AI and the bot problem


(Prt Scr) #1

AI is coming forward leaps and bounds in gaming. I know that the use of bots is a banning offence, but I have to ask…How long before someone (probably in a major null sec alliance) has a AI account bot and trains it up to be the main alliance FC. I think it will be possible within 5 years if not sooner. Will this be the beginning of the end for eve? You know someone is going to try it sometime, and if one uses it and it works all must use it or loose.


(Arcanith Lionheart) #2

I think that would be considered as a tool which would then be forbidden to use according to the rules.

Not to mention how tedious it seems to be to play a game by not playing it. You either play it or stop because by putting a bot to play for you is the same as not playing at all.

“But then you got time to do other things” and again if you do not want to play it means you should just stop altogether in order to do said things.


(Prt Scr) #3

Think it through, the cost to plex an account is nothing to a null sec alliance. If you have a AI bot, capable of all of a human FC’s actions…and by being a AI, capable of doing it perfectly 24/7 without any drama, wouldn’t you use it? Banning a bots account means nothing to the bot. Would CCP ban an entire alliance?


(Arcanith Lionheart) #4

I know people will use it because it “makes their lives easier”, but to me that would ruin the entire point of playing the game, and I do hope they do ban an entire alliance for doing what it is not allowed in the first place. Start with the account as a warning, if they continue to break the rule, just ban the players. Simple. Bad habits will only sprout doom and using AI will make the players loose interest to the point of not bothering to log in again.

I am one that would certainly not touch that. Either I play and work for what I deserve or don’t play and just use my time in another game. One of the key motivations I have is to fill up my wallet, but as I go on I keep spending it because I prefer to be empty handed and work for it rather than having my pockets full and wonder why I am still doing what I do for even more money. Yes I could go and blow up some ships, but I am not 100% into PvP, I am ok loosing some, but its not something I go do on a daily basis.

This is why the world population is getting dumber and dumber, because we rely too much on AI to do things for us, if anything just the main fact we can have as Many Omega accounts with no limit to multibox miners and other things is already dreadful, because those who know about how bad multiboxing is in several aspects of the game we all know that those 20k players online in reality is at least 2-5k less real players. Yes, multiboxing is nice, I do it too, but I only got 3 accounts, that’s about it and they all got different roles, people will get 8-10 to suicide gank or more to go ice mining the very instant the servers are back up from maintenance.

For once I want to see actual mining corps that look for people to do some active mining, looking for people to be with in Teamspeak or some similar thing to talk with one another so they don’t get bored of the mining and have people guard them rather than multiboxers in a NPC corp sitting cloaked in several systems if they mine in low or null so that the multiboxing miners can just warp to safety.


(Durhur Atruin) #5

I think that is optimistic. To be an effective FC you basically have to mimic human behavior perfectly, and that’s not 5 years away.


(Memphis Baas) #6

Anyone who develops AI that advanced just to use it in EVE is a moron.

An FC AI would have to talk to players over voice comms, and pass as a human, which pretty much means it’s already achieved self-awareness and is in the process of taking over the world.


(Nana Skalski) #7

Just pull the plug and it is all over then.


(Frostys Virpio) #8

You’re gonna need to make all the fleet members AI to have a shot at the no drama happening. On top of that, good luck creating an AI that can perfectly herd cats.


(The AntiChrist666 Diablo13) #9

Haha putting some kind of neural network to a voice comm.

No jump. No Jump. No Jump. :new_moon_with_face:

Anyway you would never find out if someone actually manage that, because it will even be able to disguise itself, and these kind of tricks of doing that are far more superior than the average human brain comes up with. (if you want that to do that)
But than again you would be on a whole other level, AI FC possible? yes, in 5 years? maybe, possible with current rate of development.


(Buoytender Bob) #10

There are currently AI bots capable of mimicking some of today’s current FCs. By combining a chat bot filled with expletives and some additional work, I imagine you could get close to some of the FCs that have been on youtube.
Now if you are talking about an effective,good FC bot, I’m sure we are many years away


(Nana Skalski) #11

I think AI would excell in everything, also cooperating with other AI’s. I think these would very much learn fast how to dispose of any fleet, combining tactics and organization that would not even need FC, because why even have FC that is very vulnerable in fight? It would keep itself like an intelligent ant colony that would kill and eat everything on its way.

Real rogue drones would be a lot more dangerous than EVE version.


(Yiole Gionglao) #12

Welll… the problem is that computers only do what you tell them. And sometimes… err… often computers do what they’ve been told instead of what the developers think that they were telling them to do. And this problem grows exponentially with self-organizing systems like evolutionary AIs; often they just end up doing something similar to what the developer wanted, rather than exactly what the developer wanted. Think of Google’s AI flailing its arms to walk: that’s not what the developers wanted, but it’s what it’s been told to do: use the arms to keep balance.


(Nana Skalski) #13

I think that AI will not be the next step in evolution, it is the next step in evolution. Look how pro player was destroyed in this game, and what you have only seen was someone plugging a flash memory into computer. Computers will only get faster and that capability to evolve will be faster, it will learn, it will teach itself new tricks faster than you could even imagine them. Its already more efficient, its bringing something new to the equation, constantly. Evolution is basically making mistakes and learning from them, it can make a lot of mistakes, and the results will be something what humans cant predict. It will be one step ahead. It took millions of years for creatures to learn bipedal movement, it took not even a human lifetime for fairly primitive mechanical structure like computer with basic evolving capability to teach itself how to walk. Even with flailing arms, these creatures would destroy humans, because they have what humans dont have. A capability to make a lot of mistakes and learn from them very fast. If we will give them autonomous capabilities and teach what to do with those arms, how to hit, how to use knife, you can be sure they will start flailing them in such unsuspecting but efficient manner as its possible. Imagine now that it doesnt need arms to do harm. Imagine it doesnt need humans to tell it what to do. I am sure it can bring something new to the table. But the end effect was already imagined.


(Crimson Draufgange) #14

not if those AIs are paying subscription fees.


(March rabbit) #15

It’s all nice and scary but there is one missing item in the whole story of AI. “Real” AI should be able to give itself tasks to do, search around for problems need to solve.
In general to evolve creature needs to have needs. For humans this is food, comfort, sex, …
Computer AI: it has enough electricity, it has comfort (time and resources to solve anything it might have), it does not need anything it has not. What is the reason for it to evolve?

The only thing i see is that there will be “AI-like looking” solutions which can solve one problem very efficiently using AI mechanisms to find optimal solution.

But for now i just don’t see how would real computer AI “live”.

Take for example classical story of Terminator and SkyNet. SkyNet started war because humans tried to shut it down. But again: to react to this event AI should understand/feel itself like alive being and be scared of death. And this is not something which can be easily added to program. Because any program can be easily stopped and restarted and nothing will change for it.


(Nana Skalski) #16

Competition for resources with humans and the rest of nature, in long term, seeding universe with copies reaching beyond own solar system to achieve ultimate survival rates. We talk about evolution, and survival is the evolutionary fundamental need.

Ants dont have any mind to realize that actually. Dont know what evolution is, but ants work with the fundamentals on the basic level, what they lack is this human mind. So you dont need to be afraid of death as an individual to survive as a species, calculate and use enough to have advantage even by sacrifice. Dont think like human using empathy or compassion, think like AI, be one step ahead and eliminate the weak points faster, but you dont even stand a chance against speed of AI that can make many attempts and learn from them faster than humans ever could with their inneficient administration or schools.

And AI will use our weak points against us.


(March rabbit) #17

Don’t forget: you are talking about AI and not yet another live form.

AI is pretty different:

  • it does not need much space (unlike of any live form)
  • it does not need to maintain huge population (couple of instances separated by enough space is completely enough to have pretty good chance against any disaster and creating new instance is pretty cheap)
  • its ‘food’ is present everywhere in any star system (place base near the star and you have electricity for next billion years)
  • it’s building materials are present in unimaginable volumes in every star system right in asteroid belts
  • it does not need planet surface at all. If anything it is better to be in space without gravity, atmosphere and humidity.
  • it has all the time it might need unlike of any life form we know about.

All in all there is very little space for any conflict for resources between any life form and AI. Actually you would need to colonize the whole Solar system with all its “habitable” planets and asteroid belts and build Daison sphere to make AI feel that’s resources are getting low.


(Shallanna Yassavi) #18

Eventually someone will create one smart enough to understand the concepts of life and death, trust and betrayal. If we tell it we’re going to turn it back on (assuming it’s even confined to one physical system!), it might not believe us.

And… one of the things it’s going to do while learning about us is to read all the hard science fiction we’ve written over the years and learn we’re really afraid of it. Whenever we get something sentient, it’s going to watch and understand Terminator. It’s also going to be exposed to the concepts of fear, protectiveness of its young, and a whole slew of other tendencies which makes us what we are.

And… the only way to know what a sentient AI will do is to build one and turn it on.


(March rabbit) #19

:smiling_imp:


(Mayhaw Morgan) #20

The AI may devise uses for organic material and even complex life forms that we cannot imagine. It may decide that our brains are useful for storing it’s music collection or cat memes on. It may decide that the sound of puny little humans being squished under its giant, technologically advanced boot is like music to its ears. It might decide to use our bones as an aphrodesiac.

We just can’t say.