Ecology, Nukes, & Information
My politics were founded long ago on a concern for apes – orange, black, brown, or silver haired, and the largely hairless and excessively chattering ones. Conserving ecological systems untouched or least touched by man were the most radical politics of the 20th century and remain so in the 21st. The best news is the small efforts taken in the last hundred years revealed by conserving a little, nature’s resilience can revive even greatly devastated ecologies. Key is keeping enough, unfortunately we continue to fail to do that.
Presently, it’s tragically amusing watching the first big backlash to the little traction ecological thinking ever gained. Amusing in the sense, that over my life, most environmental thinking never much penetrated the body politic. Tragic in that such limited movement in the correct direction can be so quickly back-footed. Nonetheless, it was easy to see at some point this backlash would come and environmentalism, such that is, would be easily rolled.
Intricately entwined in all things ecological is energy. Energy is the magic elixir of the universe, nothing moves or grows without it. Nothing. Having energy at our disposal is a wonderful, in ways, magical thing. For the vast majority of the history of homo sapiens, our main energy sources were restricted to the sun, the harnessing of human and animal labor, and the burning of wood. With the dawn of agriculture, ten-thousand years ago, organization of these energy sources began a reshaping of the global landscape, upsetting and replacing very ancient, pre-homo ecological systems. By using these energy sources, what we refer to as civilization, whether the dynasties of Egypt in Africa, the Qin in Asia, Rome in Europe, and the Maya and Incas in the Americas, developed. Two-centuries ago, with the addition of a new source of energy, the burning of fossil fuels, humanity’s reshaping and destruction of existing ecological systems increased exponentially.
Today, it’s difficult to think the mass harnessing of electricity as a technology is barely a century old and what an extraordinary event in the history of humanity it has been. A half century ago, physicist Richard Feynman said, “I think the discovery of electricity and magnetism and the electromagnetic effects, which are finally worked out, the full equations were worked out by Maxwell in 1873, is probably the most fundamental transformation, the most remarkable thing in history.”
That's elec - tric - tric - tric - tric - tric
Today, the relatively recent electrification of American society is not just taken for granted, but literally considered a human right. Completely lost in societal ignorance is what's involved in this mass electrification and how it is all accomplished is a choice. The development of the electricity industry was and is contingent, not determined by some imaginary notion of progress or any law of nature. However, once any given technology is developed and then implemented, its initial structure and processes begin to determine the future. A technology’s initial adoption can lock in an artificially deterministic future, constraining not simply a given technological structure and the society it creates, but thinking itself.
Cheap and plentiful has always been America’s sole electricity concern. Damning all ecological concerns, today, we rush to madly expand American electricity generation. This started a few years ago with the wrong-headed notion of electrifying American car culture. Now can be added the ridiculous notion of a frantic powering of the Tech industry’s newest innovation, deceptively marketed AI, which from an energy perspective is nothing more than a gigantic suck. To meet this demand, AI's advocates call for a nuclear renaissance.
Once again, concern for problems with nuclear energy are swept aside. The inability to dispose of nuclear “waste” and the ability for such waste to provide raw material for nuclear bombs are the two greatest. Also extremely problematic, nuclear electricity generation is controlled by a relatively few people. In the US, all nuclear energy is controlled by the federal government, the Nuclear Regulatory Commission, usurping local and state control, creating an even greater centralization of political power.
Daily an announcement of new nukes being built is made. Each day another headline is offered of Microsoft, Amazon, or Google opening a shuttered nuke or building a new one, including nuclear technologies that don’t yet exist. Like plenty of fantasies peddled from the Tech industry, it’s difficult to tell what’s real from what’s bullshit, half the time its proponents can’t tell either. This has always been an inherent feature of Tech. In so many ways, Tech is a perfect fit for the Age of Trump.
An article from the FT glaringly exemplifies possibility passed off as certainty, "Tennessee Valley Authority agrees to buy nuclear power for Google data centres." The Tennessee Valley Authority (TVA) is a New Deal entity, a “federal corporation.” It’s leadership is appointed by the president and approved by the senate. TVA was created to dam the Tennessee river and bring the then new magic of electricity to the peoples in and around the Tennessee Valley. The FT article states, “Kairos, a small modular reactor company, is building a reactor in Tennessee that will supply 50MW to federal utility TVA, which provides power to Google’s data centres in Alabama and Tennessee.” Kairos has yet to build one nuclear facility. In their marketing material they state, “Our mission is to enable the world’s transition to clean energy, with the ultimate goal of dramatically improving people’s quality of life while protecting the environment.” Phew, that clicks pretty much every feel good 21st century Tech-marketing box doesn’t it. So much so, one could even call it Boomeresque in its naivety.
They will build “a novel advanced reactor technology that aims to be cost competitive with natural gas in the U.S. electricity market and to provide a long-term reduction in cost.” Novel is always good when you’re talking about a technology like nukes, first and foremost it means untested. Even better is “aims,” which isn’t quite “we can do this,” but more, “we hope to be able to do this.” Let’s just say these small nuke reactors have yet to be proved both technically or their cost. We do know nukes have proved far more expensive then their first disciples’ marketing slogan of “too cheap to meter.” This is why its good, with TVA, to have the infinite DC pocket book in the middle. One is reminded a few years ago when the old oil majors started buying up the disputably profitable oil shale plays. An industry wag commented, “It’s good they’ve got a lot of cash flow.”
Even better, in the same article, data center behemoth Equinix (“270 scalable, AI-ready data centers strategically located in 75 major metros around the world”) announced they “signed a 500MW procurement deal with Oklo and pre-ordered 20 reactors from Radiant Nuclear.” Neither Oklo or Radiant has yet to build a reactor. The US Department of Energy states, “Radiant Industries has completed the front-end engineering and experiment design phase to test a prototype of its Kaleidos microreactor at Idaho National Laboratory." – design phase! The FT does at least remind their readers, “The US has brought only three reactors online in the past 20 years,” or you can say only four in last 30 years, but who is counting?
This mad rush for electricity, not just nukes, Louisiana just approved three new natural gas generators for a Facebook data center, is not simply nuts, but in old Valley-speak, it’s largely vapor-ware, a promising of things you can’t yet or ever deliver. There’s absolutely no way all this energy is going to come online in the short amount of time our AI oligarchs promise. The ability to build a data center is a lot faster than providing the electricity needed to power it. Make sure that’s in your calculations for whatever AI stock you’re holding.
Relatively new compute technology has developed atop an industrial system only two centuries old. This industrial system has been and still is overwhelmingly fueled by fossil-fuels. The process has fundamentally altered innumerable ecologies across the planet, in cases inflicting serious damage we already know of and others we are yet to fully understand. Yet, we remain not simply reluctant but determined to not deal with any of this. In regards to nukes, it seems a little irresponsible, to say the least, to go full bore creating tons of new waste that will remain toxic for tens of thousands of years in the name of progress conjured up by Tech-fantasists sitting eighteen hours behind their screens. Conveniently, the Valley’s announced a solution as Forbes breathlessly assures, Nuclear Waste To AI Fuel: US Startup Turning Radioactive Trash To Gold – call it 21st century alchemy.
Not only has compute technology adopted the industrial mindset of unlimited wasteful energy use, they headlong and mindshort advocate tyrannical consolidation and centralized information control beyond anything previously possible. AI’s centralized, brute force compute power, requires massive inputs of energy. We are assured this is the exclusive architecture for an information technology future. However, this architecture is only one way to structure information technology. Among many other issues, its energy requirements make clear it isn’t a good one.
Key to doing things differently is understanding the greatest component of information technology is information and as Norbert Wiener, who provided some of the first equations and processes for AI, stated, “Information is neither energy or matter.” However, information requires energy and matter to be utilized. We need better understanding on how to utilize information, most importantly, the design, the architecture of any information system, determines the quantities of energy and matter necessary for the information to be utilized.
Instead of conceiving information as just another industrial product, information can be seen as a lever utilized to enable us to use less energy and less matter. As that old Greek Archimedes said “in the Doric speech of Syracuse: ‘Give me a place to stand and with a lever I will move the whole world.’" However, in order for information to be used as a lever, it needs a functional, robust politics and it is obvious to the even most myopic of us, besides depravity we have no politics today.
Developing a politics of information requires processes and organization of information not just in machines, but across human society. It is quite clear our old industrial and agrarian institutions and many of their values are incapable of providing such organization and processes. Our ancient government structures are completely incapable of dealing in any effective way with the tsunami of information being unleashed, while industrial values of limitless production and hyper-consumption are not the values of levered information. The so-called laws of supply and demand, the identities of producer and consumer, are not the foundation for an information era requiring first and foremost the individual as citizen, as a participating member of innumerable political bodies, distributedly networked facilitating creation, editing, communication, deliberation, and finally collective decision making – this is a social information architecture, a political fulcrum for information as lever.
Nothing written in this short piece should be taken as anti-technology, but a questioning of how best to use any given technology or even if it should be used at all, nuclear weaponry as an example of the latter. Many of the industrial values we hold dear and others going further back will not help us. As someone who well understood the obstacles of reforming entrenched systems, Mikhail Gorbachev, the last General Secretary of the Communist Party of the Soviet Union said of our present collective predicament, we need a "new paradigm of development…to set in motion a value shift in people's minds.”
The final marketing campaign of our wannabe AI lords is we will all have to work less. Great. Far be it for me to in anyway oppose getting rid of the drudgery of endlessly repetitive jobs, be they on a factory floor or in an office. But this will not be done well without us taking a much grander view of life and the world which bestows that life, an appreciation of the innumerable ecologies of life. Doing that requires a little thinking and as the wonderful Hannah Arendt wrote,
“For if no other test but the experience of being active, no other measure but the extent of sheer activity were to be applied to the various activities within the vita activa (active life, a full life might be a better translation), it might well be that thinking as such would surpass them all.”
After all, thinking is the magic all us apes share.