De-Industrial Techno-Optimism

Can an energy scarce future still be sci-fi?

Longtime readers will know that John Michael Greer is one of my favorite writers on esoteric and political topics. And in neither case is this because I agree wholeheartedly with his positions; I, for instance, am more Platonist in disposition than he and his more 20th century strain of occultism. I am also a bit more techno-optimist than he. That said, I enjoy where these differences cause me to reflect, and his most recent blog post, a retrospective and evaluation of the peak oil phenomenon, has caused such reflection.

To give as brief a recounting as I can: The steelman of the peak oil hypothesis is that as the most accessible petroleum sources are tapped out, the cost of extracting less and less accessible or conventional fossil fuels increases until at some point there is no profit margin left; yet the practices, industries, and ways of life set up in the meantime are utterly dependent on energy being under a threshold of affordability and abundance. Fossil fuel stagflation thus leads not just to economic depression within our conventionally acceptable frame of reference, but to the destruction of those complex ways of life.

I find that hard to refute. Any person experiencing this society as an under-30 or even under-40 year old is witnessing our parents’ way of life stagflate rapidly out of our grasp. And most of this generation, of course, is ignorant to the underlying economics and is flocking to play blame games that will not actually make oil more abundant, but just might collapse demand by killing millions of people.

Now, that said, I do think there is hope for a better way of life without having to go full Amish. I will thus give what I hope will be a fairly realistic account of why I am moderately techno-optimistic for the future.

First, let it be noted that I appreciate the reasons JMG presents in his works to believe that the future will look more victorian than sci-fi, and I also think there are reasons to believe such a world would be psychologically and socially healthier than the alternative. A future of unlimited energy leading to runaway Jetsons-style abundance that freezes the postwar social and political order in place forever is probably more likely to lead to gruesome behavioral sink than to human thriving.

But, just as JMG contends that there is a world between apocalypse and sci-fi, I contend that there is still so much low-hanging fruit in the way that we live, organize, and use energy, that a world of declining fossil fuel availability still has potential for astonishing advancements in concrete quality of life for human beings — just maybe not in the form of permanent doordash-powered VR goon pods on the surface of Ganymede. And really, at the stage of material abundance we are already at, what is there left for the 1960s style “gee whiz, I don’t have to wash clothes by hand anymore!” vision of leisure and luxury to offer us? We are already at the point where young people don’t have any idea what is worth doing or struggling for.

JMG’s reasoning for foreseeing a low-tech future hinges on a few additional points of pessimism beyond just the economics of limited fossil fuels. In particular, he is skeptical of the economic viability of nuclear & renewable energy, LLMs, and space travel. In his view, these things cannot exist without subsidy, and increasing scarcity down in the world of raw necessity will eventually hamstring R&D in these sectors.

I am ambivalent on nuclear energy, and I tend to agree that the current implementation of renewables seems in large part to be just a laundering scheme for burning fossil fuels in their manufacture, transport, and maintenance. But I also believe that in this particular instance, a penny saved is a penny earned and we are not yet close to the peak of efficiency and distribution for renewables. Their relative economy will only keep improving the more expensive fossil fuels become, and it will be precisely at the point that profit rather than subsidies are what make them desirable that the greatest efficiency gains will be incentivized; at this stage, subsidies in fact seem to incentivize sloppy and inefficient renewables — like wind, with those ridiculous ecosystem-disrupting turbines. Even so, I will still contend that two gallons of gasoline spent building solar panels, hydroelectric plants, or electric cars are better spent than one gallon burned in a combustion engine; in the former case, some part of that energy is conserved even if the short term opportunity cost is uneconomical.

Now, taking this alone, we are still left with a best case scenario where renewables slowly bring us to the point where we can sustain a static level of energy consumption indefinitely; Optimistically, the equivalent consumption of the early 21st century. Pessimistically, that of the middle 20th century. But of course nothing can truly be taken alone, and I consider these alternative energy sources to be mostly a mitigating rather than absolving factor anyway.

Next, what do we make of LLMs? I do agree that LLMs are unlikely to yield true consciousness or even “superintelligence”, whatever hopefuls mean by that. Even so I still think JMG is mistaken to undersell them. For one thing, frontier labs now seem to be at the point where they are able to use their current models to develop their next models more quickly. Each iteration of the model does indeed make fewer mistakes than the last. Take, for example, how earlier image models failed to produce accurate hands; current models quite consistently depict hands accurately. Many apparent mistakes and much of the “slop” floating around is a result of old people or third worlders using old models for cost reasons or because they don’t know any better.

Now, if scaling is reaching diminishing return territory, and the current tech will not produce intelligences capable of independently inventing new tech or optimizing social coordination or whatever else ASI is hoped to do, where does that leave us?

Same as with renewable energy, there is still juice left to be squeezed from modern models in terms of efficiency and how best to apply the capabilities it does have. This has been neglected because the focus is still on scaling — and it will change as soon as scaling stalls.

The primary reason for optimism about LLMs, in my opinion, is in applications to biotechnology. Biology involves such a staggering volume of high resolution information that LLMs are almost uniquely and providentially suited to helping parse for less minutiae-tolerant human minds. As an example, high resolution genome and biochemical mapping can allow experimentation to happen at an exponential rate via digital simulation, much faster than cell cultivars or animal testing could. Promising applications can be sifted through much faster to find the actual gems. This is a technological capability with a potential quality of life impact extremely disproportionate to the amount of raw thermal or electric power necessary to make use of it.

Now, if you’ll allow me to exercise a bit more imagination here than many of our scientists have – biology is also orders of magnitude more energy efficient than machinery is, and imaginative applications of it can replace many things we now rely on machinery for. Heating, cooling, and even housing, for example, can theoretically be solved with specialized tree cultivars modified for extensive shady canopies, wood quality, or carbon density. Already we have found and started to cultivate microbes that break down plastic; might this same line of inquiry not result in a way to turn existing plastic waste into biofuel? Plants already yield a wide variety of highly specialized and complex chemical byproducts from their simple alchemy of taking in sun, air, and water. Salicylic acid, for example, is a miracle drug in many ways. Penicillin and Vitamin C are both usually made using mold. Theoretically, biomanufacturing is limited only by what raw atoms are available.

Truly, biology is the real “solar power” and in my opinion the key to getting any farther on the “kardashev scale.” It may be unlikely that LLMs will unlock “zero point energy” or clean nuclear or whatever else; but it’s entirely possible that they help unlock several biological and ecological “technologies” that can keep humans thriving much better than our current industrial world does.

Finally, regarding space travel, I also think there is still more squeeze to be had from space industry – note I say “industry” rather than “exploration.” While it is certainly not something that we can rely on a free market producing, I think the subsidies being dumped into it are building up enough of an economy of scale for launches that mining asteroids will become feasible. That will unlock such a glut of metals that reduced input costs will keep certain heavy industries solvent even with increasing energy prices. And one of these industries will be building renewable power sources!

Besides for the technologies that JMG directly comments on, I will add that I think there is also promise in robotics and automation. It seems to me that within the decade we will have a level of automation for various jobs where the lifetime energy input of a machine, between building it, running it, and maintaining it, will be less than the lifetime energy input of a human and especially of a human in modern luxury conditions where the cost of growing and shipping food, fueling a personal car, and moving water, gas, and electric through a house, all must be factored in.

I actually view population decline as highly favorable for human thriving with all of these factors coming together. I think the optimal future for human happiness is quite attainable: A smaller human population that lives longer, makes use of biotech and ecotech, and lives either a rural, robot-assisted agricultural life or a tight and orderly urban life reminiscent of the ancient city-state. If long distance travel is harder and slower we will probably psychologically and socially benefit, and as long as starlink and solar panels are around there will be enough network connectivity that there is a floor to what knowledge can be lost and forgotten. Local LLMs can even allow completely off-grid people to have an extremely valuable repository of knowledge.

Unfortunately, as has almost always historically been the case, our greatest obstacle is not technology but coordination. The longer we use price gouging and mass migration to maintain the postwar status quo, the less thermal and social capital we will have to execute the necessary pivots, and the more likely it becomes that change comes through violence and anarchy. But perhaps that was always the only way for things to happen and the end result will remain along the lines I have described. Who knows!