Friday, May 19, 2023

You may assume my pronouns

 It seems every organization in the West is obsessed with the idea of self-professed pronouns.

I get it. If someone is gender non-conforming, they should profess their pronouns.

If someone can be considered gender non-conforming, they should certainly be allowed to profess their pronouns.

But if someone is gender conforming, they should be able to say: "You may assume my pronouns."

I remember a gentleman who worked at my mother's employer who had an unusually soft, high, and feminine voice. On the phone, many assumed him to be female. In person, no one made that assumption.

I have seen athletic women with very short hair paired with athletic males. I assumed them to be heterosexual and not lesbians. Again, not unusual.

Certainly there are transgendered individuals who present in the gender they identify with.

My point is, 99% of the people present as the gender they identify with.

The pronouns debate is a fringe debate. Let those who want to profess their pronouns.

Do not encourage or force those who accept others assuming their pronouns to profess them.

Wednesday, March 29, 2023

Brutal Efficiency

In a December 2006 interview with CNET, Sun Microsystems Chief Technology Officer, Greg Papadopoulos repeated the 1943 statement by IBM’s then CEO, Thomas J. Watson, that world only needed five computers. Papadopoulos was referring to the large service providers which were just starting to emerge. 2006 was also the year Amazon Web Services, now synonymous with cloud computing, released its S3 storage service and its EC2 compute service.

Papadopoulos also noted the large service providers, due to their scale and their investment in automation, were capable of driving “brutal efficiencies.” The web-scale services (web search, e-commerce, etc.) drove very high levels of utilization, and Papadopoulos believed the service providers would follow that model. That is exactly what happened with the hyperscale public cloud providers. They drive extreme levels of efficiency through secure virtualization and continuous capacity management. As a result, hyperscale service providers are now the standard for IT efficiency.

In the past, these levels of utilization and efficiency have been difficult to achieve in on-prem organization IT. VMware provided the hypervisor software that drove a wave of consolidation and efficiency improvements, but efficiency gains have stagnated since. The inability to operate on-premises organizational IT in a highly efficient manner is a large driver of moving on-premises software to SaaS providers, and on-premises compute to cloud providers. But in most cases, the costs of the “lift and shift” of heavy, traditional applications to the cloud proves more costly than operating them on-prem.

Another issue is current organization sustainability goals require new considerations about IT efficiency. In fact, in some cases, migrating on-prem software to SaaS, and lifting and shifting custom applications to cloud providers is just being done to outsource the electrical consumption for an organization so they better meet their sustainability goals.

But what happens when the two are in conflict? When the cost of running customer workloads in the cloud is higher than on-prem, but there is a desire to maximize the efficiency of IT to meet sustainability goals? The answer is private clouds and on-prem IT must operate with similar efficiency goals as public clouds. Another consideration is if an organization has real-estate consolidation initiatives that mean owned data centers go against the organization’s real estate strategies. This usually means owned IT resources are hosted in colocation facilities. Also, for those organizations looking to build a true hybrid cloud, there is the desire to move owned IT resources to cloud connected colocation facilities. But unlike an owned data center, where there might be plenty of available space, every square foot of a colo costs money. So, improving efficiency reduces colocation costs.

There is another factor driving the need for improving on-prem IT efficiency. Newer, denser CPUs and memory are consuming more power. Straightforward “one for one” replacement strategies will force either fewer servers per rack, or power and cooling investments in the data center. The cloud providers have no problem configuring servers with hundreds of cores and terabytes of RAM, then loading dozens of virtual machines from many different customers on the same server. But many traditional IT shops fear high consolidation ratios due to the “too many eggs in one basket” philosophy. Of course, the number of eggs that can be tolerated in one basket does grow over time, but not at the rate of Moore’s Law.

IT organizations need to look at VMs and servers the same way storage administrators looked at thin-provisioning on all-flash arrays. When less expensive hard drive systems dominated organizational data storage, it was easy to just thick provision everything. After all, it ensured performance and minimized issues and management efforts. But all-flash was considerably more expensive per TB, so thin-provisioning was necessary. Performance of all-flash was not an issue, so thick provision eager zeroed VMs, done to maximize performance as data in a VMDK grew, was no longer necessary. But it did impact management. Thin-provisioning was scary. What happened if something went wrong? What happened if there was a runaway data writing process? Could it fill the capacity of multiple thin-provisioned volumes and take down multiple apps? But for the last 8 years, all-flash arrays have been used and managed within IT organizations. At its optimum, it means a thin provisioned VM on a VMware datastore, on a thin provisioned LUN on the storage array. So, there is an experience base in “thin everywhere” and “thin on thin” (VMware thin provisioning on storage array thin provisioning) operations.

With each generation of CPUs increasing low-level virtualization features, and increased instruction level parallelism, both at a Moore’s Law rate that exceeds the growth of software ability to consume it, we should be seeing higher vCPU to core ratios. But increasingly, we are lower vCPU to core ratios due to the desire to avoid performance issues. While VMware memory sharing (transparent page sharing) is not used often due to security concerns, VMware memory overcommit features are safe and well understood, but are likely underutilized. While memory sharing is off by default on ESXi, it is on by default in VMware Cloud on AWS, as is ballooning and memory compression. VMware Cloud on AWS seeks to drive very high levels of efficiency. In essence, there are equivalents of “thin provisioning” virtual CPUs on physical CPU cores, thin provisioning virtual RAM on physical RAM, and thin provisioning virtual networks on physical networks. Another term for thin-provisioning in these cases is oversubscription, and we manage oversubscription with tools like QoS. Tools similar to QoS exist in storage (VMware Storage I/O Control, storage array QoS, etc.), CPU, and memory as well (VMware resource allocation shares, reservations, and limits, etc.). But we need deep visibility into storage IOPS, CPU usage, and memory consumption if we want to drive higher levels of oversubscription in these resources. But we must if we want more efficiency.

CPU, hypervisor, and network consolidation and virtualization features have increased dramatically over the last decade, affording business IT customers the opportunity to significantly increase consolidation, including higher vCPU to core ratios.

While lower than 50% CPU utilization at the host level is typical in VMware environments, it is also not unusual to also see VMs over-configured with vRAM. This often is due to ISV recommendations, which are often over-specified to ensure expected performance.

What is needed is visibility into the virtual and physical infrastructure to identify inefficient configurations, adjust them to eliminate inefficiencies and drive higher levels of utilization. A visibility tool must constantly monitor the environment, because after an initial “right-sizing”, reducing allocated resources to only what is needed, resource requirements may change and may grow, requiring later adjustments. The good thing is IT management tools have improved significantly over the last decade to allow efficient “as a service” approaches to business IT.

The incremental improvements in IT efficiency of the past are no longer sufficient. The potential for significant improvements in IT efficiency now exist. When properly implemented, the right tools allows both lower costs and the achievement of sustainability goals.

Tuesday, August 30, 2022

When did pessimism become a virtue?

Pessimism and worry have become virtues. Too often the pessimist is a pessimist because he or she feels being pessimistic represents clear-eyed, honest, realist, and educated conclusions. The pessimist looks at the optimist and assumes the optimist is naïve, deluded, ignorant, or worse they just don't care. The kids have come up with a term for this: The "doomer".

"Doomer" Wojak meme

This is not ideological. It happens on the left and the right. Some on the left are pessimistic about the future due to climate change. Some on the right are pessimistic about the future due to crime, or morality.

Related, somehow we got to a point that equated worry with care. We reached a point where the more concern someone has for a problem, the more worry a problem causes them, it must be because they care more. And those who are not worried, those who are ... carefree ... must be so because they do not care about this important thing that we all should worry about.

He is not a crank, he just cares more than you do.

The problem with this is worry leads to pessimism leads to apocalyptic worry. Environmentalists like Greta Thunberg, Extinction Rebellion, and the Sunrise Movement all represent apocalyptic movements. They literally believe the world is ending. In that sense, they are not very different from the apocalyptic movements of Jim JonesPeoples Temple or the Heaven’s Gate movement of Comet Hale-Bopp fame. And we know how those two ended.

Apocalyptic movements are dangerous. Moral calculus changes when you think your remaining time on Earth is measured in a short period of time. That is why eco-terrorism is a thing. That is why Earth First! literally tried to kill people.

Likewise, on the socially conservative right, we see people who think the decline in morality portends the end of times. Increasing sexual liberation, starting with the sexual revolution, the legalization of abortion, continuing with the gay rights movement, and now the transgender rights movement, triggers more concern. In the past there has been nihilistic right-wing terrorism such as Timothy McVeigh bombing the Murrah Federal Building. We have seen right-wing terrorism in the form of attacks on abortion clinics and assaults and murders of abortion providers. It may actually be a positive that the Supreme Court overturned Roe v. Wade because it may defuse some of the more violent tendencies of the most extreme elements of the socially conservative right. But at the same time, it now incites elements of the political left.

Of course there is worry that is purposeful, just like fear (worry is a form of fear). Worry is why we strap our children into car seats, put safety caps on electrical plugs when we have toddlers, make sure our smoke detectors in our homes work, express concern when our children struggle in school, and polish our resumes when our employer has severe economic struggles. This is not the healthy worry I am speaking to. I am talking about existential worry the future is lost.

Similarly, Abrahamic theology suggests worry is a negative emotion, which while not directly a sin, leads to sin because it reduces faith in God. Even a secular person should realize worry is corrosive.

And don't get me wrong. I probably have more concern about the future than most. I am concerned about a looming energy crisis and a grain shortage that together might cause a depression in Europe and famine in the developing world. But that to me is concern, not worry. I get more worried about missing a flight than $1,000 a month energy bills. At least for now. Come back in January, 2023 and maybe I will be worried.

What led to me writing this? It was the last few episodes of the the All-In Podcast. Co-host and venture capitalist David Friedberg lamented the continuing topics of inflation, gloom, and doom. Friedberg is the eternal optimist on the podcast, and his optimism is both refreshing and motivating.

Optimism is a virtue. Pessimism is not. Nor is worry.

Friedberg is interested in things like nuclear fusion, carbon capture, new food growing technologies, etc. Just today I saw an article on another breakthrough in manufacturing milk from precision fermentation. This is not "nut milk" or other substitutes. This is a molecule for molecule equivalent beverage, where a fermentation process replaces biological processes. These kinds of breakthroughs, along with other bioreactor technology, may be the very thing that solves the agricultural greenhouse gas problem. These are the technologies that interest Friedburg.

Likewise, I read a story where some people working on geothermal energy are looking at using the powerful lasers used in nuclear fusion to improve their drilling capability. The big deal here is this would allow deep drilling everywhere, so you could literally drill under an existing coal fired power plant and turn it into a geothermal plant. Cheap and plentiful geothermal made possible by laser-enhanced drilling. Nuclear fusion research dollars paying off with clean energy from another source.

Then there is carbon capture, or more accurately carbon-dioxide capture. This would be game-changing, because not only would it solve the CO2 problem, it would provide a source of CO2 that could be used to make renewable synthetic fuel.

There are many reasons to be optimistic. There are more reasons to be optimistic than pessimistic. Riva Tez noted "If you convey an optimistic idea about how the world can be, you’re also suggesting that there’s a responsibility for us to be able to get there", and that pessimism "reduce[s] people's agency to solve problems because there's no point."

This is why we should default to optimism. Even if pessimistic, one should "fake it until they make it" with optimism, if only to motivate the best out of others. That is virtue.

UPDATE, March 2023

Two excellent Substacks have been published in the last month which also speak to this. The first is by Noah Smith on February 22nd:

Don't be a doomer

The second is by Sanjana Friedman on March 3rd, posted at Mike Solana's Pirate Wires Substack:

Collapse Support: The Doomsday Prophets of Reddit


Thursday, March 24, 2022

UFOs are a Thing Again

Just over two years ago the US Navy released video taken in 2004 of what would later be called the “Tic Tac” UFO. The UFO intercepted by the US Navy rekindled interest in the idea of extraterrestrial visits. Most people attempt to consider “close encounters” with extraterrestrials in terms of assuming the UFO represents a highly advanced, interstellar capable, extraterrestrial civilization and working backwards. They assume the UFO is extraterrestrial in origin, represents an advanced civilization, and attempt to assign cause and meaning to it.

I suggest a different approach. Instead of an advanced, interstellar spacefaring society, and working backward (“Why would they explore us?”), I suggest we start with our own, non-advanced, not interstellar spacefaring society, and work forward.

First, we have to ask, what would it take to be interstellar spacefaring? A good first step is harnessing enough power for interstellar travel. That means we have to first assume harnessing power is feasible. For this mental exercise, I will always assume such advanced technology is feasible, after all, the presumption the Tic Tac UFOs are extraterrestrial suggests interstellar travel technologies are feasible.

The first technology is harnessing nuclear fusion in a net-positive way. One has to ask, when will nuclear fusion be harnessed by humanity? In 20 years? 50? 80? With each increase in time, the probability increases. While the joke is nuclear fusion is always 20 years away, 20 years from now might be optimistic. We could assign an 80% chance. In 50 years, perhaps the chance is 90%. By the dawn of the 22nd century, it might be reasonable to say the probability is over 99%.

The next technology to consider is faster than light travel (FTL). The current FTL concept that gets the most traction is the Alcubierre Drive, a concept proposed by physicist Miguel Alcubierre. The Alcubierre Drive is a “space warp drive” concept, which creates a sort of of local wormhole, which is being researched by NASA. The problem with the Alcubierre Drive is it requires impossible amounts of energy. The energy requirements have come down as physicists have refined the concept. Regardless, the power required will likely be significant.

That brings up the best known conceptual energy source more powerful than nuclear fusion: Antimatter. It appears Gene Roddenberry was onto something when he proposed the idea of a “space warp drive” powered by antimatter in his 1960s Star Trek television series.

Let’s assume the same timeline for both harnessing antimatter for power and refining an Alcubierre Drive. Do we believe it is possible in 180 years? How about 280? 380? How about 480? What if we assume an 80% chance for 180 years, 90% for 280, 95% for 380, and 99% for 480? By the dawn of the 26th century, we will likely be an interstellar species.

But even if we assume 280 years for perfection of FTL travel, the dawn of the 24th century, we can extrapolate some ideas.

The first is, any FTL spaceship in the 24th century likely would not require a human crew. The concept of “the singularity”, a point where General Artificial Intelligence passes the Turing test, and where humans will be augmented physically by machines and cognitively computers, will likely occur sometime before the end of the 21st century. Computation will be limited by Moore’s law (which will become a hindrance, rather than an enabler in the future), but new forms of computing, such as quantum computing, are likely to emerge. Nanotechnology is another area ripe for advancement over the 21st century. And biotechnology could impact both computing and nanotechnology.

While Gene Roddenberry got warp drive and antimatter right, he clearly go starship crew sizes very wrong. Instead of crews of 400 (the “Original Series”) or 1,000 (Star Trek - The Next Generation), a crew of a few dozen would be more than adequate. The idea of a holographic doctor (Star Trek Voyager) on board a starship is much more realistic in a 24th century timeline. But realistically, small, fully autonomous starships make more sense.

Keeping the mass small would greatly reduce the power required. Building a starship “as small as possible”, but “as large as necessary” would be the most likely path. This means the size would be dictated by the largest component, as nanotechnology would shrink things like control computers, sensors, recording devices, etc. to very small sizes. Most likely, the FTL drive and antimatter containment system would dictate the size of the spacecraft.

While Star Trek solved the long-distance communication problem with an FTL communications mechanism, this is likely the most difficult problem to solve. More likely FTL spacecraft will be like ancient sailors of old, requiring a return to their home port to tell of their tales. This would be another case for very small FTL spacecraft, which would likely be capable of higher FTL speeds due to their low mass.

Another unknown beyond fusion, Alcubierre Drives, and antimatter is the domain of harnessing gravity. Physicists believe gravity, like light, exists as a wave and a particle (graviton), but gravitons are yet to be discovered, and therefore harnessing them is not possible yet. But assuming it is possible to harness gravity, it could dramatically change slower the light travel. It would it be possible to protect the inside of a spacecraft from acceleration forces. It could allow craft to reflect gravity, to travel like an air hockey puck, changing direction at high speed. Not by being hit, but simply by pointing a graviton beam. This would be especially useful not only for traveling within a solar system, but within a planetary gravity well, that is, landing or flying close to a planet. Instead of fusion based thrusters, needing hydrogen fuel proportional to the gravity of a planet, a lander which harnessed gravity might be able to reflect the planet’s gravity in a way that would make it work at any level of gravity.

I think you can see where I am going with this. A very small robotic spacecraft, capable of flying into a planet’s atmosphere, harnessing gravity so its is capable of moving without the need for visible exhaust or aerodynamic surfaces … it sounds like a Tic Tac UFO.

The Tic Tac UFO makes more sense than the ship from Close Encounters. A mothership is possible, but it would likely be small, and its smaller ships would be even smaller. “Greys” or similar alien species would stay home. “Alien abductions” would have to be done by robots, but honestly, if we assume Star Trek tricorders and medical scanners will be real, why would there any need to abduct a human to learn about their biology?

Do I believe the Tic Tac UFO is an extraterrestrial spacecraft? I don’t know. But I do think in 300 to 500 years, if interstellar travel is possible, human beings will produce small, autonomous interstellar spacecraft, not giant starships of today’s science fiction.

And I think this approach of projecting forward gives a better idea of what to look for than attempting to project a cause onto a phenomenon.

Saturday, January 01, 2022

Why Energy is Everything

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.

Tuesday, October 27, 2020

There will not be a "Return to Normalcy" anytime soon

I wrote these thoughts on Facebook in response to a post endorsing Yascha Mounk's Atlantic article advocating voting for Joe Biden as a means of reigning in the "Illiberal Left".

I think the original poster's analysis was correct. People can be defined by, motivated by, and empowered by what or who they hate, and a Trump reelection would further motivate the illiberal left. However, remember the abhorrent 1998 add that said a vote for the GOP meant another church would burn? People like Mounk appear to be saying, in part, a vote for Trump means another federal building will burn. This is like "giving in" to hostage taking and threats of terrorism.

Furthermore, the only way a vote for Biden avoids increasing the power of the illiberal left is if part of the plan is to vote to create a divided government. In other words, if the goal is to avoiding an emboldened Illiberal Left by voting for an arguably centrist Democrat president, the best way to ensure the Democrat president does not stray from the reservation is to preserve Republican control of the Senate, perhaps also the House, and state and municipal offices. Therefore, those who propose Biden as a firewall or bulwark against the illiberal left should not only endorse Biden, but also endorse voting exclusively GOP down-ticket. If those promoting Biden as a bulwark did that, they would add far more credibility to their argument.

With respect to Biden, the idea he is a centrist firewall against illiberalism does not hold up if you read his campaign web page, consider his planned policies (reinstate the Title IX "Dear Colleague" letter, rescind Trump's anti-CRT EO, pass the Equality Act), or consider many of his comments, such as ones around youth transgenderism. 2020 Biden is not the Biden from the 1980s and 1990s. In some ways he was to the left of Obama during that administration.

Many are voting for Biden because they believe Biden will usher in a "Return to Normalcy". By that, most are thinking the Obama-Biden era of 2015 (the year before Trump declared his candidacy). Others think it will be a return to February 2020 (before the pandemic hit the U.S.). This is naive thinking.

Regardless of who wins, the Successor Ideology (Wesley Yang's term) marches on. Either more peacefully with Biden, or more violently with Trump.

Human nature suggests a critical mass of people will need to feel impacted by something before they push back. The Successor Ideology eventually will impact enough people that there will be a backlash. That will be the inflection point—not the 2020 election. Somebody is going to lose a job. Somebody is going to be conflicted enough to quit a job. Somebody is not going to get a job. Some divorced mom is going to see her aspiring teenage athlete daughter lose an athletic scholarship, and see her hopes and dreams for her daughter crushed. Someone is going to be called a racist one too many times. Someone is going to have to declare themselves a racist one too many times. Eventually a pushback begins. Eventually people organize to pushback.

Rod Dreher writes about the possibility of China's "social credit" system coming to the U.S. Some have written that the Internet Archive has been pressured to remove items from its Wayback Machine web site. What happens when not even your social media posts can get you fired or prevent you from being hired, but even long deleted social media posts are available, because of archiving sites? What happens when others (those who belong to the right group) are able to scrub their pasts including any archives? What happens when the general public has access to some people's social media, but others social media is restricted from public view? We have seen prospective and current college students held to task for social media posts made when they were 14 or 15 years old. We has seen other people held to task for who they choose to follow on social media, with motive assigned to the reason for following. It is reasonable to assume social media follows have been used against someone in an employment situation. We are not far from the ability to gather someone's Twitter follows and assign an ideological and personality profile to the person based on it.

I think we are in what we might call a "Long War". One measured in decades. I believe the start of this global populist/anti-globalist left+right movement goes back to the 1999 WTO protests (perhaps further back to Ross Perot's anti-NAFTA independent/Reform Party run in 1992). I think we have another decade to run. Wokeness and a claimed "democratic socialism" is the endgame of this "New Left". The "Great Reset" is the endgame of the new neoliberal centrists. A "New Right" has not yet fully defined itself. It may be socially reactionary against wokeness. It might be a further evolution of the anti-globalist somewhat socially conservative populism of Ross Perot and Donald Trump.

Another way to look at this is through the lens of the Strauss–Howe generational theory, or "Fourth Turning". In this model, the entire cycle is about 80-years, divided into four roughly 20-year "Turnings". The Third Turning is an "Unraveling." The Fourth Turning is a "Crisis." Per Strauss and Howe, the Third Turning started in the early to mid 1980s, which puts is in a Fourth Turning now. The problem is, I am not so sure. In some ways things still feel like an unraveling., and it is hard to believe we are approaching the end of a Crisis turning. However, that depends on when the Crisis started. Did it start in 1999 (WTO protests), 2000 (contested election), 2001 (9/11 attacks), or did it start in 2008 (financial crisis)? 2008 makes more sense, and fits with my theory the current Fourth Turning will not see resolution until the second half of the decade of the 2020s at the soonest, but probably more towards the end of this decade.

The idea we have another decade of angst to live through is too much for many to accept. They want a return to normalcy as soon as possible. A different president. A vaccine for COVID-19.  A reliable, compounding increase in the S&P 500 index funds in their 401ks driven by continued cost optimization from the globalization of supply chains and labor markets. Their comforting property tax deduction returned, even if the tax cuts meant their total taxes went down. Readily available cheap, but questionably documented day laborers to handle repainting and landscaping. The costs of all of this put in a can and kicked down the road for future generations or Modern Monetary Theory to deal with. They want ideological and financial comfort food—tonight. This desire leads to a false hope that the resolution is near. It leads to a form of Magical Thinking: This just can't go on any longer. It is like the person who posted on Facebook in mid-March at the start of the lockdown: "We are only going have to do this for two or three weeks until we get a vaccine." Yeah, tell me how the SARS vaccine is going, now, eighteen years after that pandemic started.

I believe the 2020s are going to be another 1960s. People think the current chaos is wrapping up, or that this can be wrapped up, with a Biden win, as if changing the president changes our reality—a sort of quantum superposition. The truth is, it is just starting, regardless of who wins. There is no way out, there is only a way through, and with apologies to Robert Frost, both roads are rough, and neither will make any real difference.

Wake me up in 2030.

Wednesday, October 21, 2020

The Coming Airliner Consolidation

The COVID-19 pandemic severely impacted air travel. While air travel is starting to recover, it will take years to fully recover. Much of the recovery at this point is in leisure travel, not in business travel. Businesses have moved aggressively to work from home and video conferencing technology even for meetings across town. This technology will result in less demand for travel until travel for in-person meetings is viewed as a competitive advantage.

The pandemic occurred at a moment of a worldwide economic peak, and a major recapitalization effort on the part of the airlines. The pandemic offered the opportunity for airlines to accelerate the retirement of older aircraft ahead of schedule, while deferring the delivery of newer replacement aircraft.

But the broader and longer-term effects are different. In some cases, entire sub-fleets of aircraft are being retired, primarily due to the length of time expected for demand to recover, and because of accounting issues. An aircraft which is owned outright can be retired and sold on the secondary market. An aircraft over 20 years old is fully depreciated, and while this offers no operating cost reduction through expensing depreciation, these are cash-flow cows, operating at very low expense costs at least until their next major maintenance check ("D-check"). Instead, a new approach to accounting is taken. Older aircraft approaching D-checks are retired ahead of schedule to rationalize fleets. Newer, but fully owned aircraft, especially small sub-fleets are sold because they have some value on the secondary market. Leased aircraft are retained, because the lease payments must be made regardless of if they are flying or not.

The upstream impacts to the aircraft manufacturers are the next consideration, and where things will change considerably.

In good economic times, "thin" routes can justify their own dedicated investment in nonstop routes and equipment. In bad economic times, these cities are served via consolidation and connections. As a result, the pandemic sees less value in aircraft at the small end of a category. Aircraft like the 737-700, A319, etc. in the domestic space, and the 767-300ER, A330-200, and 787-8 in the international space. At the same time, lower passenger demand on the middle routes means these aircraft have ideal capacity points for routes formerly flown by larger aircraft. But often, these are smaller subfleets, so there is pressure to serve the thin markets using a different approach.

My expectation is we will see the manufacturers try to wrap up current production of the small capacity variants of their major product lines, and they will not introduce new small models. For example, there are only 14 orders for the Airbus A330-800neo, the upgraded version of the A300-200. The A330-800neo is designed to directly compete against the 787-8, however, the 787-8 has pretty much made its production run, and most new orders are for the 787-9 and 787-10 variants. The 787-9 is the "sweet spot" for the 787 series, with similar range to the 787-8, but greater passenger capacity, lower manufacturing costs, and with the 787-8’s bugs worked out. With the shift of 787 manufacturing exclusively to Charleston SC, there is the possibility of wrapping up production of the 787-8. Boeing can steer demand towards the 787-9 using pricing and financing. It can also fill 787-8 demand with used aircraft traded in, coming off of lease, or early lease return aircraft.

Regarding the Airbus A330-800, the same is true for Airbus. They can try to get open A330-800neo orders converted to A330-900neo orders, and they can leverage trade-ins and lease returns of A330-200s to fill A330-800neo demand. The main value proposition of the A330-800neo is extreme range at relatively low passenger load, and that market is a "good economy" market negatively impacted by the pandemic.

The same logic applies to the planned Boeing 777-8X. The 777-8X is seen as a follow-on to the "C-market" (ultra long-haul) 777-200LR. But the 777-200LR did not sell well. Boeing has already pushed back the 777-8X, and it is possible Boeing could cancel the project.

An obvious question is: "What about the dedicated freight airliner market?" Dedicated freighters operate on different economics than passenger aircraft, and as such do not demand the latest models with the best efficiency. The Boeing 747 has continued production as a freighter, but will wrap up over the next few years. The Boeing 777-200LR, which was not popular as a passenger airliner, but has sold well as a dedicated freighter. Similarly, the Boeing 767-300ER, being retired right now out of passenger service, continues in production as a dedicated freighter. 767 military variants (the KC-46A aerial refueling tanker) remain in production.

Airbus has not had the success with dedicated freighter variants Boeing has had. Airbus offers a dedicated freighter version of the A330-200, but there are only a few outstanding orders.  Airbus also offers an aerial refueling tanker version of the A330-200, with a number of outstanding orders. Airbus is offering passenger to freighter (P2F) conversions of its A320 narrow body and A330-200 and -300 wide body aircraft. Conversions could also impact the market for dedicated military variants.

Boeing has always had a strong passenger to freighter conversion business, and other companies also offer passenger to freighter conversions of airliners. There are currently proposals for 777-300ER passenger to freighter conversions, but with the pandemic accelerating retirement of 777-200ERs (in favor of more efficient 787s and Airbus A350s), there is a possibility of a flood of used passenger aircraft on the market, depressing acquisition prices, making passenger to freighter conversions very economically compelling, and negatively impacting the new build dedicated freighter market, especially the long-term prospects of the 777-200F.

Finally, on the wide-body front, talk of further stretches of the A350-1000 and 777-9 have been talked about, but with the pandemic reducing overall demand, both of these seem less likely.

Therefore, it is a reasonable assumption for the decade of the 2020s, major wide-bodied airliner production will look like this:

Airbus: A330-900neo, A350-900, A350-1000

Boeing: 787-9, 787-10, 777-9X

Dedicated freighters and military variants can keep a line in production even when significant economic downturns cause passenger orders to be deferred or canceled. In the case of the Boeing 767, it will remain in production for another several years. If more orders are made for 767-300F or KC-46A aircraft, production will continue longer.

In the narrow-body airliner space, the smaller end of the series is also likely to be impacted.

Airbus presents a very interesting case, due to its acquisition of Bombardier’s C-Series airliner (now the Airbus A220). Airbus ended production of the A318 in 2013. Airbus has offered an A319neo, but orders have been very limited (~1% of total A320neo family orders), with fully half of the A319neo orders from one airline, Spirit Airlines. Spirit’s order was made in late 2019, with the A319neo portion finalized in January, just prior to the pandemic’s impact. Will this order survive? Would it make sense strategically for Airbus to steer A319neo demand to the A220-300? After a wave of A320neo orders, the current "hot" model of the A320neo family has clearly become the A321neo, which is targeting the Boeing 757-200 replacement market.

The biggest challenge for Airbus trying to steer demand for the A319neo towards the A220-300 is the A319neo shares a common pilot type rating with other aircraft in the A320 family. But, if it can overcome the cost impact of this, it opens up increasing opportunities for the A220 family.

Assuming Boeing recovers from the 737MAX issues, and is able to shift production exclusively to the MAX and wrap up the 737 Next Generation line, it is reasonable to assume the majority of the 737MAX line will be the -8, -9, and -10. Only two airlines have ordered the -7, Southwest (a large 737-700 customer), and WestJet (also a large 737-700 customer). Like the Airbus A319neo, the 737MAX-7 orders represent just over 1% of total 737MAX orders.

With the Boeing 757 aging (the last model was produced in 2004), the 757-200 replacement market is becoming a factor. As a result, like the A321neo, the 737MAX-10 is becoming popular.

Unlike Airbus, Boeing has nothing like the A220 to shift 737-700 demand to. Also, Boeing is likely to get some military and business jet business for the 737MAX-7. Airbus is now promoting an Airbus Corporate Jet (ACJ) version of its A220, which may compete directly with the ACJ version of the A319. Military versions of the A320 family have not been significant, as European militaries have been more likely to use commuter or business jet platforms for smaller use cases, and the A330 for larger purposes.

Regardless, the A319neo (if it survives) and the 737MAX-7 will represent very small portions of the production of these families, more like the A318s and 737-600s of the prior generations. As such, their production runs may be limited, and production wrapped up at some point early in the production runs, allowing manufacturing optimization to focus on the larger variants.

For the A220, the A220-300 clearly dominates the orders, being the "sweet spot" of passenger capacity, range, and operating costs.

The reasonable assumption for narrow-bodied airliners for the decade of the 2020s will look like this:

Airbus: A220-300, A320neo, A321neo

Boeing: 737MAX-8, 737MAX-9, 737MAX-10

The Boeing 767-300ER represented a unique passenger capacity and range capability which is not easy to replace. Aircraft like the Airbus A321LR and XLR will support about 170 international two-class seats, while the 767-300ER supported about 215, and the next step up, the 767-400ER, Airbus A330-200, or Boeing 787-8 support about 240. The gap between 170 passengers and 240 passengers is fairly significant. Next are the range considerations. The Airbus A321XLR has a range of about 4,500 nautical miles, compared to the 767-300ER’s 6,000nm. While 4,500nm may look fairly long-ranged, one has to consider traveling westbound against winter headwinds, which will reduce that range down to about 3,600nm vs. the 767-300ER’s 5,000nm. This makes a route like the 3,980 nm Charlotte, NC to Munich, Germany route (think BMW automobile related business) likely require a refueling stop in the winter. The options are a connection either in Europe or in the US to a larger aircraft.

The prospect of the pandemic forcing the aircraft manufacturers to consolidate production to fewer models, may allow both Boeing and Airbus to consider new options by the middle of the 2020s. Boeing, if it gets the 737MAX issues behind it, consolidates the 787 at Charleston, and gets the 777-X out the proverbial door, could again look at its proposed New Midsize Airplane (NMA). If 767-300F and KC-46A production is continuing, it could again look at a derivative of the 767, perhaps with a new wing and certainly with the latest generation engines, but this seems unlikely. Alternatively, the shuttered 787 line could be restarted with a shortened 787 with a new wing and engines for this market—but the line could also be retooled for an all-new airplane. Or, most likely, a new airplane could be designed (the proposed Boeing "797"). If the 767 line wraps up by mid-decade, and the Everett 787 line is shuttered, it means the 777-9 would be the only aircraft manufactured in Everett, and the prospect of a new design is increased. I think the most likely outcome is an all-new design proposed in the latter part of this decade.

Likewise, Airbus has the option of looking at aircraft between the A321neo and the A330-900neo. The prospect of an "A322" stretch of the A321 to go after the intra-Europe and US domestic market is possible. Also, while Airbus is rarely mentioned in conjunction with a middle of the market (MoM) aircraft like Boeing is with the NMA, it is certainly possible Airbus could consider something to go after the market once dominated by the 767-300ER.

In summary, I expect the following to happen in the airliner industry. I do not expect the 777-8X or the A330-800neo to be produced. If the Spirit Airlines A319neo order is withdrawn, I do not expect Airbus to continue to offer the A319neo. If it does go forward, I expect the A319neo to wrap up production shortly thereafter. I expect the 737MAX-7 to wrap up after the Southwest and WestJet orders are fulfilled, with the possible exception of business jet and military variants. I expect Airbus’ dedicated freighter and military variant product to wrap up in the next few years, and given the glut of low-time wide-body aircraft on the market, I expect Boeing’s dedicated 777 and 767 freighter orders to dry up, resulting in those lines wrapping up by mid-decade. Likewise, if there is not a follow-on order of KC-46A military refueling tankers, it is possible the 767 line wraps up by mid-decade. I expect the 787-8 variant to wrap up in a few years as well, possibly with the transition of the 787 line to Charleston. I also do not see a market for stretches to the A350 and 777 beyond the A350-1000 and 777-9X. Finally, I expect these consolidations to set up the aircraft manufacturers to explore new designs in the latter years of the decade of the 2020s.