The Packers are generally on minimum wage. When something isn’t in stock, the computer automatically offers an alternative. The packer can either follow the suggestion, or use their judgement. If they blindly follow the computer, they can’t be criticized. If they use their judgement, and a customer complains, then they are blamed. They really have no reason to stick their neck out, even if the computer is being idiotic.

The key difference is that your thinking feeds into your word choice. You also know when to mack up and allow your brain to actually process.

LLMs are (very crudely) a lobotomised speech center. They can chatter and use words, but there is no support structure behind them. The only “knowledge” they have access to is embedded into their training data. Once that is done, they have no ability to “think” about it further. It’s a practical example of a “Chinese Room” and many of the same philosophical arguments apply.

I fully agree that this is an important step for a true AI. It’s just a fragment however. Just like 4 wheels, and 2 axles don’t make a car.

I design build and operate broadcast equipment. A good chunk goes onto UAVs. I’ve built small quads, and I’ve played around with equipment fully capable of some of the more complex tasks. E.g. live 3D mapping from an airborne capable computer.

I’m also friends with several people who used to design and build military equipment, including radar systems. Military tech is a weird mix of amazingly high tech, stupidly simple hacks and long lifespan versions of off the shelf technology. I’ve a fairly good feel for how hard or easy a good chunk of the bits are to build. Most of what I suggested I could personally design and build, or easily commission, given some time, a reasonable budget, and access to restricted resources as required.

In its simplest form, chaff is just tuned lengths of mylar foil. As it flutters, it glitters in a radar beam. This creates a large noise floor. While modern military chaff is more advanced, the old stuff will still cause problems for modern systems. It’s not trying to hide a tank, or pull off a missile’s lock. It’s trying to swamp the signal from a tiny, mostly plastic, drone.

I’m also not saying to reinvent the wheel. Chaff is now a fairly niche defence tool. It’s hard to use while advancing, and gives away your position. It also needs to be integrated with other countermeasures to be useful. It is still a fairly solved problem however. It’s cheap to make, quick to deploy, and available in bulk, if required.

Most modern military equipment isn’t expensive due to its inherent nature. It’s expensive because it’s a niche product, and the buyers have deep wallets. The same game plays out in broadcasting. A £100k camera isn’t that much better than a £5k one. It is better however, and buyers are willing to pay for that difference.

The reverse is also true, as Ukraine is proving. 100 $1k drones are more useful than 1 $100k, ultra capable, drone or missile. The point of a swarm is to allow multiple cheap systems to do the job of a far more expensive weapon.

Easy for a remotely advanced military force.

An explosive drone is easy. Just a small amount of high explosives and an electronic detonator.

Strobe lights could just be an overdriven LED. It just needs to dazzle optical sensors for a few seconds.

Chaff is just lightweight foil. It’s effectively an oversized party popper. It’s job is to help overwhelm radar based tracking.

Software is the hardest bit. At the same time, many computer game ‘AIs’ are good enough at this they need to be dumbed down significantly. It would be more specialised, but only needs to be written once, then rolled out to a fleet.

Batteries would be a swarms limiting factor. Single shot lithium would likely be the bulk. 5-20 minutes of flight, then it’s dead. Disposables would likely need to be moved into position by other means, either a dedicated transport drone, ground transport, or air drop. Your transport doesn’t need to stay in the combat zone however, it can bug out and be reused. Larger more specialist systems would land and loiter to save batteries, and/or be fuel cell powered.

Reliability is handled by numbers, losing 10% is fine, when you have 20% extra.

Computing requires would be met by something like Nvidia’s Jetson range. They are designed for low power, low weight AI processing. Putting a tflop of computing power in the close Comms loop would be simple. The controller would be the most expensive part of the swarm. Not only would it need enough power, both computing and electrical, but also significant Comms capabilities. Radio links, with optical backup would be the workhorse. With a mesh setup, including dummies to help hide it’s location. This is similar to how the display drones work. An expensive hub, serving a cheap swarm.

While none of this is “easy” for a random guy in a shed, or a terrorist in a cave, it’s child’s play compared to a lot of the tech the US can deploy.

For when you REALLY need every advantage you can get with the upcoming meeting.

Potentially to skirt driving time limits?

Many lorry drivers are paid by the trip. If they get stuck in traffic, they are losing money. They are also required to take regular breaks, to avoid fatigue. If they jammed the GPS, then the company can’t prove they didn’t take their break, and worked through, to make up time.

It also allows for disallowed detours. “Sorry boss, I was stuck in traffic for over an hour”. In fact they went for a pub lunch, on the clock.

Nvidia GPUs aren’t the only way to run a machine learning type system. They are just the easiest to use, currently. China has also been developing their own AI optimised chips. Though I don’t know much/anything about them.

This is a good example of how AI can be used well.

Current AIs are effectively fuzzy pattern detection and matching engines. This one can sift all the data coming in, and spot patterns that previously corresponded to problems. It then flags them for human interpretation.

The AI chunks the vast sea of data. A human is then involved to sanity check what it has found, and react accordingly. E.g. a pattern appears that often precedes a broken rail within a month. A human can check the subset of the data, and schedule a maintenance team a week later. Conversely, a pattern that leads by hours would require an immediate response.

Please show me a built up area where you have clear line of sight for 3 miles. 3-500m would be optimistic. You would have 10s to 100s of 15cm drones. They would flit around bins, cars, buildings and through windows. A racing drone can pull 4.5G of acceleration. It can spin that in a fraction of a second.

3G is enough to cover 300m in around 5 seconds. That’s also assuming it is going from a dead start. If it can build up speed before entering line of sight, it would be even quicker.

Even worse, they could easily spend 30 seconds to manoeuvre around you. The sensor package drones (cameras, lidar etc) playing peekaboo, to snatch data. By the time they move, they’ve built a complete 3D map. They know every blind spot, every area the gun can’t target. Your gun will go from nothing to shoot, to too many targets in a second or so. Most will just have extra batteries. They exist to draw fire. A few will have payloads designed to target your defences. Others will have payloads aimed at breaking up your situational awareness.

If you engage the micro drones, then your firing arcs will give windows for heavier elements to engage you. If you don’t, then the armed micro drones will damage your defences or block your sensors, to create the same effect.

Gun drones are perfectly viable. They just can’t fire well while flying. (At least not more than 1 shot) The current prototypes have to land and anchor themselves. They are currently machine guns, for area suppression, though anti-material would be viable.

The gun drone is also not in the smoke cloud, it’s behind it. The smoke, chaff, strobes etc are just to break the ability to counter target it. You don’t need to just saturate the cloud, but the whole area behind it.

As for the smoke, it’s not 1 cloud. A drone’s advantage is hyper mobility. A swarm would easily attack from multiple directions. Your gun is now required to saturate multiple clouds at multiple angles. 1 might be hiding something nasty, or 2 or none. Smoke (or chaff etc) drones would be dirt cheap, as would simple distraction drones.

To fight it, you would either need to put up a wall of shrapnel, which would quickly deplete a mobile weapon, or get accurate targeting data. Both could be viable, depending on the situation, but it’s risky.

As for engagement ranges. A drone swarm would be cut down by advancing over a large open area. I fully agree on that. It would also struggle engaging fixed defences. That changes in a city, or forest, or mountainous area. A patrol or convoy could be encircled by a swarm in seconds, engaging from multiple sides simultaneously.

Your gun can fire 10 rounds a second. That’s 50 rounds in 5 seconds. 200 micro drones, hitting from all sides could easily overwhelm it. Most don’t even need a payload, they are $10-20 decoys. 1 clean hit on your gun however, and it is potentially disabled. At that point the more expensive stuff can potentially attack with impunity.

If it fires big, heavy rounds, then they are slow and of limited numbers. You then bait it at range, or swarm it. If it uses lighter round, to get higher speeds, or more shots, then you use a different platform to soak its shots.

You’re also likely vastly overestimating the final engagement ranges.Right now its long flights at relatively high altitudes. A properly designed drone swarm could hug terrain to close, or be deployed early and loiter on the ground in cover.

A good chunk of the swarm would also be small. 10cm would be big enough to carry just enough teeth to not be ignored. They would also be nimble as hell. It would be a numbers game.

As for the use of smoke. You use 3 or 4 types of drone. A smoke bomb lays down cover. Camera drones fly through and around it to triangulate on your gun. Finally a sniper platform drone moves out of cover and shoots blind, using the camera drones feeds. A coordinator might be required to sort the data. Critically, only cheap, disposable drones are exposed to fire.

The key is that you can mix and match drones on the offensive. Your defence needs to be able to react to all of them.

The game will iterate further. A machine gun works against current drones. It can be countered however. E.g. use a ducted drone, with a few layers of Kevlar facing the gun. It doesn’t need to win, or even survive. It just needs to soak up the fire. The other drones rush in, either behind it, or from various angles.

Even things like chaff and smoke can mess up targeting for long enough to rush in.

I suspect it will be more subtle even if it’s only battery life limited. Huge swarms will also struggle against fixed defences. More likely it will be used in ambush. E.g. air deployed near an enemy convoy, or swarming from rooftops and windows onto an infantry unit. Counter deployment will have to be seconds to stop the lead elements. Potentially with heavier reinforcements flying in.

I’ve personally got visions of a Boston dynamics dogbot with a harness full of drones. 1 button press and a few dozen micro drones swarm out, with larger ones launching as needed.

I could also see facial recognition drones being deployed from a predator drone, like cluster bombs. A little akin to the bots used in the film minority report. They swarm a building or block, and try and identify all the faces they can find.

The key thing however will be battery life. Multicopters are power hogs. You need around 40% battery to get maybe 5-20 minutes flight times (depending on how the manoeuvre). Longer times can be achieved , but requires larger systems with higher costs. Is 1 system with a 2 hour flight time worth 20 smaller systems only good for 10 minutes?

I’ve worked with drones of various sizes. Bigger and more expensive ones are more capable, but hard to make bullet proof. If you can remote off their sensors and weapons into cheap, more disposable systems, it makes sense.

A big drone, like a predator, drops a package into an area. Mid sized multicopters provide local computing power and coordination. Small planes provide fast loiter surveillance. Small multicopters with cameras give more accurate coverage. For attack, you have what amounts to a hand grenade with props. Protection takes the form of similar disposables. A flying strobe light to mess up optical tracking. Chaff bombs to mess up radar tracking. Smoke to obscure the high value units.

A lot of these I could throw together myself, given a few weeks, and a few grand. What part wouldn’t be easy, for a large and well funded military r&d team?

It’s worth noting we are at the start of an arms race. It will iterate all over the place.

For example, smoke and chaff deploying drones would make defensive fire harder. Anti air can be either baited (and so depleted) or rushed. Lasers can be shielded against, at least for a time. Jamming can be countered with line of site communications.

In turn, each of these can be countered.

A key thing of note is that your solutions are heavy. Fine for defending a static target, but problematic when dealing with defending a mobile unit etc infantry of transports. In those situations an extremely rapid, focused highly dynamic response would be required. The obvious way to deploy those fast enough is to have them automated and airborne, aka a drone swarm.

I might be completely wrong, current drone warfare is akin to the invention of the smoothbore musket. How it will develop remains to be seen (for better or for worse).

Apparently I’m an idiot. I saw a modified version, mixed in amongst other legit videos.

I’ll try and remember to dig it out later. It was a sales demo from a weapon’s company. I can’t remember exactly which one it was, but the implications scared the shit out of me.

I’ll admit, I’m willing to pay a couple of £ for the convenience of having them available on steam.

The only reliable counter to a drone is likely another drone.

I suspect Peter F Hamilton got it close, in the Confederation series, with WASPs. They are space based weapon platforms. They carry a mix of offensive and defensive subsystems, and operate with swarm logic.

I could easily see a larger drone carrying a swarm of 1 shot micro drones. When close, some would be sacrificed to get better sensor data, others would go on the attack. Conversely, a defensive target would launch their own swarm. It’s goal would be to stop the attackers getting a good shot on a high value target. It might also counterattack, either against the mother ship drone, or backtracking to find the launch site.

Jamming would also be part of this. A jammer could easily cut off the swarm from external data sources. Live satellite or remote surveillance systems would be cut. Point to point lasers are far harder, as are burst transmissions. Local sensor drones could easily punch short range data back, or paint targets, until they are destroyed by defensive systems.

GPS is useful, but not required for operation. Inertial guidance, and ground tracking cameras can easily maintain a good position sense, while completely RF passive. This is also already normal on many toy drones.

You would also want to jam it over a large area. That jamming is akin to a “kick me” sign, in neon lights.