Balloons and Autonomous Weapons
Let's delve into a future where weapons can think for themselves.
The Chinese Balloon incident demonstrated that:
Networked public discussions are great at holding those in power accountable (the administration tried to hide the incursion from the public).
Networked discussions are also very good at digging into issues and advancing political decision-making — networked outrage forced the US administration to take action against it (blow it up).
However, in an environment distorted by swarm-driven Gleichschaltung (a dangerous method of hyping threats to generate social cohesion and justify muting/eliminating any opposition), networked tribalism can quickly turn any incident into an existential threat (in this case, bringing us closer to a new cold war with China).
Thinking about Autonomous Weapons
For me, the balloon incident is more valuable as a jumping-off point for thinking about the future of autonomous weapons* (weapons that think for themselves) —something we all should be thinking about, given recent developments in AI. With this in mind, here’s what made this balloon interesting:
It moved slowly, using wind currents to deploy itself globally. This made it somewhat stealthy (which, with work on reducing the radar signature, could be improved upon). This maneuver allows undetected penetration in uncontested environments and is even relatively effective in well-defended zones like the US.
It had the potential, using modern techniques, to “maneuver” without a propulsion system. New algorithms and AIs (Google developed one) allow high-altitude balloons to utilize the differences in the directionality of winds at different altitudes to steer themselves. Also, advanced versions could use a “wing,” distended below the balloon, to leverage the differences in the direction of these winds to steer themselves aggressively (even moving at right angles).
They are cheap, easy to deploy, and relatively expendable (in comparison to alternatives), making them perfect for uncontested environments and swarming. Once deployed, they can be very effective for surveillance and communications (see the range below).
* The Joint Chiefs of Staff hired me some years ago to think deeply about the future of autonomous weapons.
From the analysis above, it should be clear that autonomous balloons are useful;
for surveillance (better images than satellites),
as a communication platform (with line-of-sight lasers, they can provide bandwidth that is better than an undersea cable),
and as a weapons platform (for releasing autonomous drones or smart munitions).
Now, let’s dig into autonomous weapons and the future of conventional warfare.
Autonomous Weapons
So far, most of the military’s thinking on autonomous weapons has focused on the following:
a human in the loop (developing human/AI/robotic hybrids) and
swarming (using inexpensive and numerous autonomous weapons to overwhelm a target by converging on them from multiple directions).
These concepts leverage the intelligence of autonomous weapons in three ways: speed (it can make decisions faster), maneuver (it can move and navigate without human control), and coordination (it can coordinate its actions, like insects, with other autonomous weapons).
While those discussions are interesting, they aren’t revolutionary concepts that utilize weapon autonomy in ways that revolutionize modern warfare. Without much effort, I think we can do better. Based on the concepts gleaned from the balloon incursions alone, let’s develop a revolutionary idea for autonomous weapons.
