We have no such AI capable of that.
I say that is unknowable for us ordinary folk.
We don't know what is currently being worked on in secret, just waiting in the wings.
This is normal in the industry.
So without info to the contrary, we must look at what current AI in the public sector
can do, & envision how it would be applied militarily. A computer has already
thoroughly beaten a top professional go player. This was an AI hurdle jumped
decades before expected. An autonomous UAV would be similar in difficulty, &
relatively cheap compared to a replacement for the F-35 or F-22.
It's a natural direction for development to proceed, therefore it's already in process.
I was talking about lag between operator and UAC.
I addressed that. Telepresence has no lag if the UAC has sufficient autonomy.
No it can't as it requires a sentient mind. AI does not have that.
With no objective (relative to an autonomous weapon) definition of "sentient"
this cannot be argued. But whether it's AI or something more predetermined,
computers could do the job. It's all about how soon the can (or already do) &
how soon we'll accept their deadly autonomy.
It also will not have compassion, empathy, nor pain. The AI will never see a human as a being like another human can. If some child has an AK-47 the AI will kill that child. The human may not.
Weapons don't need compassion, empathy or pain.
They need only do their job.
As for child soldiers, it would be possible to identify them as not adults.
Children are smaller & have different proportions. But this is a choice.
If a child is fighting against us, trying to kill us as a combatant in war,
then I find it acceptable to kill them.
Which is why UAVs are in low intensity CAS and Recon
That they're limited to a particular function, doesn't mean they're inherently so limited.
No it isn't. Dog-fighting is gun combat. Air combat is any form of combat which is what missiles fall under. Guns have a limited range far below missiles.
There's no point arguing about the definition.
I define dogfighting as combat between fighters by whatever means work.
guns, missiles, ECM, chaff...they're all part of the mix.
And....
Air-to-air missile - Wikipedia
Air-to-air missiles are broadly put in two groups. Those designed to engage opposing aircraft at ranges of less than 30 km are known as short-range or "within visual range" missiles (SRAAMs or WVRAAMs) and are sometimes called "
dogfight" missiles because they are designed to optimize their agility rather than range. Most use
infrared guidance and are called heat-seeking missiles.
I was pointing out the idea was already tried and failed.
What failed?
We have no such technology.
While I don't see a UAV needing it, a missile can pull 60 Gs (IRIS-T).
As for the synthetic mind (software), I can't say whether we do or don't
have current technology. But again, it appears achievable.
China and Russia still have yet to catch up to US current UAV use let alone using in a fighter role.
Their being behind could give them an advantage. We're wedded to our current
view of things, & our major weapon inventory is geared towards threats de jour.
To be cost effective, the enemy could focus upon leap frogging us in not entirely
predictable ways. We have the advantage in conventional warfare, eg, naval
power, air force, littoral force. But if attack is on their minds, it would be a different
battlefield. Think in terms of missiles, UAVs, space, electronic & warfare....&
perhaps nuke, bio, or chem warfare.
Do consider deployment targets if it came to a conflict with China or Russia. Those types of wars UAV will save a few lives in a conflict that will claim millions.
I consider that, thus advocating a more peaceful foreign policy.
But not in weapon system design, which is about winning.