AI and future of war
Anyone have any good insights on how it could change war ?
We already have computers that are impossible to beat in chess so what about super computers that can simulate every possible move in a war and predict best possible outcome .
>Anyone have any good insights on how it could change war ?
yes
You don't want to know what's coming. No one with any notoriety has successfully predicted it publicly—for good reason. It's worse than everything you could possibly imagine. Worse than your entire family being hunted down by drones through the forest or whatever SciFi nonsense you thought up. You don't even understand your own question, so you could never figure out the answer on your own.
>t. knower. No, I won't tell you. My only advice is that you seek God's mercy and forgiveness.
It's AI ideal partners to pacify and placate entire societies at a time, isn't it?
There would be no point in that. They don't want us alive at all.
That doesn't make sense, all they care about is more wealth and capital. If you kill everyone you have no one to make so grunt work
>all they care about is
How do you know what they care about?
Best way to cozy up with elites
>Anyone have any good insights on how it could change war ?
Instead of some dude a robot that looks exactly like Reimu will shoot you in the chest.
For a serious answer, OP, less than you think in most areas but a lot more in others.
Computers can do Chess because there's a finite amount of potentials. You couldn't really do that with calculating human actions in war, however, you can absolutely feed them shitloads of wargame data and decisions.
It's worth noting here that what everyone is calling AI has fuckall to do with actual AI, and everything to do with extensive scripting.
Anyway, you'd most likely see much more use in the prioritization of multiple targets either for defensive systems, or stuff like the LRASM. Logistics will also become immaculately streamlined. You might also be able to achieve some level of object permanence in sensor systems if a target/vehicle/whatever gets behind something your sensors can't see through.
Biometrics would also benefit, but most people don't even know that's a thing in the military, let alone how important it is, possibly allowing extremely positive ID without having to stick someone's entire head in a portable photo booth. If anyone remembers those things, fuck those things.
Weather prediction would also probably benefit from AI, which is yet another item routinely overlooked.
All this, of course, to say nothing of the benefits of an "AI" personal assistant/secretary/orderly.
The fact that chess, a deterministic game, still hasn't been solved and will continue to be unsolved in the foreseeable future is also telling about current AI capabilities.
It's flat out not going to happen until people work out how to program non-binary computing, and I don't just mean fuzzy logic.
Quantum computing?
They're bad at solving chess but very good at a surprisingly narrow selection of algorithms that binary computers have a hard time with.
>then why are quantum computers so hyped up
they are a meme and meant to be used in conjunction with conventional computing. When the time comes they might have a place in chess but even then it will probably be limited and the game will remain unsolved most likely.
t. guy with a dangerous amount of half-knowledge
-automate target detection/collection [ESM, GMTI, camera, thermals, planes, satellites, hidden ground sensors, vehicles, infantry's reports, spies, anything else]
-automate cross-referencing of sensors to find the exact coordinates, weed out decoys and prioritize a list
-automate mission planning to bomb/arty the targets
-fire automatically
-do counter-battery, counter-counter-battery, and AA automatically
-request resupply and plan the routes for it automatically
-all of the above happens at machine speeds
Future war will be a slowly moving web of AI vehicles which periodically bursts into volleys of fire. There will be a few commanders to sanity check the AI, a few mechanics, and some legacy infantry who are retained to mop up afterwards or do the edge cases the AI can't handle.
This, and there is no other answer right now. I will summarize even more - identify targets, determine right fire solution, send fire solution.
The "hot" military development areas right now are: Small satellites / sat sensors, 3d printed munitions and cheap loitering, target identification. The latter is much more difficult than people think. They watch videos about a person being identified in a gif and think it works. In reality, there are a number of constraints - the camera has to be static or far away enough where the background seems static, it can't rain, there can't be fog, cloud cover can't move to create shadows. This is all solved by LIDAR, but it's expensive and emits. The fact that PrepHole doesn't discuss even this surface shit regularly pretty much confirms that we are all 50 year old boomers here with no clue on modern developments.
What do u mean identify ? Like through satellite images or just photos in general
Before: Satellite thinks it sees something. Command calls another command, talks to someone, checks some shit. AF sends drone. Drone operator actually looks at target and sends a crappy low-res picture to analysts. Analysts try to see if it's an S-400, a blow-up doll S-400 or a dragon dildo. They decide it's an S-400. They call a military lawyer to get permission to engage. They get it. They call another command to ask for fire mission. Send coordinates manually or even speak them. Etc.
Future: Sat sees something, identifies probability of real target, auto-sends drone, drone auto-confirms target, data about situation and target relayed to another computer, all compiled, if go, all coordinates get sent to integrated MLRS/artillery system, which automatically picks the right munition and hits shit. Or more likely, there is a 3d-printed loitering munition hanging out.
Hit on target in 2 minutes instead of 2 days.
Very, very bad idea to not have a human in the loop somewhere. Preferably at the very end with a go/no-go button.
You can put a human in there and they will to say they did. In reality, this will still result in a ton of false positives, at least at first, but maybe always.
That's the way automatic systems work already. For example, the laser anti-drone system being developed finds target, locks on, and gives an operator a clear, zoomed-in view of the target, and says "Fire or on fire? Third option - pick the part of the drone you want to smoke". This is all as of 5-10 years ago.
Same with drones. They see movement in window, zoom in on movement, take a still image, track it, send infantryman a go/no-go decision. Again, 5-10 years old.
>We already have computers that are impossible to beat in chess so what about super computers that can simulate every possible move in a war and predict best possible outcome .
EMPs and signal jammers/face scramblers, marvel 2099 (and other cyberpunk lit) explain what you’d need
AI memeticians conduct soft cultural & socioeconomic war on targeted demographics -- foreign & domestic -- in order to promulgate idioms compatible & contributive to a human-assigned (but robot-interpreted) agenda. Most "wars" will be something like every cellphone in Guatemala suddenly suggesting TikTok videos of a carefully sculpted CGI actor engaging the audience and incepting the idea that they should overthrow their corrupt government, meanwhile also performing an end-run on Guatemala's entire stock market with the dual aims of profiteering while crashing their market... you see where this is going.
Man, I gotta write a postcyberpunk novel.