My story is set in the 2100s. Everyone on Earth has a computer implanted in their brain that they use to interface with the technology of the era. A side effect of this is that most violent conflicts begin (and end) with a "psychic" conflict using the implants. This drives a lot of the concepts in my world.

Obviously, the best way for the bad guys to defend their HQ is to fill it with autonomous robots that can't be fought this way.

Is it a given that such technology would exist? Can I do something in my world that would make such technology:

  • Nonexistent?
  • Scarce (so it makes a good boss battle)?
  • Easily defeated with the right tactics ?

To add some clarification about my agents:

  • These are the equivalent of FBI agents. They aren't hackers.
  • Hacking works in this world pretty much like it does now. It can't really be done with no prep in real time.
  • They have both mental and physical combat training
  • 7
    $\begingroup$ I am wearing my tinfoil hat when I visit this world. $\endgroup$
    – Willk
    Feb 15, 2017 at 20:38
  • $\begingroup$ Have you considered that the psychic war you describe essentially comes down to 'my firewall is better than your firewall'. I'm certainly not letting anyone put one of those things in my head until I know it's unhackable. And of course we will all $\endgroup$
    – pHred
    Feb 16, 2017 at 8:57
  • 1
    $\begingroup$ ... be able to trust the HeadWindows OS by your favourite MegaCorp to be completely secure. $\endgroup$
    – pHred
    Feb 16, 2017 at 9:02
  • $\begingroup$ @pHred These devices have been around for almost a century, and they are put in at birth. It's considered a human right and parents can't deny their children an implant. It's how the world works. These things are well vetted. For reference, I'm writing a cyberprep story rather than cyberpunk. Things are generally pretty good. $\endgroup$
    – Mystagogue
    Feb 16, 2017 at 15:58
  • $\begingroup$ So noted. If everyone has one of these in their head I'm not so sure autonomous robots are the best defence. You could build on the EMP concept (cf- Marooned in Realtime. Vernor Vinge). That would deal with defence in depth. What about purely mechanical traps / defences. Unsubvertable and resource intensive so the dude with the most money has the best local defences. Is your world post scarcity? $\endgroup$
    – pHred
    Feb 16, 2017 at 17:59

7 Answers 7


If technology continues to develop as it has over the past ~15 years, robotic soldiers (that likely won't look humanoid) will respond more accurately, precisely, cheaper, and drastically, drastically faster than any human could by the late 20xxs, let alone the 2100s.

We've already seen leaps in autonomous drivers. In 2004 for the DARPA Grand Challenge not a single vehicle completed the course successfully. In 2005 23 of the 24 vehicles surpassed the furthest distance traveled int he 2004 race, with 5 completing the race.

Whenever soldier robots begin real development we'll likely see equally drastic improvement. It'll only be a handful of years--maybe a decade--before we go from prototypes to machines that can respond with gunfire (or whatever weapon) with greater precision and accuracy and do so in a small, small fraction of the time that humans take even to focus on the target. Humans, quite simply, won't be able to compete.

Not only would robots be the go-to against organics if they exist they'll be the go-to in general if actually troops are needed (they might not be).

So how can you stop this?

Stop them from existing, or at least wide-spread.

Make it against some internationally held laws, with fear of robotic soldiers so dearly held that anyone who tries to move forward with the effort sees globally-united efforts at stopping them. Some rogue nations and particularly well-off criminal empires might have them, but they only do so in very limited numbers and under considerable amounts of secrecy.

  • $\begingroup$ This is more or less what I've been thinking, especially about the DARPA challenge. Making their use or manufacture a major crime should make them pretty scarce. $\endgroup$
    – Mystagogue
    Feb 15, 2017 at 19:32

You ask about ways to make a particular technology nonexistent, rare, and common. I think the real key here is not to dig into the autonomous robots at first, but to dig into the implant and why people get the implant in the first place. Surely this is a major world-defining feature of your story, and it's worth fleshing it out.

First off, I'd like to introduce you to my first generation psychic-proof robot:

Now astute readers of this answer will note that the Brick v1.0 is not very mobile. However, everyone will have to admit that it is completely resistant to all psychic attacks via the implant.

Brick v1.0 is, of course, a straw man argument. Nobody is going to defend their HQ with an army of bricks. However, it turns out to be a very useful tool for exploring the issues with these autonomous robots. Our evil overlord would really like their robots to be able to do their bidding. The ability to communicate with the robots is essential, and if you think about it, that is probably the singular reason to have implants in the first place. You're better at conveying your intent to a robot psychically than you are with voice commands or a flash drive.

So I would recommend using this as your primary slider for adjusting how valuable autonomous robots are. If you want autonomous robots to be common place, weaken the power of the implant's ability to convey meaning to other robots, and naturally people will not be as dependent on them. They'll accept autonomous robots more because the autonomy really doesn't hurt them all that much. On the other hand, if you want these autonomous robots to be ultra rare, make the ability to convey meaning psychically into a very powerful ability. If there's a monumental advantage to using psychic communication, you will find people use it.

Once you've dialed in how common you want autonomous robots to be in general, you can fine tune how common they are for evil overlords. To do this, you need to identify a subtle weakness in the psychic system and have your evil overlord exploit this to bring good-dooers down to a level where autonomous robots can actually be a threat. There might be an issue with electromagnetic interference. Or perhaps there are positions where no amount of communication between devices helps, and might makes right (of course, autonomous robots would be designed to be mighty). Or maybe people are just plain lazy now, and literally wont lift a finger if they can help it.

Or, if you want to be even more nefarious, make them psychic, but have them not act on any psychically delivered instructions. Have them lure the good-doers into believing that they have psychic control over the autonomous robot when, in actuality, it's just playing them.

By tuning in the general societal treatment of the implant first, and then making the evil overlord's approach an outlier, you'll make the effects of these implants feel more realistic. Of course, this does mean you have to be creative enough to create a convincing evil overlord outlier, but that's where the fun is!


I think this question and Nex's answer make these drones out to be a bigger problem then they probably would be since electronics are highly sensitive equipment and a tiny static shock in the right place can destroy them.

Basically what I'm saying is that things like EMP, electrical, sticky (splatter goop in a large area to gum up robotics), and a number of other crafty small ordinance would be developed to combat these drones.

Yes, they would likely be hardened to EMPs and electrical attacks, however, they'll still need moving parts and visual capabilities so anything that messes with these would be deadly to them. They'd probably be able to recover eventually, but if you can stop them from killing you for a bit, you can take it out of commission even without completely destroying it. Lastly, humans are notorious for coming up with lots of ways to solve problems whereas an autonomous drone is going to follow a very basic set of commands and can thus be foiled through ingenuity like luring it into traps, etc. and even with lots of different possibilities programmed in, it's just a matter of finding a combination of circumstances it isn't programmed for and eliminating it.

  • 1
    $\begingroup$ Nice answer, aside from the drones will be dumb thing. Network them, add a deep-learning algorithm with the proper reward function(s) (e.g. dodge any incoming projectiles; kill anyone not in the database) and you've got an enemy that learns each time you kill a drone $\endgroup$
    – dot_Sp0T
    Feb 15, 2017 at 23:00
  • $\begingroup$ @dot_Sp0T If they're networked, then they're sensitive to hacking by a large variety of attacks. You could then develop ways of scrambling their communications and then the others wouldn't learn. Basically that addition opens them up to WAY more attacks. $\endgroup$ Feb 16, 2017 at 13:55

Cool scenario. Psychics. Robots.

A true AI robot like R. Daneel Olivaw in I Robot does not need to be told what to do. If something of this ilk wanted to kill you it would be big trouble.

An autonomous robot which is not intelligent needs an interface: to update program, upload commands / routines etc. The robot needs to be told what to do and then it runs through its routines. Possibly it can cope with many variations on routine but not an infinite number. This leads to 2 counter measures.

1: Hack the interface and tell the robot to do something else. Commandeering enemy robots is always welcome. From Art of War Hence a wise general makes a point of foraging on the enemy. One cartload of the enemy's provisions is equivalent to twenty of one's own

2: Give the robot a scenario so far outside its routine that it does not have a response. The strategy exemplified with Monty Pythons Confuse a cat Ltd. Subtitulado! https://www.youtube.com/watch?v=qGRgFpT-zkM


You could always equip yourself with another sort of tropy future bad guy...The Nano Bot! Make it move fast and inject itself into gaps in the killbots armor. From there you have a range of options, such as the Nano bot gums things up, welds critical joints, or even interfaces with the machine AI to give the hacker psychic control. Then that killbot goes and kills a bunch of other killbots.

I like the idea of legal hurdles being in the way of creating killbot armies. If you outlaw kill bots, only criminals will have killbots. That sets you up nicely for bad guy boss battles and such.

You could create scarcity by making the necessary battery material required for an energy intensive killbot ridiculously expensive or difficult to create. If you need 30 kilograms of pure gold to make the necessary contacts, then this thing is going to be priced well beyond most ordinary bad guys.


It seems to me if the technology exists to create autonomous robotics then a true fully sentient AI is not far away, and if it can be done someone will do it. Probably several someones.

A sentient AI or two roaming the internet, would likely be impossible to delete, but governments could likely keep it mostly contained with ever evolving firewall's and "anti-virus" type software. Like the flu, some people get sick every year but it's more nuisance than plague. Between vaccinations and evolving immunity it's not really a big deal.

With that kind of AI roaming around intelligent robotics might be at higher risk for the AI to try and commandeer them. Building intelligent machines would then become a risk. That would keep the relative intelligence of such autonomous sentries to a minimum, out of necessity they would need to be easy to shut down or take out if necessary, but still dangerous enough to be worth building. You would end up with dumb networked machines, or intelligent but deliberately stand-alone machines, either of which a group of humans working together effectively should be able to deal with.


Robotic Scarcity

We've had the technology for some basic battlefield-robotics for a while, and yet things as simple as unarmed camera-drones still aren't commonplace. It's not about them being vulnerable to hacking, or ethical concerns...

It's all about the about money.

Consider, for a moment, that the average Private in the US Army makes less than 20,000 USD per year. Even a Staff Sergeant makes less than 30,000 USD in annual salary. That means that paying the yearly salary of an entire fireteam (Four people, usually several privates with at most one Junior NCO) costs around 90,000 USD - Less than a mid-to-high-end luxury car.

For comparisons sake, a single Javelin missile and command/targeting unit runs nearly 200,000 USD. A single shot for a vaguely high-tech weapon costs as much as 10 men. I don't have any hard figures on infantry-scale combat robotics, but I can only assume they'll easily reach into the millions of dollars.

Unfortunately, life is cheap.

You might call it dystopian, but our world is already in a place where it's cheaper to pay people to die than it is to build robots and save lives. Now, I assume your evil crime-lord/ mad scientist has the funding to afford robot soldiers. Let's also assume that they're tougher, faster, more heavily armed and more psychically-resistant than your average non-hero. They can take a headshot and keep fighting, and they never miss (Unless they're aiming at protagonists.) Perfect fighting force for every situation, right?

But.... For the price of just one of those robots you could probably pay several dozen if not hundred undesirables to do your bidding. The more poor and destitute they are, the more cannon-fodder you can afford to simply throw at the problem. As an added benefit, if they die you no longer have to pay their wages, whereas your robots would require an astronomical repair and maintenance budget.

Anti-robot tactics

As Nex Terren pointed out, these robots may have super-human abilities. However, since we're in a world of cybernetically-augmented psychics. By our standards, everyone is a superhuman. Therefore, I don't think it's a guaranteed win for the robots.

An important question to ask is what these robots were meant to do. Perhaps Nex Terren is right, and the battlefields of the future will be completely robotic. In that case, our big boss might have black-market military surplus, in which case the robots are built to fight... Well, other robots. While it's gruesome to think of what an anti-armor weapon would do to a human, a non-armor-piercing light machine gun is simply a better weapon to use against people. If this is the case, the tactics could be as simple as 'Trick them into firing, and attack while they're reloading.'

Or maybe the robots are stolen riot-control droids, or hacked construction units with a very.... Liberal definition of what requires sawing and welding. Maybe they're custom built sentries that shoot anything without a transponder on-sight. I could speculate robots and countermeasures all day, but it's very hard to come up with tactical level counters to a force you know nothing about.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .