Correct, the video someone elsethread shared had a guy saying 90%ish of the Russian casualties were drones but only a slight majority of Ukrainian casualties were from drones. Still supports my overall point though -- Russia has a very artillery-heavy force, no one can claim that they didn't invest enough in artillery, and yet still they are getting more kills with their hastily-assembled drone force.

I agree drones aren't going to get much faster or quieter. I think their range and EW-resistance will continue to improve (e.g. due to AI) but other than that they'll stay pretty similar to today. Oh, the price might go down too once they are produced at even greater scale.

They won't rely on dodging autoturrets to hit targets; they'll rely on overwhelming. Suppose you have a perfect autoturret that never misses, but N drones are flying at it simultaneously. What is the smallest N such that at least one drone will get through? I'd be interested to see an analysis that takes into account typical drone speeds, turret rotation/targeting times, and typical distance the drones can creep up before being fired upon (depends on terrain).

 

Both problems are severe. Not enough money is being spent on drone procurement, and what money is being spent is being spent inefficiently. I make no claim about which is worse.

Idk, but I have talked to some people in the USAF and the stories I hear are discouraging. Also I hear that the cool US drone startups sent their stuff to Ukraine and were humbled, the FPV kamikaze designs that actually worked best at scale were mostly homebrew by volunteers. Also, in general, I have come to strongly suspect that classified US military R&D programs are wasteful boondoggles just like the nonclassified ones; the tech is probably pretty great in some sense but (a) takes several times longer to develop than SpaceX would take if they ever became a weapons manufacturer and (b) costs orders of magnitude more. Like, Boeing is responsible for some military R&D and also for the atrocious Starliner system. Why should we think they are doing a much better job for the military than they are for NASA?

Yeah I agree that the physics favors the autoturrets over the drones. I don't think there will be silent drones and even if they are, visual identification will probably work well enough anyway. (There will totally be fixed-wing drones that turn off their propellers and glide silently towards the target btw...)

But even if you have a perfect autoturret, it can probably only take out, like, 10 drones before one gets through and kills it. So your autoturrets can't cost more than 10x the cost of a drone... so, like, $10k. Also, even if you have a $5k autoturret that can reliably take out 10 drones before they close the distance, the drones are way more mobile and so can concentrate force, retreat, etc. and thus will have a huge role to play even if they generally stay away from autoturret-defended areas & even if every vehicle has an autoturret on it.

So I think drones are well worth investing in on the current margins, even if we assume that autoturret tech will advance by leaps and bounds and achieve perfection in the next few years. Which is a very generous assumption.

 

Oh, also, minor thing that isn't particularly relevant: The 40km drone I heard about wasn't a quadcopter, it was fixed-wing. So, it could probably "go quiet" during the terminal dive towards the target. However, I buy your overall claim that it should be possible to after a few years of R&D make autoturrets that cost-efficiently defend against ~all small drones. (Physics based argument: You don't have to fly, so you can rest your weight on the ground, so you can have more weight, so you can have armor, bigger guns, etc. for the same amount of money. You lose mobility but for defense that's probably fine.)

Also, I would say drones have already revolutionized military affairs. 70-90% casualties caused by drones? Despite the war being a static war of attrition with large amounts of artillery on both sides, the kind of situation that historically would have led to artillery being the star of the show? (As indeed it was in the first year or two of the war)

Just imagine if you could teleport the UAF of today back into summer '22. They'd absolutely wipe the floor with the Russians. Likewise Russia would have achieved their original maximalist objectives if they began the war with the force they have today.

Thanks. I skimmed the transcript and didn't see any mention of drones, though I was only skimming so perhaps I missed it. I'm lazy right now and not particularly motivated to dig deeper but if anyone has a timestamp I'd be grateful.

As I describe in this old post, I do expect anti-drone-drones to be really important. (Analogy: Flak cannons weren't the only line of defense in ww2, or even the primary one; fighter planes were.)

OK, thanks. I'm halfway through the interview so far and am not convinced. He doesn't really address my cruxes, e.g. he doesn't provide (1) or (2) above as far as I can tell, or anything close. To be clear, I agree that anti-drone defenses should be heavily prioritized and mass-produced, and that after some number of years they will 'catch up' to today's drone technology and be a reasonably effective counter to it. I describe what this world looks like in this post btw. But on the current margin, both small drones and defenses against such seem woefully under-invested-in by Western militaries. Like, several orders of magnitude less spending than would be optimal. It's embarrassing that Ukraine is producing millions of drones a year, and the USA is producing... thousands? Tens of thousands? 

That said, I do also think that the main war the US needs to gear up for is a war over Taiwan, and that war to me seems like one that will mostly take place in the sky and seas surrounding the island, and hence won't see much role for small drones, or ground forces in general.