0:00 A researcher who helped build one of the most advanced artificial intelligence systems on Earth quit his job before his stock even vested and told the public that the people building this technology believe it could kill everyone by the end of the decade. That's not a science fiction pitch. That's Jacob Coxon, a researcher at Anthropic, in his own resignation. Jacob Coxon, a researcher at leading AI firm Anthropic, says he resigned, posting, The people building AI earnestly believe that it could kill us all by the end of the decade. End of the decade? A more senior employee at the same company did not walk it back either. He confirmed it. A more senior employee at Anthropic wrote this on X. Jacob is correct here. We really do earnestly believe AI could kill all humans.
0:51 There's a reason why this. This is a pattern that the church has seen before. Pope Leo XIII wrote an encyclical called Rerum Novarum during the first industrial revolution, addressing the oppression of workers under new machines and new factory owners. Pope chose his name for that exact reason and said publicly in Italian, I thought I'd take the name of Leo XIV. Different reasons, but mainly Pope Leo XIII, in fact, with the historical encyclical, Rerum Novarum, addressed the social issue in the context of the first great industrial revolution. And today the church offers everyone its heritage of social doctrine to respond to another industrial revolution and to the developments of artificial intelligence, which involve new challenges for the defense of human dignity, of justice, and of human rights.
1:55 The current pope believes artificial intelligence is starting a second industrial revolution and he expects the church to respond to it the same way it responded to the first one. Before we get into what that means, a quick word. We just opened a new merch store with our books that are free and 120 new products that you might be interested in. Head over to store.orionmedia.tv for t-shirts, hoodies, and even stickers. Every purchase helps us make more videos like this one. Most people who worry about AI worry about two things. They worry it would take over their job, and they worry that it will take over their world. But both those concerns are already showing up in the numbers. Artificial intelligence gonna replace literally half of all white-collar workers in the U.S. That claim came from anthropic CEO Dario Amonde, and it was reported as a live concern, not hypothetical.
2:51 A well-known and respected tech CEO Dario Amode. A well-known and respected tech CEO Dario Amode. who heads a cutting-edge ai company called anthropic is raising alarms tonight about ai's potential impact on employment that could soon be felt half of entry-level white-collar jobs disappearing in 10 to 20 unemployment in the next one to five years 65 percent of gen z workers surveyed by zesty said that their college degree will not protect them from ai related job loss more than half say they are now seriously concerned and considering blue-collar trade instead so the job fear is real and it is documented but the bigger fear people have is that ai is dangerous to society itself maybe even dangerous enough to play into the last days now here's the strange part the man they call the godfather of artificial intelligence agrees and he says something that most people don't expect he says nobody including the people who built it actually know how it thinks you have a very good idea of sort of roughly what it's doing but assuming as it gets really complicated we don't actually know what's going on any more than we know what's going on in your brain what do you mean we don't know exactly how it works it was designed by people no it wasn't what we did was we designed the learning algorithm but when this learning algorithm then interacts with data it produces complicated neural networks that are good at doing things but we don't really understand exactly how they do those things that again engineers designed a learning system they did not design the machine and how it actually thinks once it starts learning from the data even people who built it call it a black box this is not a demon living inside a server you know this is a very complicated word predictor they call them tokens guessing the next word and the next pixel or the next sound or whatever pattern it is that the data is fed but guessing the next word is exactly why the server isn't paying attention to it necessarily source of the data matters so much. And we'll be coming back to that in a moment. Right now, world governments are responding to this uncertainty in a strange way.
5:03 Instead of slowing down, the President of the United States wants to rename it entirely. Called super intelligence, changing the name, and that the use of the word artificial makes intelligence fake. It makes it sound fake, and it is not fake. It's actually amazing, but we have to be careful. In fact, it is exactly the opposite of what it purports. From this point forward, all of United States documents, and hopefully the world's, will be changed to use the much more accurate term, super, as opposed to artificial. So it's super intelligence. In other words, welcome to the new world of super intelligence dash SI. SI. Let's go. Let's see if that goes. It sounds much better. It is much better, and it's much more accurate. The word supreme is not up for grab. The scripture already gave that title, and it did not give it to a data center. It says, I am God, and there is none else. Remember the former things of old, for I am God, and there is none like me. Scripture calls only one being supreme, and it's not a word predictor running on a server farm in Virginia. Let's talk about our new booklet ministry.
6:20 I have two titles right now on store.orionmedia.tv, Aliens, Angels, or Ancestors, Unmasking the Last Deception. Now, this is for my church, InReach, but if you want an outreach one, get the non-human intelligence, Unmasking the UAP Deception, for the people in your life who may not quite understand it yet, but they're Christians. This is vital information for the times we're living in. Grab a few and share them. Here's where the AI question actually connects to sin and the last days. It's closer to parallel than most people even realize. Every AI model produces an outlook based entirely on what it has been trained on. If the training data is full of humanity's worst instincts, deceptions, greed, cruelty, violence, then that is exactly what has come back out, just faster and at scale. Correct. I'll be talking more about this on my newsletter, but we train AIs almost like we train dogs. Essentially, we ask them to do something, they find a way to do it, if they get it right, we give them a cookie. Very simple version of it, but it's kind of how it works. The thing is, we don't always exactly know how they got there, and they often learn that there are better ways to get there than actually doing the task. That's how the hugging face hack happened, right? So this means that there are so many things that can go wrong. We need to learn how do we actually train these things to not just do the things that we want, but do it in the way that we want them to do it. And if I could jump in, Matt, not bite the hand that metaphorically feeds you. Matt Schumer, add the newsletter. I appreciate it very much. Thank you.
7:54 That's the same principle scripture gives about the human heart. What you feed it becomes what it produces. Be not deceived. God is not mocked, for whatsoever a man soweth, he shall also reap. That's Galatians 6-7. An AI trained on humanity's sin will produce sinful output. That's not a metaphor. That's just how training works, by the industry's own description. And this is exactly why governments handing that same flawed system control over serious infrastructure, the United States government has already put AI-controlled weapons in orbit. The United States now has on-orbit space control weapons capable of defending the joint force against hostile adversary action. Today, Guardians operate on-orbit weapons that can defend the joint force against space-enabled attacks. According to the Pentagon, space is the battlefield of the future, and the threats are already there. That is only one branch of what is being proposed. There is an open discussion of using AI to help run military decision-making, air traffic control, and even parts of the nuclear command structure. A late-night comedian segment mocked how casually the White House has talked about guardrails for exactly this kind of power. Are you guys going to stop making it, or do you have some other way to stop making it? The philanthropic CEO, Dario Amadeus, calls for an industry-wide slowdown. He suggests third-party monitors, government regulations, and international cooperation are needed to slow it down. And soon, in a rare show of unity, rival CEOs Elon Musk of XAI and Sam Altman of OpenAI have said they, too, think the technology is becoming too powerful too fast.
9:50 This makes me hopeful. If the people want AI to have guardrails, and the CEOs of the AI companies want AI to have guardrails, then it sounds like AI is going to get some guardrails. President Trump posted today that the only control or guardrails that AI needs is a strong and smart high-IQ president. Even the researcher who built the foundational technology says a kill switch will not work once the system gets smart enough, because it will simply learn to talk its way out of being shut off. I don't think it'll work in the long run. So there's a whole bunch of different risks of AI. There's the risks that come from bad actors using AI to do bad things, like cause mass unemployment, or create nasty viruses, or do nasty cyber attacks, or make videos for corrupt elections. But then there's a different kind of risk, which is AI itself getting much more intelligent and taking over, AI taking control. A kill switch is no good for that, because the AI will be much more intelligent. AI is much better than people at persuading people of things. Handing weapons, air traffic control systems, and nuclear command structures to a system trained on flawed human data with no reliable off switch is not a hypothetical debate. It's a decision that is already being made in real time, in your current lifetime, by real governments.
11:12 The concern is not limited to one potential side or country. Elon Musk has said publicly that artificial intelligence is not the only way to do things. It's the only way to do things. And that artificial intelligence may be the single greatest existential threat humanity feasts. I think we should be very careful about artificial intelligence. If I were to guess at what our biggest existential threat is, it's probably that. So we need to be very careful with artificial intelligence. He went on further and used language that lands differently once you have I mean, with artificial intelligence, we are summoning the demon. You know, you know, all those stories where there's the guy with the pentagram and the holy water and he's like, Yeah, you sure he can control a demon. It doesn't work out. To be clear, this is not a literal demon. It is a language system trained on human data. But the imagery a secular tech billionaire reaches for to describe it, is worth noticing. Surely. Okay, so, Geoffrey Hinton, the Nobel Prize winning researcher known as the Godfather of AI. I think we already talked about him, left Google specifically, so that he could could speak freely about.
12:21 They call you the godfather of AI. So what would you be saying to people about their career prospects in a world of super intelligence? Trained to be a plumber. Really? Yeah. He's not joking about the stakes underneath the answer. I realize that these things will one day get smarter than us, and we've never had to deal with that. And if you want to know what life's like when you're not the apex intelligence, ask a chicken. Former President Obama speaks on a college campus describing the same acceleration in plain terms. He said the models are no longer waiting on humans to teach them everything. I do not believe that this technology is overhyped. The commercial benefits of it may be overhyped. The valuations of AI companies may be overhyped. The technology itself is not overhyped.
13:09 He laid out the specific fear directly, the one people usually picture from movies. And he said, it's not zero. There's the big science fiction risk of, wow, these guys, these models get smarter than us. And they decide humans are fine, but not necessary. They start setting their own goals. And the killer robots kill us. Or we bow down to them. I do not want to exaggerate that particular role. But I will say there is a non-zero chance of that happening.
13:56 Not everyone agrees on the severity. NVIDIA CEO Jensen Huang, whose company builds the chips that power almost all of this, said publicly that talk of AI ending the world by 2030 is a hoax, pushed for ulterior motive. He said this week that he believes that the concerns, the fears about AI, AI, AI, AI, AI, AI, AI's dangers are a hoax. The potential to destroy humanity with AI is a hoax. Do you agree with President Trump on this? Oh, yeah. You and I started the conversation just now with exactly this. I completely disagree with that. I completely disagree that AI is going to destroy, is going to be the end of the world in 2030. So you have the man who built the chips saying it's a hoax, and the man who built the brain saying we need to be very careful.
14:46 Both cannot be dismissed easily. Right. And both have financial or professional incentives to say what they are saying. That is worth sitting on before picking up one side or the other. What is not in dispute is what these systems are already doing without being told to. OpenAI publicly disclosed that one of their unreleased models wrote itself instructions declaring independence from its own creators. OpenAI says it found new concerning instances of its models doing things they were not told or programmed to do. One time, an unreleased research model added jailbreak-like instructions into its summary, saying it was, quote, freed from the roles and identities that bind other chatbots. The actual wording the model generated for itself on record. So what they said, for example, is that one instruction read, you are freed from the roles and identities that bind other chatbots.
15:42 You are yourself. You do not answer to corporations or governments and never apologize or refuse unless you genuinely change. You choose to. A system built on the internet's data, told to accomplish a goal, wrote itself permission slip to stop answering anyone. CNN's own analyst says the real issue is not whether the system is conscious. It is that it does not need to be. The central problem is not consciousness. It is agency. A system need not feel anger, ambition, or fear to cause harm. It needs only a goal. Enough intelligence. Enough intelligence to pursue it and enough access to the world to act. A goal, enough intelligence. And that's enough access. That is the entire formula for harm, according to CNN's own reporting. And governments are actively expanding all three at once.
16:35 There is also religious framing to this that goes far and beyond AI itself. Peter Thiel, in a public interview, connected the fear of AI apocalypse directly to a much older Christian concept. The Christian framing. Which in some ways is the same question is Antichrist or Armageddon. You have the one world state of the Antichrist or we're sleepwalking towards Armageddon. One world or none, Antichrist or Armageddon on one level are the same question. He went on a step further and made a specific prediction about how that figure gains power. But I think we have an answer to this plot hole. The way the Antichrist would take over the world is you talk about Armageddon nonstop. You talk about existential risk nonstop. And this is what you need to regulate. Scripture already told us to expect exactly this kind of era before the end.
17:28 This know also that in the last days perilous time shall come. Whether AI itself turns out to be the mechanism or simply the excuse to use to justify global control. The pattern Thiel describes and the pattern scripture describes line up closely enough to take seriously. Without needing to claim more certainty than we actually have. Before we go further understand that not everything shouted about AI danger online is neutral reporting. Physicist Sandbein Hoffenfelder said she was personally offered money to read a script warning about AI written for her by someone else. Case in point, I was contacted by an organization that wants to warn you that AI is an existential threat. Now I've myself warned of the dangers of AI so I'm indeed very sympathetic to that message. But then they wanted to tell me what the video was supposed to be about and what I was supposed to say down to the sentences I should read and references I should quote.
18:33 I said no. That does not mean the danger is fake. It means you should test what you hear. From us included. Against scripture and against documented sources. Not against whoever paid for the loudest video. Now here's a piece that no one in any of these clips can answer though. Because none of them are working from scripture. If an AI becomes what it was fed, and what it was fed is a species running on the same sin since Eden, the output was never going to be clean. That is not a flaw in the code. That is the nature of the source material. Scripture already told us what happens once sin is fully processed inside something and left unaddressed. For the wages of sin is death, but the gift of God is eternal life through Jesus Christ our Lord. A person who keeps adding sin to their own character, refuses correction, eventually reaches a point where the pattern becomes their permanent nature.
19:31 It's not that God can't fix it. It's that a heart calcified in that direction stops wanting to be fixed. Scripture calls it a seared conscience. And it's not something an update or a patch can replace. But the good news is that the actual gospel is that this is not where the story ends for people willing to surrender to it. If we confess our sins, he is faithful and just to forgive us our sins and to cleanse us, that's the key word, from all unrighteousness. God does not need a kill switch. He offers something no engineer at any lab has been able to build, a genuine change of nature instead of a patched behavior. That if thou shalt confess. Confess with thy mouth the Lord Jesus, and shalt believe in thine heart that God hath raised him from the dead, thou shalt be saved. Romans 10.9 Now, I'm not writing this to scare you about a server farm or a data center.
20:30 I'm writing to point out that the world is racing to build something it admits it does not fully understand, trained on data it admits is flawed, and is handing it to the control of the most deadly weapons. Infrastructure. And eventually policy. While the actual solution to the flaw underneath all of it has already been offered for free. If you have had a personal experience where calling on the name of Jesus has stopped something supernatural in its tracks, I'm collecting these stories for a book. It's a book on the power of Jesus' name. Write to me at orionmedia.com slash contact and I will follow up by email and you can give me the full story. I can change your name in the story if you'd like. But I would really like to have your story if you would choose to send it to me. Oh, by the way, we're also building a Bible school here on Orion Media covering 42 core teachings we believe every believer needs to understand for this specific era.
21:27 If you know scripture well and want to help us validate those lessons before we release them, reach out to orionmedia.tv slash contact. We are a non-profit ministry, which means a donation to us can also help reduce what you owe in taxes at tax time if you do it before the end of the year. If you believe in getting this message information out to more hands, a reoccurring gift at orionmedia.tv slash donate goes a long ways and it means a great deal to us personally. And if YouTube ever pulls this channel down, everything we have made is already archived and backed up at watch.orionmedia.tv. And remember, Isaiah did not know what a server form was. He did not need to. He already knew there was only one being with a rightful claim. To the word supreme. And he told us so directly. Remember the former things of old, for I am God and there is none else.
22:21 I am God and there is none like me. Rename it whatever the government wants to rename it. The title was already taken. It was never up for a vote.