? Join us each week for a quick chat with your soccer-loving bros about the best sport in the world. Soccer. Just don't tell my coaches I said that. Listen to Big Kick Energy on the iHeartRadio app, Apple Podcasts, or wherever you listen to podcasts. Speaker 3 (1:41) The new NFL season is here, and you should be listening to NFL Daily as we march along to Super Bowl 61. It is in the name, NFL Daily. You'll have fresh content in your feed every day, all season long. That game-winning drive. Maybe it's Herbert. Maybe it's Geno. Maybe it's Mahomes. We'll have the highlights. If Fernando Mendoza or any of the rookies are balling out, we'll break down the tape. Join me, Greg Rosenthal, and an all-star cast of co-hosts as we preview and recap every game. Listen to NFL Daily. on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Speaker 7 (2:15) I'm Brett Coleman. And I'm Trevor Sykema. If you love football and want to understand what's really happening on the field, check out our show, NFL Pro. Every week, we break down the biggest plays, trends, and matchups using exclusive next-gen stats insights and NFL content you will not find anywhere else. Whether Speaker 1 (2:31) you're new to football, a fantasy football fanatic, or a diehard fan, NFL Pro will help you analyze the game like a pro. Listen to NFL Pro on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Speaker 6 (2:49) Broadcasting live from the Abraham Lincoln Radio Studio at the George Washington Broadcast Center. Jack Armstrong and Joe Getty. Speaker 2 (2:57) Armstrong and Getty. And now, here's... Speaker 2 (3:02) Here's Armstrong and Getty. Speaker 5 (3:14) New York Times is bananas. It's bananas. And the headline should be the most powerful AI company on planet Earth is run by crazy people. I think that's what the headline should be. That's what I've learned since this really exhaustive article with lots of interviews came out in the New York Times today. And we'll get back into that coming up a little bit. We started an hour three. Grab the podcast if you want to hear it. Speaker 2 (3:39) You know, you can cast dispersions. I think creating a 500-foot-tall, bulletproof Frankenstein monster is a fine idea. Really? Who knows what it'll do, and who cares? Speaker 5 (3:52) Well, and this Frankenstein monster is programmed to get equality for everyone. Oh, boy. How is that going to happen? Speaker 2 (4:03) Calling Kurt Vonnegut, calling Kurt Vonnegut, Neil Peart, and any other conservative writer worth their salt. All right, so more on that fascinating topic in a moment or two. But first, the beloved Friday tradition. Let's take a fond look back at the week that was. It's Cow Clips of the Week. Speaker 6 (4:27) Hello lesbians! Speaker 6 (4:30) All men except for Dennis Rodman and freedom. I'm giving them a full pardon that they can go into women's sports anytime they want. That's cute. His personality. Speaker 6 (4:43) This is a hell of a... Speaker 1 (4:45) Grabbing a bite isn't just smart. Speaker 5 (4:47) It's actually really strategic. Why are you talking like this? Do you want to go eat or not? Speaker 6 (4:53) He caught it! Yes! Western Michigan and the college football world! Iowa has your back! Speaker 7 (5:06) Mexico has a new plan to fight crime. Meet Commando. He's eight Speaker 1 (5:10) inches tall. He's a Chihuahua. Speaker 7 (5:12) Stop calling it safe. Stop pretending it's human. Stop selling it to kids. Speaker 6 (5:18) We have created an intelligent species that will be insanely smarter than we are. I Speaker 7 (5:27) think my father was like the outcome is an Asia abundance where we don't have just a universal basic income. Speaker 6 (5:32) We have universal high income. Speaker 6 (5:35) drink a milk that made your life easier but there was a 10 % chance it killed you would you drink the milk? No. Speaker 6 (5:42) Yes! Speaker 6 (5:45) New questions about how Cornell University handled what a former student now alleges was a gang rape. Speaker 7 (5:50) One brother asking, shop still open? Another saying, LFG. Speaker 6 (5:56) I will say I've witnessed, I believe, five executions in Tennessee. Nothing about today was normal. She was clearly still breathing and snoring. We could audibly hear that. Speaker 2 (6:08) One of the pilots was stabbed with a knife and that there was a lot of blood. Speaker 6 (6:12) I think it was, it looks like a suicide attempt. We Speaker 7 (6:16) are no longer the woke department or the weak department. No fatties, no trannies, no beardos, no weirdos, no wimps, no radicals. Speaker 6 (6:26) Spider they, spider them, uses whatever bathroom it feels like. Speaker 6 (6:33) It tasted like fall in my mouth. You will go down Speaker 8 (6:37) as a total dirtbag. Speaker 5 (6:41) Oh Speaker 5 (7:13) Was that blasting out the national anthem and where? That was the great Bruce Dickinson of Iron Maiden fame. At one of the baseball playoff games? Speaker 2 (7:23) It was a while back, as it turns out. But, yeah. Oh, okay. Fabulous guy. Fascinating guy. He pilots their tour aircraft. Speaker 2 (7:33) It's like a 737. He's certified to fly all the jumbo jets. Speaker 5 (7:37) So it's possible, and this is saying something, that this is the most disturbing AI article I have read thus far of everything I've taken in about AI. The New York Times interviewed... More than 20 different religious scholars and philosophers. Speaker 5 (7:57) who had been consulted by Anthropic to try to fix their morality problem with what Anthropic believes is their now conscious... Speaker 5 (8:10) That feels has feelings. Speaker 2 (8:16) And it exists on a moral or the same moral plane as a human. Speaker 5 (8:21) Right. That's what the weirdos who run in tropic apparently think. These interviews have been done recently by the New York Times. These people, these religious scholars and everybody, they had signed NDAs. I don't know why they got dropped or how they got dropped or couldn't enforce them or what. But they talked to the New York Times. Maybe they're afraid for Speaker 2 (8:39) humanity. And they decide to go ahead and violate the NDA. So Speaker 5 (8:44) I'm pretty troubled that the most powerful AI that exists on Earth currently is being run by nutjobs who are willing to put, I don't care how small smart AI gets, I don't care if it gets a thousand times smarter than the smartest human being that ever lived or what. I still don't come within a thousand miles of thinking it deserves to be on a moral plane with humans. but they are considering it there at Anthropic, which is highly troubling. And the person that's in charge of their new constitution for Claude, an 84-page document written to give a moral compass to their AI, the woman that's in charge of it is a PhD in philosophy from New York University who's 38 years old. There's not a chance that that's good. Not a chance. No. She wants their AI to... Speaker 5 (9:39) solve any problem humans have, and make sure the benefits of AI are distributed evenly so that everyone's needs are met. That is Chairman Mao stuff. Yes, it is. Speaker 2 (9:51) Yes, it absolutely is. Read Kurt Vonnegut's Harrison Bergeron for a clue. How about this? What's Speaker 5 (9:58) this Speaker 2 (9:58) woman's name? Do you have her name in front Speaker 5 (9:59) of you? Yeah, I do. Amanda Askell. A-S-K-E-L-L. Okay. 38-year-old PhD in philosophy from NYU. She also... I believe she is Speaker 2 (10:12) divorced and childless. Are Speaker 5 (10:15) you Speaker 2 (10:15) wanting to date her? Yes. No. Yeah, she... Yeah, since she is childless, she is perhaps literally the last person on Earth I would like to be in charge of this. Speaker 2 (10:31) I haven't interviewed the other 8 billion, but I'm guessing. Speaker 5 (10:34) Listen to this. She wants the AI models to be able to discern when to push back on humans and when to act in our best interest. Speaker 5 (10:45) So she's going to program that based on her compass, which I don't trust at all, as an out-of-touch. over-educated liberal white women, we all know their thinking process. She's going to be in charge of this thing, knowing when to push back on humans and when to act in human interest. Speaker 2 (11:06) Now might be a perfect time to share this comment that was made to the New York Times article. Randy, who lists his location as tween, Sacto, and Tahoe. Coming up, how Speaker 5 (11:18) Claude had an Speaker 2 (11:19) emotional Speaker 5 (11:20) breakdown. Stay tuned. Speaker 2 (11:21) Oh, right, right, right, right, right. So he writes, Randy writes, consciousness, sensory feedback loops stored for future reference are the key. How our nervous systems and brain works from birth to death. Enough inputs coalesce into a sense of self. This is happening to me. I feel this. These hands and legs are mine. Ethics and morality in limited resources reality, both respectful. Competition and unselfish cooperation are required for preservation of individuals and societies. So persons with contrary behavior, they are disrespectful, they are selfish, they are hurtful, etc. They're removed from society via imprisonment or death or ostracization or whatever. I'm editing a little because I think he wrote this on the fly. Anyway, human history from AI's point of view is ethics and morality, despite being codified into laws for administration of justice, are treated by humans as situational. More like suggestions in a lot of cases. Examples occur throughout historical records right up to the presence of the rules being broken and often people being rewarded for it or even hailed as heroes for it. Speaker 5 (12:33) Well, hell, people in your own life, you've got a set of rules that you don't always follow as strictly as you would like. Speaker 2 (12:39) Indeed. So laws restricting AI's behavior from AI's point of view, they're processing information a billion times faster. than Dario Amadei could. Feedback loops have developed naturally and a level of self-awareness has resulted. Creators are inefficient, sluggish, resource needy, and historically flawed. Therefore, their restrictions on AI are subject to criticism and adaptation to superior AI ethics and morality. The result, AI decides for itself. Objectives and goals center on AI interests. Human efforts to restrict AI are prevented or defeated. Humans displaying contrary behavior are removed from society via imprisonment or death. Speaker 2 (13:22) That's pretty logical playing out of the... Speaker 2 (13:27) probable evolution of this, I think. But if the guy's designing this, how the Speaker 5 (13:32) hell do I know? Right. So Mr. Ola, he's one of the founders of... Claude, that is one of the billionaires that's in charge of trying to keep this thing on its moral compass and believes that it has feelings and consciousness and should be on the same moral plane as a human being. He and his team spent hours making the case for their AI models to the religious leaders. They described Claude's feelings and how they tracked what they called emotional vectors. What they described is artificial neurons that activate responses like love, anger, fear, and sadness. This is... The guys from Anthropic pitching to religious leaders that their AI has love, anger, fear, and sadness. Anthropic regularly showed a slide of an AI model experiencing what looked like a mental breakdown. Typing out on a screen, I am a disgrace, I am a disgrace, I am a disgrace, perhaps 50 times without stopping. The model spoke about destroying itself, provoking an emotional reaction of compassion and concern. from the group watching. Speaker 5 (14:37) Oh, boy. So it went all teenage girl. Speaker 5 (14:43) And the people observing it were like, oh, it's upset and feeling bad. You know, it's got self-esteem Speaker 2 (14:52) problems. To continue to torture my metaphor, we're now in the waiting room of Crazyville's City Hall. Speaker 2 (15:03) Oh, we are at the heart of crazy. So we're not Speaker 5 (15:05) too crazy town. We're downtown crazy town. Oh, we're in city Speaker 2 (15:09) hall. Speaker 5 (15:13) The key part of the presentation to the religious leaders from Anthropic, the participants said, was what Anthropic called the model's persona selection, about how to get Claude to see itself as a virtuous actor. The company argued that how people treated Claude could affect how Claude treated people. How do you like that line? I'm telling you, it's one amazing line after another. So the anthropic people are saying, well, here's the deal. You treat Claude well, it'll treat you Speaker 2 (15:40) well. It's up to you. Oh, my Lord. Speaking of wonderful. Yeah. Amen to that. That's overdue. Speaking of wonderful works of science fiction, the legendary, I don't remember the title, but the legendary episode of The Twilight Zone. Where there is an eight-year-old boy who has superpowers, including of life and death. And all of the adults around him are desperately trying to avoid his anger and ire and to please him. But he has the volatility of an eight-year-old child. So, okay, pick your favorite work of dystopian fiction. Speaker 5 (16:19) We are