zenpundit.com » Blog Archive » Thoughts on CNAS “Preparing for War in the Robotic Age”

Thoughts on CNAS “Preparing for War in the Robotic Age”

[by Mark Safranski, a.k.a. “zen“]

My reading at CNAS, which had once been frequent, declined with the waning of the Abu Muqawama blog. While formerly I usually scanned through CNAS reports on a regular basis after reading what Exum and his commenters had to say, toward the end I only visited when Adam and Dan had new posts up.

At the gentle nudging of Frank Hoffman, I decided to read the latest CNAS product;  I’m pleased to say with the release of ” 20YY:Preparing for War in the Robotic Age by Robert Work ( CNAS CEO and former Undersecretary of the Navy) and Shawn Brimley (CNAS Executive V.P. and former NSC Strategic Planning Director) CNAS has rolled out an intellectually provocative analysis on an important emerging aspect of modern warfare.

Work and Brimley have done a number of things well and did them concisely (only 36 pages) in “20YY”:

  • A readable summary of the technological evolution of modern warfare in the past half century while distinguishing between military revolutions,  military-technical revolution and the the 80’s-90’s  American “revolution in military affairs“.
  • .
  • A more specific drill-down on the history of guided munitions and their game-changing importance on the relationship between offense and defense that flourished after the Gulf War. 
  • .
  • An argument that the proliferation of technology and information power into the hands unfriendly states and non-state actors is altering the strategic environment for the United States, writing:
  • .
  • “Meanwhile in the 13 years since the last 20XX game, foreign nation-state C41, surveillance and reconaissance systems, and guided munitions-battle network capabilities have become increasingly capable.  Indeed, these systems now form the very robust and advanced “anti-access and area denial”  (A2/AD) capabilities envisioned in the 20XX game series. The effect has been that the dominance enjoyed by the United States in the late 1990’s/2000’s in the area of high end sensors, guided weaponry, space and cyberspace systems and stealth technology has started to erode. Moreover the erosion is now occurring at an accelerated rate.”
  • .
  • Positing the near-future global proliferation of unmanned, autonomous, networked and swarmed robotic systems replacing( and leveraged by diminishing numbers of) expensive manpower and piloted platforms on the battlefield and altering the age-old relationship between a nation’s population base and the traditional calculation of its potential military power.
  • .
  • An argument that “warfare in the robotic age” will mean substantial to fundamental shifts in strategic calculation of deterrence, coercion, the use of force, operational doctrines and the evolution of military technology and that the United States must prepare for this eventuality.

This report is well worth reading.  In my view there are some areas that require further exploration and debate than can be found in “20YY”. For example:

  • While the power of economics as a driver of unmanned, autonomous weapons is present, the implications are vastly understated. Every nation will face strategic investment choices between opting for simple and cheaper robotic platforms in mass and “pricing out” potential rivals by opting for “class” – fewer but more powerful, sophisticated and versatile robotic systems.
  • .
  • The scale of robot swarms are limited primarily by computing power and cost of manufactureand could be composed of robots from the size of a fly to that of a zeppelin. As John Robb has noted, this could mean billions of drones.
  • .
  • The US defense acquisition system and the armed services are ill-suited for fast and inexpensive introduction of robotic warfare technology – particularly if they threaten to displace profitable legacy platforms – as was demonstrated by the CIA rather than the USAF taking the lead on building a drone fleet.  Once foreign states reach parity, they may soon exceed us technologically in this area. A future presidential candidate may someday warn of  a growing ” robot gap” with China.
  • .
  • Reliance on robotic systems as the center of gravity of your military power carries a terrific risk if effective countermeasures suddenly render them useless at the worst possible time (“Our…our drone swarm….they’ve turned around…they are attacking our own troops….Aaaaahhhh!”)
  • .
  • The use of robotic systems to indiscriminately and autonomously kill is virtually inevitable much like terrorism is inevitable. As with WMD, the weaker the enemy, the less moral scruple they are likely to have in employing lethal robotic technology.
  • .
  • For that matter, the use of robotic systems by an authoritarian state against its own citizens to suppress insurgency, peaceful protest or engage in genocide against minority groups is also highly probable. Is there much doubt how the Kim Family regime in north Korea or Assad in Syria would make use of an army of “killer robots” if they feel their hold on power was threatened?
  • .
  • International Law is not currently configured for genuinely autonomous weapons with Ai operating systems. Most of the theorists and certainly the activists on the subject of  “killer robots” are more interested in waging lawfare exclusively against American possession and use of such weapons than in stopping their proliferation to authoritarian regimes or contracting realistic covenants as to their use.

All in all “20YY:Preparing for War in the Robotic Age provides much food for thought.

7 Responses to “Thoughts on CNAS “Preparing for War in the Robotic Age””

  1. deichmans Says:

    A blast from the past:


  2. J.ScottShipman Says:


    “Machines don’t fight wars, people do — and they use their minds.” John Boyd 

  3. zen Says:

    So far. Yes.
    Ai could change the truth of that Boydian maxim, at least tactically ( “Machines don’t start wars, people do…)

  4. J.ScottShipman Says:

    There is an agility and adaptability required in combat that machines, in the near-term can’t (if ever) provide. Further, battle-space awareness depends on bandwidth—a fragile connection to be sure…btw, concur with your modification. 🙂

  5. Lynn C. Rees Says:

    Since the early 1990s, my default imagining of robotic warfare has been “utility fog“, a sort of “modular self-reconfiguring robotic system“. These would be yet another expression of the more general military phenomenon of Materialschlacht, the battle of material. It might look like a swarm of the unlamented Rocky Mountain Locust:

    …with one famed sighting estimated at 198,000 square miles (513,000 km²) in size (greater than the area of California), weighing 27.5 million tons, and consisting of some 12.5 trillion insects – the greatest concentration of animals ever speculatively guessed, according to Guinness World Records.

    It might look like a dust storm. It might look like a slight rustling motion across the ground. It need not consume everything, only selected parts of the target like organics, etc. It would be cumulative and attritional, like the miasma of robotic mines that will blockade Chinese sea commerce in the not too distant future, deployed by rail gun or simply drifting into position along the North Equatorial Pacific, hidden among the debris generated by China’s westernization modernization efforts. Like worms, they will win tactically by slowly chewing infrastructure. The critical strategic issue will be whose replication infrastructure pumps out more foglets of the right macrophage/T-cell flavor than its opposing polities.


    This will make near-term warfare, at least its jeng (often translated as “orthodox”) leg in the taxonomy used by Swun Dz and later Chinese military writers, much closer to Delbrück’s ideal “war of exhaustion” (Ermattungsstrategie). In such a scenario, just-in-time on-demand agility of the sort popular in some parts of the American military establishment, leading with our glass jaw, will be less of a factor. The sorts of platforms we currently deploy, which, if accurately labeled, would be shipped to war zones with the label FRAGILE prominent on their shipping container, would have the life expectancy of china dolls in a tornado. The American strategic tradition that prevailed in the first half of the twentieth century (“more”) would be more ideationally appropriate than the current fashion (“fancy dancing”).


    The implications of killer robots for our fundamental political economy, one based on the vestigial killing potential of massed riflemen and the need for elites to bid for their loyalty, I leave to the imagination of the reader.

  6. Grurray Says:

    Zen,  here’s some more information about your twitter question.
    Russia ran some military exercises last fall where they mobilized tens of thousands of troops preparing for a scenario where insurgents seize power in urban areas
    The conclusion is that Russian military capabilities are improving, they are practicing for events exactly like Ukraine, and they are reviving Soviet mobilization models. Whether or not they can pull it off is the open question, but there’s no one currently that can force them to back down.

  7. T. Greer Says:

    that is pretty freaking scary.

Switch to our mobile site