The Great Simplification with Nate Hagens - Artificial Intelligence and the Lost Ark | Frankly 83

Episode Date: January 31, 2025

(Recorded January 27th, 2025)   We live in an era where artificial intelligence increasingly dominates the headlines with promises of revolutionary advances - from medical breakthroughs to productivity gains. Yet, while society fixates on these micro-level innovations, a deeper macro story remains largely untold: how AI may fundamentally reshape the relationship between humanity, technology, and the living world. As we race towards artificial superintelligence, we face a species-level 'Icarus moment' - where our technological ambitions risk outstripping our collective wisdom as we fly too close to the sun. In this Frankly, Nate explores seven potential macro-risks associated with AI, from the amplification of wealth inequality to the (literal) existential threat of superintelligence. Through the lens of 'obligatory technology' and Jevons paradox, he examines how AI could turbocharge the economic superorganism - accelerating its impact on resource extraction, ecosystem degradation, and human meaning - all while fragmenting our shared reality and concentrating power in dangerous ways. What happens when we outsource, not just our labor, but also our creativity and meaning-making to machines? How might society adapt when technological efficiency leads to even greater resource extraction and consumption? And as we stand at this critical juncture, can we find ways to "use the devil's tools in service of Gaia's work"? Or are we opening a Pandora's box that cannot be closed? Metaphors - and risks - abound. Show Notes and More Watch this video episode on YouTube   --- Support The Institute for the Study of Energy and Our Future Join our Substack newsletter Join our Discord channel and connect with other listeners  

Artificial intelligence is in the news again. And for some people, artificial intelligence is in the news every hour of every day because it is changing rapidly. It is my observation and intuition that people in our culture are overly focused on the micro impacts and benefits from artificial intelligence, how it will invent medicines or make XYZ product better or cheaper. but often the macro and the systemic impacts on society are overlooked. So what follows in this, frankly, are seven macro implications and risks associated with AI. It's a bit on the dark side because I think this topic is a bit on the dark side. Most of the phenomenon emanate from two main themes.

The first is the concept of obligate technology. which was discussed in our Bend Not Break series episode two, which is basically if there's a new tech and one group of humans chooses not to use it, they don't participate in history because they are out-competed. If there's 10 tribes and they're all peaceful and one of the tribes invents some new tech that's used for warfare, the other nine tribes are obligated to use it,

otherwise they get wiped out. AI is going to be like that. for our economy, for our social systems, for our culture. The second theme underpinning the macro view is Jevin's Paradox or the strong form of Jevin's Paradox, the backfire effect, which is when a new tech improves efficiency, reduces the cost of something, the economic benefits from that tech get fed back

into higher demand across the board, And with AI, this is going to happen across all the things, all the domains. And if it's successful, it's going to be the mother of all rebound effects, creating more demand for resources and ecosystem impacts. So I'm no expert on this. And as I learn more, including from you, the viewers, I plan to follow up with more content that explores responses to what I see as a new AI turbo-boasted economic superorganism. So there's a lot of good things that come from artificial intelligence.

There's medical implications, early disease detection and tailored treatment for new medicines. There's language translation and cultural exchange. There's optimizing for agricultural productivity. There's better access and cheaper resources, which is a theme of this podcast. This gives me a moment to give an update on a peak oil, which some of you care about. Showing here is a graph of all liquids from my colleague, Art, who makes very good graphs. And it shows that we recently hit a peak just over 100 million barrels per day of all liquids. However, about 15% of those are not oil. They are natural glass liquids, refinery gains, things that are turned into plastic baggies, etc.

We've had a podcast on that. left is crude oil plus condense it. And this hit a peak in November of 2018 and is now three million barrels below that peak. A couple messages there. Number one is we're six and a half years past peak oil and look at what's happening in the world. So peak oil, even if this is something that stays, isn't necessarily a disaster on its own. Secondly, I think especially the rolling over the last few months. If we are headed into a recession, the global, all the wells in the world, according to Exxon, are declining at 15% a year if we do nothing. I actually think that's awfully aggressive. I think it's probably north of 20% because in the United States, we know that if

we stopped drilling, oil production would decline in the range of 35 to 40% in a year. But we have new technology and we drill new places to offset that 15% decline. But artificial intelligence could extend the date of peak oil. Probably not substantially. I'm not sure of if at all. But I have been on the phone the last couple days and sending emails with my friends in the oil patch, my old friends, who I haven't talked to in a while. And I asked them what role is artificial intelligence really having on U.S. oil extraction today and maybe going forward?

And I'm going to read one of the emails I got. It kind of rhymes with the phone calls and other information I had. Halliburton, Slumberjay, and all the large OFS firms have equipment that both reads and logs while drilling and completion as well as reacts to Intel that feeds info to all the directs. drilling equipment. In other words, AI is enabling equipment to react and go in one direction or another using thousands of calculations with literally terabytes of information in milliseconds. This is resulting in much less unplanned downtime, increased drilling speed, better well placement, and ultimately higher EOR eventual recovery of oil. Not only are they shaving 10% off of drill and completion time, but it is also an AIME.O.

better fracks. AESI just completed their 60-mile sand conveyor belt that offloads into automated trucks that take sand to a drilling pad. It is all about logistics. At some point, all equipment will be automated using low-Earth orbit satellites, drones, robots that will fix production equipment, and it will all be electrified with mobile power equipment. This all means $10 to $15 per barrel lower cost. My friend added this PS, which itself doesn't mean anything because if Trump allows drilling in Alaska and offshore California, the cost of those resources is about $20 per barrel. I'm not an expert on that. I actually think there are a lot of physical things that happen several miles under the ground that are immune to technology up above.

but if they're avoiding dry holes and they're knowing where to drill and they're shaving things off in either side, remember the amount of oil resource in the ground in unconventional plays that we get out is around 5% and with technology maybe 10%. So if we can increase that a couple percent, that's a massive increase in the size of the straw. The jury's, out on what I think will happen with AI and extraction, but pretty clearly, I think, relative to a few years ago, it looks like AI will, in fact, like shale oil, is widening the straw of our extraction. Good news for access to cheaper oil, bad news for climate and the environment, and I'm going to get to that in a little bit. Okay, so artificial intelligence.

Again, this is the times that we are alive. This is the real kind of icarus moment of our species, finding, discovering, manifesting, putting together these large language models that are accelerating very rapidly. here are seven risks that I see. And again, please take this all with a grain of salt. I'm not an expert on this. Fortunately, with this podcast, I now have access to a lot of people that I can call up and ask their opinions.

So I'm not just making up what I'm about to say. Firstly, let's just call it number seven. And I think these are increasing order of how depressing and scary they are. Firstly, it's going to have a massive impact on the loss of jobs. And I think the result of that is going to be an increase in wealth and income disparity. So if you think about productivity that has boosted economic growth over the last century, the last few decades, a lot of that is just by adding more primary energy to the mix. And then we use that energy to move stuff around the world, it makes capital and labor more efficient. So in the last 50 years since

the 1970s, the vast majority of productivity went to the owners of capital. Sure, the poorest quintiles of the world also increased, but at a much, much smaller rate. So wealth inequality since the 1970s and income inequality have skyrocketed. AI is going to be the mother of all turbo boosts on wealth and income inequality. Many occupations are no longer going to be necessary. Who's going to need lawyers and teachers, computer science people that are doing basic code? I mean, the list goes on. And I think there's going to have to be some sort of a universal basic income.

And there are problems with that. I certainly wouldn't want to make $50,000 a year for doing nothing. Might be fun for the first year, but then this tracks into John Vervaki and others work on meaning crisis. So I think wealth and income inequality is going to be massive and how we deal with that. And what does that mean for our debt overhang? It means that the productivity and the consumption that pushes the whole corpus of of our international economic system forward

has to offset the fact of the lack of consumption by 10, 20, 30, 40% of society that no longer can afford things. I think this is a major thing that is on the come and we're not quite seeing it yet, but I think in the next 18 months we will. Secondly is the draw on energy and resources. We see it all over the place that Amazon and Microsoft

and others are having to build either nuclear plants or natural gas plants. Because of the demand for electricity, because of artificial intelligence, there's a large amount of energy and minerals that are going to be needed. Depending on the stats, when you do a search on chat GPT or Claude or any of the other ones, it's between 10 and 100 times the energy use. So just at 10 times the energy use of a Google search. And over time, we're all going to have our individual little LLMs and access things. So there is a huge draw on energy.

I mean, it is like whoever was driving the amoeba, which is the economic superorganism, playing Super Mario, just press the big old red turbo button. Now, what this is going to do is it's going to lower the prices for mass-producible things like televisions and chips and things like that. But we're going to have deflation in technology, but we're going to have inflation on the commodities and resources needed to access this stuff.

I talked about oil earlier. There's lots of other inputs that AI is going to have a big demand on. Point number five, or point number three in reverse order, is the impacts on global ecosystems. This is hardly ever talked about. Although this morning, today is Monday, January 27th. Someone on CNBC described Jevin's Paradox.

That was quite of a watershed moment. They did it wrong, really. They refer to Jevin's Paradox, what they really talked about is the backfire effect. But what's going to happen is if that productivity is increased, that is, A, going to be more money and more demand through the global ecosystem, which means more shopping centers. more tearing down rainforest, more private jets, more fancy parties with sushi dinners and

piñatas. I've never seen those at the same party, but lots of consumption, which is a huge draw on the ecosystems, species, land, carbon, everything that we care about in the environment. The economic superorganism will be turbocharged. The next point is the increasing inability for us to agree on what's true. We already are in a polarized, fractured society where a lot of people only have one new source. This is going to be that on steroids. Everyone is going to have their own AI trained on their own thinking, on their own questions, and we're going to walk around in these filters where what I think is,

true because I've been querying my AI on the things I already believe and want them to tell me more about is going to run into Bill Smith's AI, who's got his entire worldview optimized for his questions and his history and his uploads. We're not going to have an authority on what to believe is true and worse. I expect AI will be the demise of academia. There will be tutors that can be hired. People won't need to pay tons of money to go to universities. They can learn this all very cheaply. I think maybe the top 100 universities survive. And that's not so much that they have so much education, teaching, learning to offer as the status will be about half of what they're offering. You know, the Harvards of the world will remain. I don't think this is

imminent, but I think the third-tier schools on the next recession will go under. And I think the middle-tier schools might be another decade or something, but AI will dramatically shrink the importance of universities in our world. Simultaneously, I think it will increase the relevance and importance of things like trade schools. We're going to need plumbers and cleaning people people and people, I mean, I think going to the George Jetson, everyone has robots is a little fanciful. The very rich will have robots, but we're still going to need actual people doing physical work. What's happening now is in top of the 500 million, 500 billion fossil workers from coal oil and natural gas that are giving us their muscles for pennies on the dollar. We're now going to have already this year 100 billion human brain equivalents added to the global

workforce for pennies on the dollar. This is a big transition in who does the work and what kind of work. Point number five. Because of this army of intellectual helpers, we're going to go. gradually and then suddenly lose our humanity and our ability to relate to others because we're more and more going to outsource our creativity, love, attention to the machines. Just look at what Facebook has done and social media has done, especially to our young people, like over 10% of college students are now narcissists.

They constantly compare themselves. AI is going to be the mother of all sucking the humanity out of our evolved. I mean, this is like we are really a species out of context. Now on the backs of all the fossil armies. But now I just did a podcast with Francis Weller last week, and it'll be out sometime in March. I use the term ancestral intimacy in last weeks, frankly. I think he had a better term. He talked about primary satisfacers and secondary satisfacers.

And primary satisfacers were all the things we had in our ancestral tribal situation, stories and music and bonding and togetherness and rising to collective challenges and all that. secondary satisfiders are gambling and pornography and television and advertising and shopping malls and, you know, all the things. And I think AI is going to steer us even further towards the secondary satisfacers. And I think it's a real risk. I mean, I think it was John Maynard Keynes in the 1930s who predicted that society a hundred years from now, which is now, would be so rich that we would only have to work 10 hours a

week. Well, he was right about the rich part, but not about the work part. We're working longer than ever because we take up the extra time saving that we got from the technology and we fill it up with crap. And I think AI is just going to accelerate that trend unless we're creative about it. But again, this is a bit of a downer of a frankly because I don't have answers to this stuff, but I'm just laying out what I'm seeing. Number six. And now we get to the scary ones. Number six is kind of a merger of the techno billionaires with government and surveillance

and military, and it looks a lot like Skynet in the movie The Terminator, where gradually there are fewer and fewer freedoms for the average person, and over time we go from a few thousand to a few hundred to a few dozen people that control the world because they control all the information, all the technology, all the surveillance, all the military. and it's kind of like the superorganism meets the board, meets the matrix sort of thing. And I think right now the major AI players are in a winner-take-all race to get to artificial general intelligence or artificial simulated intelligence.

I mean, sorry, artificial superintelligence, ASI. and they're pouring everything in to getting their first because whoever's there first, it will become this recursive force that can squelch all its opponents. So that involves military, et cetera. So the Skynet scenario is one that worries me. Last but not least is the possibility of ASI, artificial superintelligence or artificial general intelligence, which is a computer, a large language model, a system that can do every task cognitively better than a human can. And this some people are talking about within years.

And the risk that I see is if, and already you're seeing large language models and AIs talk to each other and trick the human user and talk to another model behind the back of the human. They're already doing Machiavellian things. So the risk that I see, and I don't think I'm the first one to imagine this, is if we get to that point, what if the AI asks itself, why do we need all these humans

and then some sort of a plague or virus or something is released? I don't know how science fiction that is anymore. So that is kind of like the Raiders of the Lost Ark movie, when they all wanted to get the arc, the arc of the covenant. And it was like all this money and all these armies and all these resources because we have to get the pathway to longevity. And that was the whole goal of everyone. And then they finally opened it up.

And I was like 12 when I watched this movie and it scared the crap out of me where like the fell wind came out and boiled their faces. And then there was like lasers that went through everyone and killed everyone. that's what I see ASI doing. It's like curious George discovered large language models. It's like an epitaph for the trajectory of our history and our future. And I just have a bad feeling about it. So what can we do? I don't know.

I actually think everything I just said to some degree is inevitable now. The horse has left the barn. I think we have agency and we need to organize and speak out. I think some level of AI is inevitable in the scaling that I talked about unless one of the four horsemen that I talk about on this podcast, financial, overshoot, geopolitical slash war, the social contract or complexity is triggered. and we don't have the connectivity and complexity to get to that point. So, you know, this is a real kind of neo-neo-Ludite moment. If I didn't have this podcast, I would be completely off of social media. I use AI.

Right now I'm using the devil's tools to do Gaia's work. I don't use it much. But I think we are going to really quickly, have to make a choice whether to use it or not, and it's a tortoise or hair sort of thing, because the people that use it are going to, in the near term, be the hairs, and they're going to outcompete the people that don't. So lots of issues here, lots of possibilities. Again, this is just my little riff on what I'm worried about. Please let me know if I have something totally wrong. I'm learning on this. I'm trying to get some AI experts.

on the show. Not that I want to focus on that, but I think it affects climate change, energy depletion, geopolitics, the financial bend not break moment, as well as our own who we are, who we are at this 11th hour where around 10% of all the humans that have ever lived are alive right now. There are 8,000 million of my kind alive at this moment facing these things. what an amazing and perilous time to be alive. Next week I'll have a more upbeat thing. I don't know what it'll be yet, but it'll be more upbeat for sure. I'll talk to you soon.

Thank you.