Latest / Crisis Point / Debunking AI Fears

Transcript

Debunking AI Fears · Crisis Point

• 0:00
  Foreign.
• 0:12
  Welcome to Crisis Point. I'm Eric Sammons. Today what I want to talk about is a topic on a lot of people's minds and that is artificial intelligence. Actually, I think it's more accurate to say I want to talk about some of the narratives that that are surrounding artificial intelligence in kind of the national conversation today that you see on social media, on the news and other places.
• 0:36
  I've noticed that there is a growing anti AI movement and it's starting to focus on certain aspects of artificial intelligence, particularly data centers, also quote unquote rogue AI as well as just AI eventually being becoming like sentient or something or a singularity or something like that, a kind of rogue AI which leads to that.
• 1:01
  The thing is, is this podcast episode is not going to be a pro AI podcast as much as an anti anti AI podcast because I have some serious concerns about artificial intelligence. I think there's some serious dangers surrounding artificial intelligence and its development.
• 1:21
  Yet at the same time I have seen that a lot of the arguments being made today against artificial intelligence, particularly against data centers and things like that, are often based on falsehoods. They're emotional and they're irrational. So I just want to go through some of that today, kind of not talk about, I'm not going to talk about the real dangers of AI. I will talk about that though. My upcoming book. I just got the COVID for my book this week. I'm very excited about here. I'll pull up on the screen the Catholic Guide to Artificial Intelligence Preserve Preserving the Distinction between man and Machine. I think it's a pretty cool looking cover.
• 2:00
  If you're just listening this, you can't really see it, but it's cool. So look it up and I'll put a link in the show notes to where you can pre order the book. I talk a lot about the dangers of AI in this book, but it's not what people are saying right now. A lot of what people are saying and I also want to say that it's kind of understandable that there's an anti AI movement afoot, particularly because of the statements of a lot of the AI leaders, the CEOs of AI companies, of anthropic, of OpenAI of other places. They make some claims that frankly are terrifying if they're true.
• 2:37
  The problem is they're usually not true. They're just marketing frankly to get more investors.
• 2:42
  So again, I'm not here in this podcast to defend everything about artificial intelligence. I'm not a fanboy for artificial intelligence. I do think there's a lot of benefits to society that AI can bring. And I also think there's some dangers. But I do want to debunk some of these frankly kind of stupid arguments against AI and against data centers that I'm seeing out there.
• 3:06
  So let me go ahead and start with the data center one because that seems to be the one in the conversation right now. And I think it's in the conversation right now because the upcoming election, the midterm elections, I think it's mostly driven by politics and people are taking advantage of it in order to try to get you not to vote for the Republicans, whatever the case may be.
• 3:27
  I am not a Trump guy who's like, oh yeah, data centers are great for economy and you're stupid if you don't want them.
• 3:35
  I do think there are some legitimate concerns with data centers. And I'll get into what I think the reality of data centers here in a minute.
• 3:44
  But I think ultimately a lot of the antipathy towards data centers, the anger towards data centers is really because data centers are a proxy, a physical proxy for AI. How do you argue against, how do you stop something that's kind of ethereal out in the cloud on the Internet? You know, how do you even fight against that? How do you fight against billionaire AI leaders and CEOs and things like that, These powerful companies, if you're a regular person, it's not very difficult. But if all of a sudden they want install a data center, set up a data center near you, that's a way you can fight against that. And I actually understand that. A argument that you're not really supportive of a lot of what AI companies are doing or AI and so therefore you're going to oppose them where you can, which is in the data centers. So I do understand that argument. However, it has led to a lot of misinformation, a lot of false claims about data centers and things of that nature.
• 4:47
  And I think it's mostly politically driven right now. I think some of it is probably might be from China. I'm not a guy who's like, oh yeah, it's some China psyop or something like that. But I wouldn't be surprised if they're encouraging an anti data center movement here in America. I wouldn't be surprised if they're behind some of it. I don't think they're the driving force behind it, but I wouldn't be surprised if they have something to do with it. Here's a little fun fact for you though.
• 5:11
  I helped build the data Center in the 1990s I worked for a web hosting company, one of the first web hosting companies in the world frankly. We helped develop one of the first web hosting company software programs. And we had a date, we needed a data center. We first just rented out from somebody and the first thing we did is in Balt, this was in Baltimore. We simply took over a floor of a high rise building that had been used as a bank.
• 5:38
  I think a bank had it for their computers, servers and things like that. So we converted it to a web hosting data center.
• 5:45
  And so I mean I wasn't involved in the physical building up but I was at the company then. I was involved in helping set it up, things like that.
• 5:52
  And so I helped build a data center. Nobody knew, nobody cared. It didn't do anything to the, to the area or anything that I was just a. It was one floor of a building that was basically a bunch of server racks. And so we had to spend a lot of time making sure it was stay cool and electricity was okay and all that stuff. And the Internet connection obviously was okay. It all worked out.
• 6:12
  I also was tasked with finding a data center in Northern Virginia. This is around 1999 I think maybe 1999. Around that time, 2000 around then I went down to Northern Virginia to try find another data center we could lease space in. And then I continue. When I left that company, I had my own software development company. We had to have a date, we had to use. We co located some servers in a data center. So I'd be down there sometimes working on those. So I'm very familiar with data centers and which is why I think it's a little bit funny that everybody seems often to be anti data center. I saw somebody say data centers are bad and that's all he said. It was a Catholic person who said they were an Internet crusader. And we have to realize something. Data centers have been with us for decades. Since the 1990s there probably were maybe a few data government type data centers, maybe in corporate data centers, maybe before that. But really data centers as we know them really started up in 1990s. Why? Why? Because that's when the Internet broke free and everybody started using the Internet.
• 7:17
  Because just in case people aren't completely clear about this, I this is like I've known this for 30 some years so I forget that some people might not know this.
• 7:25
  The entire Internet runs on data centers because the way it works is technically speaking I could have a computer in my house that is connected to the Internet and hosts a website and people could go to that website that on computer in my house. However, almost no websites out there do that because your connection isn't as reliable. The, the computer you're going to use probably isn't as powerful and you're gonna have downtime, all those things.
• 7:50
  So what you do is you put your website on a server, which is just a fancy name for a powerful computer.
• 7:57
  Add a data center. Because a data center has a very high powered Internet connection and it has the ability that it has all these things in place to keep it online all the time. So you have 100% uptime or at least 99.9% uptime.
• 8:13
  That's how everything runs on the Internet. Everything is. You know, when you talk about something in the cloud, I remember I was actually working on software development, Web based software development for the word cloud came out. I remember when I first started using the word cloud I thought it was kind of weird. I didn't know what it meant. But then I figured I was like, oh, that's what I've been doing for years.
• 8:31
  The cloud, it just means it resides on a physical server and at a physical data center somewhere. That's all that means when something's in the cloud, that means it's on a physical server at a physical data center. Now it's a lot more complex than this in many ways because a lot of times you'll have a website's gonna have multiple locations and things like that. But I'm trying to keep it simple right now, not trying to get too technical.
• 8:53
  And so over time, so many things have gone to the Internet that we couldn't survive without. Data centers.
• 9:03
  The way we're set up right now, it would be a literal apocalypse. If you got rid of all the data centers tomorrow, somehow all data centers disappeared. Tomorrow it would be a literal apocalypse for the world, especially in a first world country like America.
• 9:18
  Because what would happen is, and I kind of saw somebody post some of this, I'm going to go through a little bit of this immediately. As soon as that happens, you have a near total global communications blackout instantly including a large chunk of landlines and cellular. So basically you're not going to be able to communicate with almost anybody immediately if there's no data centers very quickly all payments and exchanges are frozen. A lot of 91 1. Routing and dispatch are data center hosted so you can't call for help if something happens.
• 9:50
  You're going to be. Planes are not going to be able to know where they're going. They're not going to have, you know, flight planning and weather, knowing what's going on.
• 9:57
  You're going to have everything grounded. You're not going to have any air flights, anything like that.
• 10:02
  Grid operators will fly blind. Without data centers hosted energy management, large black regional blackouts will happen.
• 10:10
  You know, you're going to basically no, no communications, no payments, no electricity in a lot of places.
• 10:18
  You know, the, the police can't dispatch people to, to problem areas.
• 10:24
  Hospitals are not going to be able to stay on. Basically, they're not going to be able to keep doing what they're doing. All their, all their systems are computerized and they're all going to be down.
• 10:33
  Like I said. So blackouts, you have, you know, police can't respond, you have hospitals down. I mean, ultimately, within about 24 hours, 28 to 48 hours, you're going to have just mass chaos, death, destruction. I mean, just. It's going to be awful.
• 10:49
  So when you say data centers are bad, what you're basically saying is like, our entire world is building something bad. And you might think that, but we can't just take them away.
• 11:01
  It's like kind of saying that cars are bad, okay, but cars aren't going anywhere tomorrow or saying electricity is bad. Okay, what are you gonna, what do we do about that? Unless you are willing to, to go full Amish, then you, you basically, you're, you're living off of data centers. And so I, I have a problem with the unqualified attack on data centers. Now, I have not talked about the type of data centers that host AI, which is. It's the same thing in one sense, but not. It's a little bit more, it's a little different in another sense. We're talking because a lot of data centers that host the Internet, you don't even know they are there.
• 11:45
  There are data centers all over the world, all over the country.
• 11:49
  And like, for example, Northern Virginia, Loudoun county, the richest county in the world, by the way. It also has the most data centers. That's where I went when I was looking for a second data center back in the late 90s. They've been the data center capital of the world for decades.
• 12:01
  But the point here is being against data centers. It's like being against electricity, frankly. And okay, if you're against electricity, you're Amish, fine, just be consistent with that. You shouldn't be watching this podcast, live out your convictions and go completely offline, completely off grid.
• 12:19
  Because data centers are necessary for modern life, whether we like it or not. They're necessary but the real, what we're starting to see is people don't like the new data centers coming out. They often go out in rural areas and more farmland, things like that.
• 12:35
  And I get that on some level. But a lot of the arguments against it are frankly just ill informed. They're based on lies or misinformation.
• 12:44
  For example, the whole water thing, that they consume all this water or something like that, that's simply not true. It's simply not true.
• 12:54
  I mean, there was this whole meme going around at one point that every query of a chatbot takes the equivalent of one bottle of water.
• 13:03
  No, it doesn't. At most it takes maybe about four or five drops of water is all it takes. And you might have been hearing, some people have been saying, and this is true, data centers in the country combined take up, use less water than all the golf courses in the country combined.
• 13:19
  It takes more to grow almonds and harvest them. And all that. Takes more, way more water to do that. Way more water, millions more water than it does to run data centers.
• 13:30
  And the truth is this is just improving. This is one of the things I always notice when people start attacking some technology, it's using some energy is you do realize it quickly gets more energy efficient because remember, the companies are actually incentivized to use less energy, use less electricity because that makes it cheaper for them. They actually want to use less. So they're developing technology that uses less and less water, less and less power.
• 13:56
  And right now, for example, new data centers, they're all using recycled water. They're using very, they have a very small footprint when it comes to water usage. You'll still see on, on X or wherever people claiming they're using Apple's water. It's simply not true. And the same thing is basically true of electricity.
• 14:13
  Those are, those concerns are way overhyped. The idea, like there was the idea that all of a sudden if a data center comes into your area, all your electrical rates are going to be hiked. No, that's simply not true. It's just simply not true. Now I will say it is possible a data center could be set up poorly. The government could give it certain, you know, not, not allow it to run properly or something like that. But I'm just saying the way most data centers are built today, these water and electricity concerns are simply overblown. They're not, they're not real.
• 14:45
  And then there's the, the whole thing about they, they don't produce jobs.
• 14:50
  Now first of all, I just want to make this very clear.
• 14:52
  Often a data center might be going into an area that has no jobs.
• 14:58
  So even if a data center came in and it produced, let's say 20 jobs and that's it, that's 20 more than you did have. That's 20 more than the town did have.
• 15:10
  So the idea that there's like a connection between land usage and number of jobs is a little bit kind of weird. Because if it's using land again, I'm fine. If, if there's like farm pristine farmland that's being taken over by land centers, you don't think it should be there, fine. But there's a lot of places in the country, like especially the old steel mill towns, old industrial, like 20th century industrial towns that are basically falling on, on hard times because industry up and left, it is decimated.
• 15:41
  Converting those areas into data centers is a net boon. It is. I mean it brings in jobs. First of all, there is all the jobs involved in constructing the data center, which is, it's all the construction jobs. In fact, I saw something, I didn't write this stat down, but it was some huge percentage of all construction jobs over the past few years have been data center driven.
• 16:02
  And so a lot of people in construction work, they're, they're, they're making a living, feeding their families, taking care of their kids because of data centers, new data centers. And it's not like all those jobs just go away. That after it's up and running a data center is like just run itself. That's simply not true. I know this from helping run a data center.
• 16:19
  You still have the maintenance of it. I mean just even things like janitorial for the data center, things like that, the servers being there, you know, you have to be there 24, seven often.
• 16:31
  Plus when you want to expand the data center, you want to update it, you got new construction maybe coming in, new jobs for that.
• 16:38
  There's a lot more than just the initial build. It definitely brings jobs. Does it bring as many per capita like per land usage as maybe a steel mill did? No, but the steel mill doesn't exist. The steel mill. You know how many jobs you're getting from the steel mill? Zero.
• 16:55
  Because it doesn't exist. The data center does. And it can bring jobs in.
• 16:59
  And in fact, it's interesting. Labor unions have actually gone come out and they're pro data center because they recognize it's not taking jobs from their blue collar workers, it's giving jobs to their blue collar workers. Jobs they wouldn't not have.
• 17:14
  Because remember when you Say, you know, take away. I, I know some people when they, they say things like take away data centers, they don't have an alternative. They're like, oh, we don't want a data center here. Our town is dying. We don't want data center here. Okay, well what is your solution then for jobs? And they don't really have one. They, or they just live in a fantasy world thinking that maybe manufacturing is going to come back or something like that. It's just simply not going to.
• 17:40
  And so these data centers can be good for a community. In fact, there's a great example that's used often, but there's other examples than just this one. Quincy, Washington is a town.
• 17:50
  Data center started coming in about 15 to 20 years ago. Their poverty rate in Quincy, Washington went from 29.4% in 2012 to 6.2% in 2024. So in 12 years the, their, their poverty rate dramatically was reduced. They have 30 data centers there, by the way.
• 18:12
  Also, an estimated 57% of Quincy's property taxes are from data centers. And that revenue funded a $120 million high school, a new hospital, a library, police and fire stations, paved sidewalks, and upgrade sewage and water treatment, all without raising the resident, the residents tax rates. All that came because of data centers.
• 18:39
  The data center directly employ around 900 people locally, which is more than 10% of the town's population.
• 18:46
  So 900 people in 30 data centers. If you do the math, that's about 30 people per data center. That's actually not that bad.
• 18:55
  But each of these jobs are estimated to generate four to six additional jobs in construction, retail and local services. Because remember, if they're not working there and there's nowhere else to work, they leave the community. But if they stay there, that means they're buying a house there, that means they're shopping there, that means they're going out to restaurants there, that means, that means all the local services have to support those 900 people working those 900 jobs that wouldn't be there if not for the data centers.
• 19:21
  So it really is, it can be a benefit.
• 19:24
  And so this boogeyman of data centers started off as just an emotional. I don't like them because I'm very uncomfortable with AI Therefore I'm. This is the one way I can resist. I'm going to be against data centers. And then people just started. They might use old studies or bad studies or just might make stuff up, frankly, about all the terrible things the data center does when it comes in.
• 19:46
  Frankly, it doesn't now I'm not, like I said, I'm not saying data centers should be everywhere. I understand certain areas. It's not good to have data centers in necessary like for example good farmland, something like that. But a lot of these industrial like rust belt towns that are just frankly dying, I think that is a great way to revive them is through data center. It's, it's the equivalent of what happened to those towns years ago Dec, you know, century ago when these steel mills and all those things came into town, brought a lot of jobs and built the community. Things like that.
• 20:20
  Data centers can do that today.
• 20:23
  Now there is one concern I have read about and I've studied and it is a legitimate one. I think it's overblown, but I do think it's legitimate and that's the noise. There is, there has been measured hum associated with data centers. I think this is overblown a bit, but I do think it should be taken into consideration.
• 20:42
  I do think potentially local communities could, could demand the data centers do something to reduce the noise pollution. At the same time I do want to say there's tons of types of businesses that produce noise. I actually live in an area that's kind of a cross between the suburbs and farm country. Like we're, the suburbs have kind of come out towards us but we're still a little bit farm country.
• 21:06
  But there's also light industrial here. There's a number of light industrial parks nearby and I can hear them if I go outside like on a weekday, like Wednesday afternoon or something like that. I can hear them with the work they're doing. Do I like it? No. Is it that big a deal? No, it's not that big a deal.
• 21:25
  So at the same time I, I, I do think it's like there is noise associated with work and so if there is a noise associated data centers, I don't think necessarily it's like okay, that therefore they're somehow anti human, they shouldn't exist because they do bring about so many benefits beyond just data centers in general.
• 21:45
  But yes, data centers should be done right. It's not saying that local government should just roll out the red carpet, give them a bunch of tax breaks, let them do it wherever they want.
• 21:54
  No, they should make sure that the data center is built properly, that they, the energy usage is considered beforehand and the water usage is going to be done the right way.
• 22:05
  Maybe noise ordinances are, you know, are put in place so it can't be too loud, all that type of stuff. I definitely think that's possible. But I really do think data centers can be something that revives a lot of towns, and it has already done that. And if you have this argument against data centers, and it's not just an emotional one, tell me, though, what you would replace it with to bring about jobs to a lot of these towns. I mean, because I think a lot of us, we don't like the fact that everything is in big urban areas. We want to have smaller towns around and things like that, where you can live. Well, let's be living reality, which is a lot of smaller towns, they have no jobs for young people, which is why young people leave them and they end up going to bigger towns, bigger cities, because that's all they can. The only place they can find a job. Well, if there's a local data center, perhaps they could get a job there where they couldn't get otherwise, a job that pays well.
• 23:01
  So I think, like I said, when it comes to the, the fears of AI, I think data centers is not one of them. I mean, it's not a real one. I think it's really emotionally driven. You see, a lot of the left is really pushing this because they just think it's a weak spot. Again, they don't care. They just think it's a weak spot against Republicans.
• 23:22
  I do think some Republicans, when they support them, they're. They don't really look at the whole picture and they're just kind of blindly like capitalism, whatever. And I'm not saying that I think there should be concerns, there should be regulations potentially associated with data centers, but I just think banning them is just, it just is really a dumb solution. And like, and it doesn't recognize. We've had data centers with us for years, and it's just a bad word now, but literally six months ago, a year ago, nobody thought anything about the data center. That was right down the street. I remember when Trump came out for data centers the other day, which, you know, he does what he does.
• 23:55
  I think it was AOC and some others, like, oh, I bet you wouldn't like it. How about days here in Mar a Lago? Well, then they found out, oh, there is a data center in Mar a Lago. Like, not very far. I mean, not very far. Just a few, a mile or two away. So there is a data center in the community. Because that's the thing. There are a lot of places. You just don't know it. You've, you haven't been trained yet to hate them until this year, even though they were all around you and you didn't care. And there's one guy who was saying he's going to put up a fuss because a data center was going to move right, right, right down the street from him and stuff like that. And somebody pointed out there'd been a data center right down the street from him for years and he didn't even notice it. And I think that's, that's the reality of these data centers.
• 24:34
  The second big fear I want to talk about, so that's data centers. The second big fear I want to talk about when it comes to AI is rogue AI.
• 24:41
  This has gotten the news particularly in the last couple months because there's been reports of rogue AI where AIs have gone, I mean they've gone rogue, they've kind of broken out of containment. And if you notice one thing about all the articles that talk about this, they do something that you, you need to be careful, you need to have certain glasses on when you read articles by AI in the mainstream media.
• 25:04
  Excuse me.
• 25:06
  And that is recognize when they are adding intentionality and thought, real thought and belief to an intention to AI. When they say like the AI believed it could do this, the I intended to do this.
• 25:22
  AI does not have that.
• 25:24
  Modern AI definitely doesn't. I don't think AI ever will. It does not have intentions, it does not have desires. It simply does what it's programmed to do.
• 25:33
  And so there really is no such thing as rogue AI in the sense that there's AI that, that goes against what it's told to do and it kind of goes off on its own, does its own thing that just simply has never happened.
• 25:47
  And in fact the funny thing is, is if you look there's a lot of, there's tons of AI out there right now. We tend to focus on the chat, the LLMs, the large language models, the chat bots, and we talk about them going rogue. Well, there's lots of AI that never go rogue. They're never even claim to go rogue. I mean one example is like self driving cars. There's never been an instance, nobody said oh my gosh, that Tesla just went rogue and decided to go drive somewhere else. Never been a case of that. The stuff like DeepMind's AlphaFold, which basically it actually earned its creators a Nobel prize for its ability to predict the folding behavior of proteins. Nobody talks about AlphaFold going rogue. There's no been stories about that.
• 26:33
  You know, game playing AIs AIs are created to play games through like AlphaGo or the DeepMind. I'm sorry, whatever the one was called it Deep Blue, the Chess and stuff like that. They've never been talked about going rogue.
• 26:45
  Why? Because AI doesn't go rogue. It does what it's told to do. The stories of AI going rogue are really stories of irresponsible programmers.
• 26:57
  What's crazy about this whole thing though is the fact that, you know who's pushing most of all the idea of AI going rogue are these LLM companies. Chat, GPT, I mean, OpenAI, I should say Anthropic. They're the ones pushing it.
• 27:12
  And we have to ask ourselves, why are they doing that?
• 27:15
  Why are they saying their own products are super dangerous and are breaking free and things like that?
• 27:23
  They're doing it because that's how they raise money. You have to remember these, the business model for these companies, they're not making money off of you right now. I mean, you can pay your $200 or $20 a month, whatever that is not breaking even for them. They need investment money. They need billions of dollars investment money to keep going.
• 27:41
  And they need to stand out from the crowd. What they need to do is they need to present a picture of AI is so powerful, it's just unbelievably powerful. And only we can contain and control. Only we heroes at anthropic or at OpenAI or at Google ever. Only we can control this all powerful AI. You need to give us money, rich people, so that we can do that. And so what they do is they release these stories about AI going rogue, their own AI even going rogue. But don't worry, if you give us a lot of money, we'll keep it contained. Even though it's, it's, it's this monster, it's getting ready to go rogue. It's simply not going rogue. First of all, like I said, all these other types of AI you hear about, none of them are there even stories about them going rogue. Why? Because they're not irresponsibly programmed. However, what's happening is here. I'm not going to get super technical on this, but I want to give at least the basics of what's happening here. It's, it's agentic AI is basically what it is, agentic AI, and that is AI agents and what these are. I'm trying to think of the most simple way, but yet correct, accurate way to explain this.
• 28:50
  An agent is basically a program, a harness, a program that calls AIs, like LLMs, ask them to do something, and then it, it takes information and then calls another one or the same one Again to do the next step. So it might be like if you think about, when you're on a chat with OpenAI with ChatGPT or Claude or whoever, you typically ask one question, it answers it and you might ask a follow up wherever the case may be. But it can only do one thing at a time.
• 29:18
  What the agent does is it has a step by step process of trying to get to a goal. It's asking LLM taking for information on that step, how to complete that step, taking that, you know, processing it and then asking for step two, doing the same thing. Repeat, repeat, repeat.
• 29:36
  Well, if you've used chat GPT or Claude or Gemini or whoever, whatever, I should say, not whoever, you know, the answers aren't always correct that it can go, it can get things a little bit wrong, but the agent doesn't know that if there's no human involved, the agent just takes what the LLM said as the truth, as the accurate, and it goes with that. And so it continues to go on this path. And so it can, it can do a lot of things. And so if the LLM says, for example, well, one way that you could solve this problem is if you break into, if you, if you break into this computer, you know, server of this company and that would give you the information you need to do it. There's no human involved here. And so the AI is just like, okay, that's, that's my next step. Then the agent says, okay, that, let's, let's do that step now because what often happens and this happened in the, the biggest story, the OpenAI hugging face incident, where hugging face OpenAI said one of its models went rogue during an internal security test and broke out of its sandbox text test environment and hacked into the AI startup hugging faces infrastructure. However, what people didn't want to report, but it was in the, it was actually in, if you look closely enough, they had turned the safeguards off. Like if you ask your LLM, if you ask your chatbot, hey, show me how to break into Amazon server. It's just not even going to try. It's going to because it has safeguards built in, but they had turned them off.
• 31:09
  So there was nothing, there was no control.
• 31:12
  Nobody was saying no, there was no human involved in this process. So yeah, it figured out, okay, the way to solve the problem I was given is to break into the hugging face servers. And so it didn't go rogue. That's the key point. Now clearly there's a danger here, it's a danger of any powerful Tool.
• 31:33
  It's a danger of any powerful tool that it can be used for good or for evil.
• 31:37
  And so I personally think from what I know about the Open AI hugging face incident, I kind of feel like OpenAI somebody there should be charged with a felony, because it's a felony to break into another server. You can't blame an AI if you literally set up the AI such that you knew it could do that and they did, did. And then it did it. It didn't go rogue. It followed its programming, it followed the directions it was given.
• 32:00
  And so I'm saying this, you might say, what's the difference? The, the reality is it's so broken. I think there is a big difference because I think the real dangers are not the programs going rogue. And all of a sudden and like, they started saying they have, like, sentience, like, they're deciding these things. They're making decisions based upon, upon, like, their own desires. They don't have desires. They just follow what the programming is and they, they stay within the safeguards they're given. If they're given those safeguards, well, they don't have to stay within any safeguards.
• 32:33
  And so I think the real problem here is not AIs going rogue, but AI companies doing things to allow their AI to, to break laws and to do.
• 32:48
  And to hack into other places. And then they act like, oh my gosh, it went rogue in a way. That kind of. It does two things. When they say it goes rogue, it does two things. First, it absolves them of sin, or more accurately, of crime. Like, hey, the AI went rogue. We didn't know it's going to do this.
• 33:07
  We're not guilty.
• 33:09
  And second, it shows how powerful their AI is for investors. So it keeps the police away from them, but it also allows them to get a lot of money. That's what the narrative is. That's what they're saying with this going rogue and things like that.
• 33:21
  Because AI does not have desires, it does not have intentions. It simply follows its programming and it follows the guidelines you give it. So if you give it no guidelines, it's not going to have any to follow.
• 33:33
  Like I said, it can be, it can be programmed poorly, and that's irresponsible. And I do think there should be some, some type of looking into these AI companies in the way they're doing these testings. I mean, like I said, I honestly think there should be an investigation of that OpenAI hugging face incident as a potential criminal action by whoever programmed it. I don't know who it was, but whoever it was for negligence.
• 33:59
  It's like if you're driving down the road and you're, and you just decide you don't care about the, you're not going to follow the, the rules of the road. You're going to go through red lights, you're going to cross the medians and you hit somebody and you kill them, guess what? You're still guilty of a horrific crime. Even if you're like, hey, I wasn't trying to kill anybody. I just decided I'm gonna just ride however I want to. Yeah, you're guilty of a crime. The same thing applies to AI companies.
• 34:26
  And so the, the main reason I'm kind of against this whole going rogue type of thing is because it suggests AI becoming sentient, which is not going to happen. It's simply not going to happen. And this then leads into this idea of the coming singularity where AI is kind of take over the world and, and they, they rule us and things like that. I already did a podcast on this a while back. I can't remember what I called it, but I, I don't want to. So I'm not going to go over all of the, the why that's not a possibility. I did that previously.
• 34:58
  But I also, I think I'll just give one argument why artificial intelligence encompasses a very small sliver of human intelligence. This is, you have to remember this, it's a very small sliver of human intelligence.
• 35:15
  Artificial intelligence is very deep, but it's not very wide.
• 35:19
  When you think about intelligence, true intelligence, like what it takes to live in society like we do, so much of it is unwritten. It drives me crazy when people say that AI today has the sum of all human knowledge within it because they have access to the entire Internet and they process the whole Internet. Do you really think the sum of all human knowledge is in the Internet?
• 35:40
  No, it's not.
• 35:42
  What's on the Internet is the sum of all the written information we have and some of the spoken put online.
• 35:52
  But there's so much more to being intelligent. Just like for example, how you read people's expressions, how you know if they're annoyed or not. Like if you're a wife or your husband is annoyed at you for some reason, they don't say anything. They might not even act like any different. Like a per. A third person who came in might not even realize it. But you know it.
• 36:13
  You know it because you have a certain intelligence when it comes to this.
• 36:17
  And there's a lot of different types of Intelligence. I wrote an article about this too a couple weeks ago for Crisis magazine about like artificial general intelligence is not coming.
• 36:28
  That's this idea. There's going to be. AI is going to have be smarter than everybody at everything. Basically.
• 36:35
  It just is not true because there's types of smarts that are not confined to a computer. I mean, one thing example I like to use is like in basketball, there are certain people, they have an intelligence to see the, the court in a way that most people can't see it. And so they can make a pass that you're like, how did they even know that person was going to be there at that moment?
• 37:00
  Why? Because they have a certain type of intelligence is a spatial physical intelligence that you don't have. You would never be able to see that.
• 37:06
  And a robot may one day be able to do some of that stuff, but the LLM sitting on your laptop can never do that.
• 37:15
  And so there's all these types of intelligence that go beyond. And so there's no sign, there's absolutely no signs whatsoever that AI is becoming sentient or anything like that. It's unfortunate that we're so influenced by Hollywood on this. I mean, a lot of our thinking on AI comes from two. Comes from two reasons, two bad foundations. One is the terrible philosophy everybody has. They don't understand what it means to be human. They don't understand what it means to be intelligent like that. And two is what Hollywood has told us with the Terminator Matrix, things like that.
• 37:48
  Even good AI like Android Data in, in Star Trek, Next Generation, it's this idea that a computer can do more than it really can. They, they always humanize them. Why? Because it's more interesting. It makes a better story. I admit that.
• 38:01
  I mean, I like watching them too, but it's not, it's not actually reality.
• 38:07
  So I just, I'll kind of conclude here again, this was not meant to be a big pro AI podcast. It was meant to be an anti anti AI podcast. I do think a lot of the arguments against AI are not very good. They're emotional, they're ignorant, they're based on ignorance, sometimes they're politically driven.
• 38:30
  I do think there are some good arguments against AI or at least being having concerns with AI. I don't want to act like I don't think that.
• 38:37
  And so the reality is we're living in a time of great change and people do fear change and they should fear change. I mean, it's like it's something new. We don't know what it's going to look like on the other side. It may very well be worse than it is today, but that's not necessarily guaranteed that's going to be worse. And I think acting like it is is just, frankly, that's kind of the Luddite behavior that you assume change is bad. Change might be bad, it might be good, it might be. You know, it'll probably be. It'll probably be like every other change has happened, especially technological change. It brings some benefits and brings some harms. AI is no different than that. And so we have to be, as Catholics, we need to be smart about this, be more in the conversation. And honestly, that's why I wrote my book. Yes, I am going to plug it. If you don't like it, you can stop listening now. But my book coming out in November, the Catholic Guide to Artificial Intelligence, Preserving the Distinction between man and Machine.
• 39:29
  The purpose of it is really to lay out kind of the AI world that we're going to live in and what Catholics should do about it, like how we should live in it. And I definitely, I have a whole, I have a whole chapter that's just dedicated to the dangers of AI, but I have other chapters that I mentioned the dangers throughout. So if you think I'm just some cheerleader for AI, wait till you read my book. You'll see I have a lot of concerns with it. But at the same time, I want to distinguish between the real concerns and the false ones. The real concerns and the more emotional ones, which I don't think are really legitimate. So. So anyway, you can pre order the book at my website. I'll put a link to it. Like I said, it's coming out in November. I'll send you a signed copy once it comes out. But the Catholic Guide to Artificial Intelligence, it's also available at the publisher site, Sophia Institute Press, and also at Amazon, of course. But you can buy it from me.
• 40:15
  Okay, I'm going to wrap it up there. Until next time, everybody. God love you. And remember the PO.