Oh you are in for a treat then. This isn't just any site. It is
Light speed lag.
Or it is the Expanse.
Yes and like most good science fiction, they make a handwave and go with how life works for the rest of it. How they use orbits and trajectory is spot on which is why so many in the science community like the books and the show.
With a N body physics simulator along with thermal and kinetic energy modeling. Not a half-assed let pretend a flight simulator emulates space engine.
Drones relying solely on tactical autonomy will lose out compared to Drones with tactical autonomy and humans in the loop to adjust for unforeseen circumstances. Hence the side that able to have human near the site of a battle will have the advantage with all else being equal.
Yes, but all else won’t be equal. The side with humans at the site will have at least one ship constrained by having humans aboard and having to carry a massive habitat, long-endurance life-support systems, supplies, radiation shielding, and enough delta-v to recover the humans in timely fashion after the battle.
Maybe it is and maybe it isn’t, depending on how good the AI is and how many months or years the operation must last for. Suppose that you can make a drone as smart as a Border collie and get it in theatre in six weeks for $27 million, whereas sending a human will take seven months and cost half a billion even without any way of getting the human back alive.
I disagree that tactical efficiency is a limited task.
The original argument that human don't need to be involved. My thesis is that if one side only autonomous drones and the other side sends both drones and humans. The side with human will win due to a having a faster OODA loop. And the investment involving humans doesn't need to be that great compared to the investment of combat firepower represented by the drone fleet.
They can't make a drone as smart as a cockroach now or in the foreseeable future let alone a border collie. But since we are talking hypotheticals then a human would able to handle the OODA loop faster and easier than a border collies smart drone.
thispersondoesnotexist.com
I don't think that's clear. If you have ever seen a border collie (or perhaps an ACD) playing chasing games with human children you will know that cattle dogs are tactical savants. Besides which, remember the psychology experiment with the gorilla suit and the basketball: human perceptions and perceptual integrations are pretty damned clunky; a robot might well react more reliably and faster.
A drone-based task force will most likely need at least one these as well, unless high-grade tactical AIs and all the hardware they need (computers, sensors, etc.) are cheap enough that you can write them off after each mission. If you can, then humans will be a huge liability because they'll be wanting to come home and the AI doesn't need to. However, if drone carriers and their C&C electronics (and their support hardware, which might not be a light as people think are too expensive to just expend, they'll need all that extra delta-vee to come home that a human-crewed command ship will.
You should totally read through Atomic Rockets. It's fabulous, plus you get to see where the authors of The Expanse got all their jargon from.
Actually they can, it was hugely expensive engineering R&D until SpaceX got into the picture. No amount of money will get you an AI yet.
Right now if there is a pattern to be found then the software can be written to handle it. So given time and studies, I have no doubt that many of the thing a border collie can do can broken down into it elements and translated into software. That underlying software (and in some cases hardware) will continue to develop to handle more sophisticated pattern similar to how collies figure out how to herd sheep in different terrain and weather conditions.
Disagree and having seen human and robotic side by side in a manufacturing plant it obvious what the strength and weakness of both are.
Both are activities involving a lot of pattern recognition. And neither set of system are able to handle why you are making a diagnosis or even what to do about a diagnosis. The same with the legal research. Pattern recognition expert system are tool that extend the productivity of a human being in exactly the same way earth moving machine eliminated the need to have gangs of diggers on a construction site.
I am able to do more in 2022 because of my programming environment improved. I can do more with less work because of the techniques and advances in understand how software framework works. In the 90s anytime I had a list of stuff to maintain there was a lot of boilerplate code that had to be typed out and tested. Then wizards came along, then generics, templates, and interfaces and so on. In the last decade, a lot of development has been made on recognizing how entire programs are structured and with pattern recognition/neural networks the creation of large section of software can be created. But they don't know why they are creating what they are creating.
Or my case how well a completely automated shop does versus a shop with heavy automation with human integrated into the work flow. Hint the latter is able to handle a wider variety of jobs at different volumes than the former. In this regard combat is no different. The future will have huge amount of automation but the best system will remain where they function as tools extending what people can do rather than complete replacement.
People are not there to coordination the interactions of a swarm. Using the real world example of drone light shows, people decide what going to be shown. Uploads the flight pattern, and the drone swarm handles the rest.
The punishment doesn't fit the crime.
Well the problem is like using poison gas after World War I. It all well and good until the wind blow in the opposite direction. It is all well and good with an autonomous drone network until the USS Kearsarge has a short and their IFF system blinks out.
History littered with automation ideas that seemed good on paper. But when implemented, actual conditions played merry hell even with the advanced systems we have today. No matter how much you think it covered exception are the norm, hence the need to have humans in the loop.
How do they know which mode to be in? If they're in war mode at the wrong time, they're a war crime in the making. If they're in patrol mode at the wrong time, they just lose to an enemy drone swarm. If they're out past Jupiter and the nearest humans are on Earth, and they require human decision-making on the matter, the encounter will be over by the time a request for a status upgrade is sent and an answer returned.
No, obviously not. I think you have misconstrued or misapplied the logic, probably disingenuously.
Ethical considerations are the only points where hyperbole is pertinent. The point of the extrapolation was that cheaper doesn't mean better when you're talking about making decisions that kill people.
Some of those depend on spaceships being watercraft or aircraft in disguise — which, to be fair, they often are by genre convention.
