The story of the last several centuries has been one of the benefits of humanity's technological advances moving faster than the negative effects.
However, the positive impacts are bursty. The negative impacts tend to be linear. It feels like we are past due for a big positive burst, however that in part is because we failed to leverage past positives.
Explaining this, energy transitions, say from wood to coal, and coal to petroleum, are bursty in nature. Accumulated pollution is linear.
The single biggest mistake humanity has made in the last 75 years was not to be more aggressive with the implementation of nuclear power generation. Only France got it right. My lifetime of just over a half a century has been marked by a near-constant series of moral panics related to energy.
One of my early memories is the 1973 energy crisis. This was triggered by the Arab Oil Embargo in response to the U.S.'s material support of Israel during the Yom Kippur War (Operation Nickel Grass).
The Arab Oil Embargo only lasted from October 1973 to March 1974. But the era of the never ending "Energy Crisis" had started.
Energy is everything. Einstein stated energy and matter were the same thing, or another way of looking at this is they are interchangeable. We typically think only of converting matter to energy (burning fuel). However, most of what is created uses energy. We smelt metals and create alloys. We saw and press lumber. We create plastics and other synthetic materials. We lithograph semiconductors. We bend, weld, shape, and assemble metal. We write and test code. We do research. We create medicines. All these acts of creation of products consume significant energy. And as technology increases in capability there will be more need for more energy. There is a very real possibility our ability to innovate will be limited by available energy.
I recently read a short book by Michael Denton, “Fire-Maker". It is an Intelligent Design apologetics book, so it will not appeal to all. But it points out something extremely important: When it comes to civilization, energy is everything. Everything.
Pottery, glass, metallurgy. The ability to take one form of matter and transform it. Fire made that possible. Combustion, or more accurately, creating heat, is the most important form of energy we have.
All of the matter we transform, from mashing potatoes to cutting timber to making steel, to making Portland cement to hammering, screwing, and welding requires energy.
There is energy content in everything, and most of that energy was from heat.
The bigger the transformation of the matter, the more energy is required. And small technology like semiconductors require significant energy.
The more matter we transform, the more energy is required. A growing world population requires more energy.
Energy is everything. It is not just the fuel you put in your car, it is in all the steel, aluminum, plastic, and electronics in your car.
Energy is everything.
We have seen increasing energy consumption driven by computation. At the same time, there have been significant improvements in energy efficiency (i.e., LED lighting), and concepts like cloud computing promise more efficiency. While these efficiency improvements free up energy for other uses, we still are often energy constrained at times (summer heat waves requiring rolling brownouts, etc.).
We have done a great job with energy efficiency. But there are diminishing returns with efficiency. We probably have made most of the gains we can. Certainly, today’s household appliances are much more efficient than those from 50 years ago. Heat pumps went from marginal technology only useful in the Sun Belt to very viable across most of the U.S. Air conditioners and refrigerators are much more efficient, despite being forced to use less efficient refrigerants due to the freon ban. But how much more is possible? We progressed from incandescent, to compact fluorescent, to LED lighting. Tyvek wraps and much better insulation mean houses are much more thermally efficient. Smart thermostats also help. But the marginal efficiency gains beyond 2020 are not likely to see the efficiency gains of the last 50 years.
Another factor which appears to have stabilized is the demand for larger and larger homes. We have seen growth per square foot throughout the 20th century, but that seems to have stalled, and may have reversed, in large part due to decreasing family sizes. There has also been a rise in blended families with the Baby Boomer generation, but those blended families live in the home only for a few years, then the children age out. Large Boomer homes intended to provide plenty of space for holiday gatherings with adult children and grandchildren are declining in demand, and it is likely the late Boomers and early GenXers will leverage nearby hotels and Airbnb to provide the temporary living space for holiday gatherings of family members.
At the same time, early and even younger Millennials are starting to move to the suburbs as they marry and have families. Another factor to consider, especially with the COVID-19 pandemic, is the need for one or even two home offices. But with an average of fewer than two children, at most five bedrooms will be required, and more likely four bedrooms will be the norm, but with a basement or some alternative space for a semi-permanent home office.
Outside of the US, the COVID-19 pandemic forced work from home may change the desired home. Europe is known for space-efficient, compact housing. Many European countries have very low birth rates. In the past, a two-bedroom flat might meet the need. But there may be a demand for larger apartments.
This could drive more energy consumption in Europe, up from current levels.
Then you look at the industrial/business demands for energy.
As Moore’s law’s marginal gains decline, the energy efficiency improvements of computing will decline with it. At that point there will be a rise in energy requirements for computing.
The demand for data analytics and AI will drive the demand for more energy.
The demand for more robotics and automation will drive the demand for more energy.
The demand for electric vehicles will drive the demand for different energy (electricity vs. petroleum).
The demand for autonomous vehicles will drive the demand for more energy.
The demand for more granular autonomous vehicle services will drive the demand for more energy.
The demand for more granular autonomous delivery services will drive the demand for more energy.
The 21st Century lifestyle will demand for more energy than the 20th Century lifestyle.
And the developing world transitioning to a 20th Century lifestyle developed world will demand significantly more energy.
This last point is very important.
What happens when everything becomes digital?
I would bet between 2030 and 2050, a country’s economic success and foreign policy influence will be directly proportional to the percentage of their energy derived from nuclear power.
Also, the first nation to achieve nuclear fusion power at scale will likely propel itself into a global economic advantage.
The cost of AI, robotics, automated manufacturing, and autonomous military drones will plummet for a fusion-powered nation state.
There is a common saying by those in the nuclear power industry: "Fusion power is only 20 years away." Over the decades, the addendum: "This time, we really mean it." could be added. However, if cost-effective, utility-scale fusion power really is 20 years away, that is roughly 2040. And given fusion should have much lower accident and security risks, one can assume the regulatory environment will be less, meaning it will be possible to build fusion power plants much more quickly than fission plants. So, if fusion is readily available by 2040, and fusion plants are easy to build, it could be a significant percentage of a nation's power generation by 2050. Fusion would be a state-of-the-art technology in 2050.
The effect of scale fusion power is the promise of order of magnitude deflation of energy costs. As all economics happens on the margins, the marginal price of generating a watt of power collapsing to near zero (after the sunk capital costs of the fusion plant) would cause significant pressure on all other forms of energy generation, which would only accelerate the adoption of fusion.
Then you must ask: "What other technologies could be state of the art by 2050?"
Obviously, artificial intelligence (AI) will be much more mature in 30 years. It is possible artificial general intelligence (AGI) will be available. Additive manufacturing (3d printing and similar technologies) will likely have matured to the point of being standard. Advancement in robotics is governed in large part by advancements in AI. Nanotechnology is unrelated to robotics, but benefits robotics greatly.
Additive manufacturing, AI, robotics, and nanotechnology will lead not only to fully automated manufacturing, but to "programmable manufacturing", "tooling as code", software-defined manufacturing in programmable factories. Factories that can be changed on the fly to manufacture different things. This has the impact of amortizing capital over a much broader output. It means the classic desire to offset the cost of expensive manufacturing equipment by leveraging locations with low land, regulatory, and construction costs will not have the same weight. The full automation of manufacturing also means the most significant variable cost becomes not labor, but energy. The possibility of small nations with limited populations not being able to have a significant manufacturing output would no longer be true, if that nation has access to low-cost power.
But for larger, more capable states, the potential of very low-cost power is more significant. The ability to rapidly manufacture military equipment in large numbers–autonomous drones, cruise missiles, combat robots–would be a national capability like naval shipbuilding.
Cheap energy has always been an enabler of significant national strength. The UK's and US's access to large coal reserves drove both country's industrial revolutions, industrial power, and national power. Coal allowed the emergence of blue-water navies. Coal was also critical in steelmaking, which was a key component of industrial products as well as military weapons. Large US petroleum reserves were important in driving military armor and airpower which were critical domains of WW2.
The lack of large amounts of low-cost, zero-CO2 energy is the limiter of progress in areas like AI and other key technologies. One only needs to look at the energy consumption of cryptocurrencies to get an idea of the energy consumption of AI computation.
Raw materials are still required. But robotics and automation offer the potential for not only lower-cost extraction, but precision extraction, and more difficult, risky, and dangerous extraction. The ability of robotic mining to extract more ore from more difficult places means robotic mining offers the potential of much larger defined elemental reserves, which will drive down the cost of these commodities.
As all capital (property, plant, equipment) is manufactured product, produced from raw materials with labor, then robotics, automation, lower costs of raw materials, and low-cost energy will result in lower cost capital equipment. Automated robot manufacturing plants built by automated construction robots will produce lower-cost automated manufacturing robots. It will be a virtuous circle.
Lower cost raw materials, near zero cost of labor, low cost of energy, and the impact these trends will have on lowering capital costs means manufacturing complex products will drop significantly.
Ultimately, the limiting cost is energy, and that is where fusion power comes in.
Regarding intermittent renewables, such as wind and solar, they will require significant energy storage to substitute for base-load power. While solar and batteries are falling in price, the primary reason for that price decline is low-cost labor. 80% of solar panels are manufactured in China. There is the very real aspect of China's human rights violation which rise to the levels of slave labor and genocide. Modern lithium-ion batteries are dependent on cobalt, most of which is mined in the Democratic Republic of the Congo using child labor, another human rights violation. While efforts are being made to remove the dependence on cobalt in modern batteries, one also must consider the sheer scale required for battery storage of intermittent renewable generated energy.
Certainly pumped-storage hydroelectricity (PSH) is an effective means of energy storage, and has been used for over a half a century, but the best locations of utility-scale solar plants is not conducive to PSH. PSH could make sense for utility-scale wind power in some locations. But the biggest limitation of PSH is the need for specific land resources, and the environmental impact.
Regardless of the method, energy storage is subject to the laws of thermodynamics, which says there will be efficiency loss. More energy is required to store the energy than the energy stored, or the energy produced from the storage.
In the near-term, intermittent renewables provide power generation that can contribute to the grid while demand is managed with dispatchable power sources such as hydroelectric and natural gas. Solar, which peaks during summer afternoons, is ideal to offset the increased demand for air conditioning during that time.
However, there was a recent study which found photo-voltaic cells degrade much faster than originally expected. This means a solar panel assume to maintain 90% of its capacity at 20 years might drop to the 90% threshold before 13 years, and be as low as 83% by 20 years. This would either necessitate more frequent panel replacement, or larger solar power facilities.
Another aspect of the developed world's economy is it is increasingly moving towards 24-hour operations, meaning nighttime energy demands are increasing and will continue to increase. In some locations, winds increase in the evenings and night (such as California's Sacramento-San Joaquin Delta "Delta Breezes"). However, long-term changes in climate may mean reductions in wind. Sacramento-San Joaquin Delta Breezes have declined significantly over the 20 year period from 1995 to 2015.
It looks increasingly likely it will require significant planning and investment will be required to build utility-scale renewable power with storage. Due to higher than anticipated degradation of solar panels and changes to wind patterns, larger and more distributed solar and wind farms will be required. And that assumes concerns about human rights violations in the case of solar panels and batteries and environmental impacts for all forms of renewables and storage can be overcome. Even if it is overcome, significant dispatchable power will be required to account for scenarios where storage is insufficient. The next phase will cost more. The low-hanging fruit of renewables has been picked. Solar in sunny places with cheap land, wind in windy places with cheap land. Cheap dispatchable natural gas plants instead of storage.
And the reality is, the moment fusion becomes viable, utility-scale renewables will be obsoleted in the developed world. And the first one there, wins the race not to "Net Zero CO2" but "Zero Cost Energy", or more accurately, nearly free energy.
There are synergies between current generations (II, III. and III+) of fission nuclear power plants, fourth generation nuclear power plants, and future fusion plants. Some Gen IV fission nuclear reactor designs use earlier generation's nuclear waste as fuel. Replacing one old reactor with a new Gen IV reactor, or adding a reactor to an existing nuclear power plant, allows on-site reprocessing of nuclear waste. The waste from these reactors is much lower in radiation. Also, when fusion becomes viable, it will require hydrogen for fuel, and extracting hydrogen from the super-heated steam of the water used to generate electricity in fission plants takes less energy than from room temperature water. So putting a fusion reactor next to an existing fission reactor makes some sense.
Wind and solar, and especially solar, are inherently decentralized, and hold tremendous promise for the developing world, much of which is located either in equatorial regions or in the sub-tropical regions of the southern hemisphere. The tropical and subtropical areas are the best locations for solar power. For small, decentralized villages in tropical and subtropical regions of the developing world, solar power and battery storage make sense. For isolated locations above 45 degrees latitude, wind power and battery storage make sense. There are some exceptions, desert regions outside of the tropics and subtropics with consistent sunshine, and areas in the subtropics and below 45 degrees with steady winds.
What about high-density population areas in the developing world? Where existing fission plants are not a factor due to security and anti-proliferation concerns, fusion holds the promise of no weaponizable fuel, and no risk of meltdown, along with the associated lower costs of security and containment.
But for developed nations with significant manufacturing and scale agriculture, energy will be everything. And that will require a significant increase in energy generation, which will likely require technologies beyond renewables. Technology marches on. AI, robotics, and nanotechnology march on. Autonomous manufacturing, farming, and distribution are coming. First to fusion matters. First to fusion wins.
