While your civil servant experience may have hinted at some of these issues, practical experience for those of us in the military suggests otherwise.
How many times has that fully autonomous capability been used operationally? Moreover, the maritime environment is in my experience a little less cluttered and therefore easier to interpret (from a C2 and RoE perspective) than the land environment.
For both sides; as I said, speed is relative.
We've had autonomous EA capabilities since WWII (research PIPERACK). Whether it's 'just a line of code' to add autonomous kinetic effects or not (which I would contest), is an enormous step change and one that can be both operationally constraining and certainly not necessarily advantageous. That's why many nations are adding 'man-in-the-loop' capabilities and in some cases even spending money to
remove existing autonomy (eg legacy Brimstone) from weapons.
Incidentally, please may I ask for the third time
@sunnoficarus/
@SOI/
@meerkatz/
@PhotEx why you keep changing your name on ARRSE?
Not necessarily. We just have to find a counter.
As I've said a great many times, 'Drones' are merely another form of Air Power.
For the foreseeable future, they will remain hugely less capable at multiple levels to manned types and reliably replicating the intuition, sensory capabilities and decision making of a carbon based life form (whether that be in a GCS or a 'conventional' cockpit) would be enormously expensive and technologically prohibitive.
A human seeing a brief puff of exhaust smoke from an aircraft he's maneuvering against will instantly know his opponent has engaged the burners and may be about to take the fight into the vertical. How many lines of software code, how many hours of bench-testing, and how many hours of trials would it take to teach an autonomous drone to draw the same conclusion or to avoid mistaking the puff of smoke for mist, condensation from wings, a missile being launched, chaff and flares, or any number of other possibilities and vice versa?
How many lines of code would it take for a computer to consider that a recent drought may mean that the object being carried by a man on the ground is a length of irrigation pipe he wishes to move to keep the crops he relies on for his livelihood alive rather than an RPG?
How many lines of code would it take for an autonomous drone to recognise a slight discolouration on the surface of some riverine water as oil seeping from a hostile patrol boat hiding in a nearby mangrove swamp as opposed to effluent from a waste pipe, or an algae bloom?
No one with any experience of Air Power is predicting the demise of manned aircraft in the next 50 plus years. An unmanned Tempest will remain highly valuable, particularly as a wingman. But stop worrying about an autonomous dystopia that we can neither afford (operationally and financially) nor possess the technology for.
Regards,
MM