#61 Japan's New Railgun Propels It Into The Future
Japan Claims Successful Naval Railgun Test, Does the Moon have military value?, The Gradients of Openness in ‘Open’ AI Systems
Today, we will talk about the ramifications of Japan’s recent naval Railgun test-fire.
Also This Week:
Dissecting Lunar Military Significance
Measuring the Openness of AI systems
Matsyanyaaya: Japan Claims Successful Naval Railgun Test
— Josiah W. Neal
Recently, the Acquisition Technology and Logistics Agency (‘ATLA’), part of Japan’s Ministry of Defense, released video footage of a live fire test of a medium-calibre ship-based electromagnetic railgun. It claims this is the first time any country has achieved a goal. The details of the test, such as time and place, were not released. It was a joint effort between the ATLA and the Japanese Maritime Self Defense Force (‘JMSDF’).
What are railguns?
For the course of human history, guns, big and small, have used chemical reactions to propel the projectile forward and towards its target. There are inherent range limits on this technology, however, as the explosion caused cannot be so large or uncontrolled that it destroys the projectile or the gun itself. The theoretical solution, for a long time, has been to switch from using chemical energy to electrical energy to launch projectiles. The technology in practical applications has been nascent, however, and the gap in capability bridged by other technologies such as rocket propulsion of cruise missiles or the use of aerial bombardment.
The science behind the railgun is fairly simple. The gun has two parallel metal rails made of a highly conductive material (e.g., copper) separated by a small gap. Electricity is supplied to one rail, and the current is conducted to the other rail by the projectile in the middle, creating a closed-loop electrical circuit. An electromagnetic phenomenon known as the Lorentz force pushes the projectile in a direction perpendicular to the direction of the current, launching the projectile off the rails. The Lorentz force is powerful and can accelerate projectiles to speeds of several kilometres per second, rivalling the best conventional hypersonic weapons.
Does Japan’s test prove that railgun technology is possible?
As reported by The Warzone, existing specifications for the ATLA’s railgun prototype show that it fires 40mm projectiles at speeds of around Mach 6.5 and has been proposed as having naval and terrestrial applications. One of the main challenges is the supply of energy that railguns need, with this newly tested one supposedly using 5 million joules. However, the ATLA has said it expects it to eventually operate closer to 20 million joules. Japan Marine United has suggested its Atago-class guided-missile destroyers could be equipped with railguns due to their advanced power-generating capabilities.
What is the likelihood of railgun deployment in future?
Provided this test is as it seems, it signifies a step into the high-tech area for Japan’s defence capability and increases the asymmetry of its defence capabilities. Railguns have been popular amongst military futurists for their many purported benefits over conventional artillery. These include more extended range and improved accuracy due to the higher projectile speed and lessened gun recoil. The projectiles are simple conductive metal rods, not explosive warheads, which dramatically reduces ammunition costs and safe storage requirements. There is also no need for explosive propellants or many moving parts in the gun itself. The lethality and versatility of railguns is much higher as well, as the core of the technology is launching a metal rod at a target at hypersonic speeds and letting the kinetic energy of the impact shatter it into pieces.
If this railgun technology proves effective, and challenges such as energy production are adequately addressed, it could have a range of applications, especially for defending against new hypersonic missile threats being developed by North Korea and China. This is a chief concern for Japan in a strategically unpredictable era. Other countries, such as India, could also look to this technology for use against hostile aerial and naval targets. This development bodes well for the Quad, who are increasingly cooperating on defence technology sharing arrangements.
Antariksh Matters: Does the Moon have military value?
— Aditya Ramanathan
Earth’s only natural satellite is estimated to be drifting away from the pull of our planet at the rate of about 3.78 centimetres per year. However, the Moon looms much larger in international politics today than it did even a decade ago. The US space agency NASA alone is expected to spend more than $100 billion on the Artemis lunar exploration programme by 2028. China has its own ambitious International Lunar Research Station (ILRS) plan, which it announced with Russia in 2021.
For all its scientific virtues, lunar exploration is only feasible because states believe it will help them gain prestige, technological dominance, and economic spinoffs. Over longer time horizons, some expect lunar exploration will generate enough economic activity to pay for itself. Yet the Moon is also said to have military value or, at least, be of military interest, though such assertions are short of any specifics. This is partly understandable. The Moon may not have any apparent military significance today, but new political and technological contexts may change all that in the future. If you’re a military planner, it makes sense to hedge your bets. Still, it behoves us to wonder what precisely a future military presence on the Moon might seek to achieve.
The Retro Future of Project Horizon
The idea that the Moon could be the site of a military base predates the Apollo programme. A 1959 US Army study dubbed Project Horizon proposed setting up a facility on the Moon operated by a crew of 12. The aims of the proposal included scientific research and deep space exploration. However, it also viewed the base as an observation post, a place “where future military deterrent forces could be located” and a location for the exploitation of lunar resources to “enhance the potential for strategic space operations” in cislunar space. In effect, Project Horizon could potentially include stationing nuclear weapons on the Moon.
Speculating About the Moon’s Military Significance
While nothing came of Project Horizon, it’s not hard to understand the proposal’s underlying logic: as a remote and desolate geography, the Moon could offer opportunities to conceal military capabilities or station them out of the reach of adversaries. But would it be of any use?
A future military facility on the Moon could achieve any of four objectives. One, it could serve as a base for future military platforms in cislunar or translunar space. These platforms could patrol routes between the Earth and Moon or be stationed at the five Lagrange points from where they could observe civil and military space activity. Such platforms could also deploy counter-space capabilities in Earth-orbit or launch munitions against targets on the Earth, though the distances involved would mean such an attack could take several days.
Two, a lunar base could act as a redundant or last-strike platform that can launch Earth-attack munitions, especially nuclear weapons. Such a base could be used to ensure a state will retain the ability to launch retaliatory strikes even after an adversary has destroyed its nuclear forces on Earth.
Three, given the relative ease of launching from the Moon, such a base could provide responsive launch services to replace adversary-disabled Earth-orbit satellites or to deploy counterspace capabilities that can target adversary satellites.
Four, a military base could protect civilian assets on the Moon from adversaries. This is an application that might come to be seen as necessary if states or private actors manage to establish large and permanent lunar settlements.
A Reality Check
There are good reasons to be sceptical of the ideas I’ve outlined above. For one, any lunar base would violate foundational provisions in the 1967 Outer Space Treaty that prohibit military bases and the permanent stationing of weapons of mass destruction. Changing circumstances could make this legal provision obsolete, but as of now, states have a shared interest in limiting military activity on celestial bodies.
Two, while future cislunar or translunar platforms may provide useful counter-space or Earth-attack capabilities, their actual utility remains highly speculative.
Three, the distance between the Earth and the Moon means any object launched from the lunar surface towards Earth will take about three Earth days to achieve its objective, a geographical reality that limits its potential use cases.
Four, any facility on the Moon will be highly vulnerable to attack, even from simple kinetic munitions. Concealment, hardening, mobility, and deception are all possible countermeasures but will be especially difficult to achieve on a distant celestial body.
In summary, the Moon has dubious strategic value at present. While Earth-orbits have critical military value for a variety of purposes ranging from remote sensing to communications, electronic intelligence, and navigation, the vast expanses beyond it are still mostly about civil pursuits.
Cyberpolitik: The Gradients of Openness in ‘Open’ AI Systems
— Bharath Reddy
**This is an excerpt from an upcoming Takshashila Discussion Document on AI Governance**
The release practices of 'open' AI systems differ significantly from those of open-source software. There are varying degrees of openness in how AI systems are released. A study conducted by Radboud University researchers reveals the large variation in the availability, documentation, and accessibility across different AI models. Additionally, unlike open-source software, significant access barriers in data and computing exist even in the fully open models.
Irene Solaiman's paper introduces a framework for grading the openness of generative AI systems. Generative AI systems are a sub-type of general-purpose AI models that generate content based on user inputs, often across different modalities such as text, images or video. The framework classifies them into six gradients of access: fully closed, gradual or staged access, hosted access, cloud-based or API access, downloadable access, and fully open. This classification helps to understand the extent to which these systems are accessible to users and developers.
Source: The Gradient of Generative AI Release, Source: Solaiman, 2023
With the fully open models, controls against misuse will be harder to enforce. However, they provide the reproducibility and independence from corporate decisions that are necessary for research purposes. One of the most widely used fully open models is BLOOM. It is a multilingual language model built by over 1,000 researchers from 70+ countries to overcome the access barriers that academia, nonprofits or research labs face to create, study, and use LLMs.
However, for most downstream applications, the various levels of convenience, customisation, ownership, and safeguards against misuse offered by hosted access, API access, or downloadable models will prove to be satisfactory. The model type selection will ultimately hinge on the trade-offs among these diverse criteria.
Another important consideration is that big tech companies also have vested interests in 'open' AI development. As a leaked memo by a Google researcher points out, "owning the ecosystem" is extremely valuable. This strategy is similar to what Google has done with Chrome and Android. He states that "by owning the platform where innovation happens, Google cements itself as a thought leader and direction-setter, earning the ability to shape the narrative on ideas that are larger than itself".
This is true of the dominant AI development frameworks PyTorch and TensorFlow, developed by Meta and Google, respectively, both of which are open-source. These companies continue to maintain them, and most AI models are trained on one of these frameworks. Meta's downloadable model, LLaMA, is also an effort at "owning the ecosystem".
What We're Reading (or Listening to)
[Podcast] Chinese Air Force Expansion in Tibet, ft. Shrikrishna Upadhyaya and Dr. Y. Nithiyanandam
[Newsletter] Wafer Wars: Deciphering Latest Restrictions On AI And Semiconductor Manufacturing, by Dylan Patel, Myron Xie, Daniel Nishball, and Wega Chu
[Opinion] Unambiguous on terror, principled on Palestine, by Yusuf T Unjhawala