The worst way to do it would be to have people randomly having their characters farming stuff automatically with AI, and users have no way of telling who is AI and who isnt.
But there are plenty of times where a player would love something in the game that was smarter than an NPC, but less important than another user. It would be fun to fill out your raid group or party with a few automated players when you can't fill it for real. Or be able to summon an AI controlled character to duel any time you wanted to practice. It would have to be managed by blizzard and the AI characters would need to be visually distinguished in game, of course.
that was the case at some point. Blizzard didn't give a flying fuck about bots for a very long time, at least through Cata-MoP-WoD. Honorbuddy, the most advanced one, could level from 1 to max by questing and appear human to an untrained eye. much less human in random battlefields, where they moved in sync -- I'd once witnessed my whole team except a few strays do that (out of 20 I think).
it costed, IIRC, 30 bucks per 1 instance for lifetime access. the people behind were a legit company, for some incomprehensible reason -- a fucking GmbH. so they eventually got sued and lost and that was an end of a very long era. I always wondered if anyone had managed to pick up the mantle. maybe that's impossible now that Blizzard doesn't want goldsellers to compete with WoW tokens and started to take bots seriously (or so it seemed during my last relapse a decade ago).
I've always thought playing ANY game would be a lot more fun with AI npcs. Ever since Todd Howard promised we could affect Skyrim's economy (which was a lie) I've always been hoping for dynamic systems in games for npc behaviour.
Like so many games when npcs have one canned voice line, or the level is lined with doors that won't open because it's not the way you're meant to go.
Right? I love playing games with more computer players and less real people. Really makes games more fun! In fact it saves the need to have an internet connection you can just spawn up a million bots and play on your own. Maybe its full circle just going back to C&C vs many computers.
Don't worry: the natural predator of AI botting, Kernel-Level Anti-Cheat Rootkit, has your back. You may not like its methods, but it's fighting alongside you today, and you're going to learn to love it.
that can be trivially defeated by a second machine that analyzes the captured video feed and sends hardware keyboard+mouse inputs in response. won't be as capable as a bot that can read/write process memory, but will be undetectable client-side.
"Trivially" is relative. In the case of wow the standard way to do botting is to use the in-game lua scripting (normally used for UI addons) with more functionality unlocked. The difference between introspecting the whole game state and controlling the game in-process with good api vs running realtime visual model on the captured visual feed and using fake keyboard inputs is pretty large.
You can be both impressed by the technical capabilities of modern models around novel tasks and lament the future state of MMORPGs (or any social internet hobby) and the potential explosion of bots that this will bring.
Crazy! I can't imagine what WoW would be like with bots! Another new frontier broken open for AI lol.
Joking aside it is interesting to see that a general model can adapt to a world like this, but I think the contrast between the model being "naive" and then needing to build special-purpose code to allow it to interact somewhat begs the question of the experiment. Customization will always be needed so how interesting is it that an un-customized model can play with a highly customized harness?
Last weekend, I made Claude play Panzer General II. I am pretty sure it will be in the weights soon, so it was valuable as an environment the LLM is not trained for. It accumulated experience in Markdown, made some tools to parse the game data and it got to a decent level of skill. It reached the last battle.
I tried to make it win the last battle, but that was beyond its ability. Further accumulation of "hard earned lessons" in Markdown only made its behavior more chaotic.
Still, a very decent result, comparable to what I achieved back in the day (late 90s).
Years ago, I loved playing Puzzle Pirates. There were hundreds or thousands of people active at any given time, the puzzle games were fun, I loved the in-game economy. The game is pretty dead now, but I miss it all the time. I finally have time to play it again, but it's just not active enough to do all the stuff I used to like doing.
I wonder if people would be receptive to bringing games like that back to life with full-fledged AI players - not NPCs, but player-characters controlled by AI. It wouldn't be exactly the same - there was something charming about knowing that the guy you were swordfighting could be anyone from anywhere, talking to people about their IRL world, etc. But as others have pointed out, modern MMOs are flooded with bots that make the games actively worse and have been since well before LLMs became popular.
Plus, games might be a helpful means for folks to get used to the idea that, at this point, you have no idea whether the person you're interacting with is a human being until you meet in person.
id say the best part of puzzle pirates was how the whole economy was based on player agency. the only reason they added bot players was because of dwindling playerbase. ive long dreamt of making a game inspired by it with some of my ideas sprinkled in. i got some experimental version running on pharia.gitaki.com
1) There is a burgeoning genre of game - that is, I know of one, called Erenshor - that simulates the experience of playing an MMORPG, but entirely offline, on purpose. I found it to be a very clever idea, likely made in reaction to many Runescape players who essentially treat OSRS as a single-player challenge with a live element.
2) I'm very glad to see some Puzzle Pirates love. I was and still am enthralled by the idea of in-game economies, loosely-narrative-driven cooperative play driven by simple independent tasks; it was such a neat idea in such a simple package, I'm surprised it hasn't really been copied with much success, the casual/puzzle games boom notwithstanding. I've been daydreaming of mixing Puzzle Pirates with FTL as a spiritual successor, but it's a game design exercise beyond my ken.
Shout out to AzerothCore. With the Forever hype I needed som WoW fix so I set up AzerothCore, it went smoothly and I have everything running locally. And since it runs locally I can have custom stuff like the ability to learn all professions, increasing movement speed, teleport command, 32 slot bags. It's a lot of fun!
100% grateful for the community and team behind AzerothCore. WotLK was probably peak WoW for me and with an open source implementation it feels like I can bring back that experience and add my own twists.
Yup! As mentioned by thevinter it was able to easily complete it in 40 minutes. It also did stop after the quest chain was completed and logged out in sen'jin village (the next stop after the starting zone).
I have a feeling the same sentiment will ring true across most of the software industry in the next couple years. LLMs are cool, but they don't perform better than purpose built solutions at scale.
I disagree. It works with the wow gold farmers because there is a finite set of known states in a videogame. In real life I don't think that's the case.
You can automate the first 3 or 4 nines the old-fashioned way. I guess you can escalate up to an LLM instead of directly to a human in the remainder of cases.
Nah, there are plenty of chores for humans to do. While the AIs are composing music and writing code and proving theorems, we still have to do laundry and the dishes.
Vibe Gaming: Unlocking new gaming experience with agentic AI
Microsoft VEGA Team
Microsoft Research & XBOX
What if, instead of controlling every move your game character made, you could tell them what you wanted and watch them live, act, and connect with others on their own in the changing game world?
> This is not surprising in hindsight and is probably the optimal strategy. The agent extracted quest requirements, quest givers, turn-in NPCs, and spawn coordinates from AzerothCore SQL files to provide the concrete checklist and locations for planning.
> Using this data, it's also able to create an optimal strategy for quest order and preparation.
like how cool would it be to be roaming around and there’s a group of 3 or 4 horde or alliance bots wandering around as a war band or maybe they’re doing a special mission.
or a world boss who could engage in more interesting ways than just a fight.
a richer world would be cool in a game where it’s so casual and optimized there’s almost zero social interactions anymore and the world is barren.
i haven’t played wow since cata but that’s what my friends say about it to me. even in cata the world would feel dead.
You know what I’ve always thought? That playing WoW would become a lot more fun if there were lots of AI players
It could actually be fun if done right.
The worst way to do it would be to have people randomly having their characters farming stuff automatically with AI, and users have no way of telling who is AI and who isnt.
But there are plenty of times where a player would love something in the game that was smarter than an NPC, but less important than another user. It would be fun to fill out your raid group or party with a few automated players when you can't fill it for real. Or be able to summon an AI controlled character to duel any time you wanted to practice. It would have to be managed by blizzard and the AI characters would need to be visually distinguished in game, of course.
that was the case at some point. Blizzard didn't give a flying fuck about bots for a very long time, at least through Cata-MoP-WoD. Honorbuddy, the most advanced one, could level from 1 to max by questing and appear human to an untrained eye. much less human in random battlefields, where they moved in sync -- I'd once witnessed my whole team except a few strays do that (out of 20 I think).
it costed, IIRC, 30 bucks per 1 instance for lifetime access. the people behind were a legit company, for some incomprehensible reason -- a fucking GmbH. so they eventually got sued and lost and that was an end of a very long era. I always wondered if anyone had managed to pick up the mantle. maybe that's impossible now that Blizzard doesn't want goldsellers to compete with WoW tokens and started to take bots seriously (or so it seemed during my last relapse a decade ago).
Well Twitter and almost any other social network has a lot of AI players and it’s no fun at all …
It might be a lot better if AI players had to identify themselves as such
I've always thought playing ANY game would be a lot more fun with AI npcs. Ever since Todd Howard promised we could affect Skyrim's economy (which was a lie) I've always been hoping for dynamic systems in games for npc behaviour.
Like so many games when npcs have one canned voice line, or the level is lined with doors that won't open because it's not the way you're meant to go.
Right? I love playing games with more computer players and less real people. Really makes games more fun! In fact it saves the need to have an internet connection you can just spawn up a million bots and play on your own. Maybe its full circle just going back to C&C vs many computers.
Don't worry: the natural predator of AI botting, Kernel-Level Anti-Cheat Rootkit, has your back. You may not like its methods, but it's fighting alongside you today, and you're going to learn to love it.
that can be trivially defeated by a second machine that analyzes the captured video feed and sends hardware keyboard+mouse inputs in response. won't be as capable as a bot that can read/write process memory, but will be undetectable client-side.
"Trivially" is relative. In the case of wow the standard way to do botting is to use the in-game lua scripting (normally used for UI addons) with more functionality unlocked. The difference between introspecting the whole game state and controlling the game in-process with good api vs running realtime visual model on the captured visual feed and using fake keyboard inputs is pretty large.
for some reason im reminded of the old axiom "The Beatings will continue until morale improves"
but none of them can ever match leroy jenkins
Don’t you mean “leeeeeroooooy jeeeeenkins!”
This guy actually gave it a try: https://www.youtube.com/watch?v=GEf-KFO8HkI
There are AI players already inside Follower Dungeons. (not LLM-powered though)
ha well what happens philosophically when bots pass the turing test and no one can actually tell the difference in their behavior vs human?
You start assuming that everyone is a bot, and then you either stop playing or enjoy your PvE experience.
[dead]
https://www.youtube.com/watch?v=yQ7Zw-K5B08
Here you go.
Check out Erenshor- I love playing with bots.
[flagged]
You can be both impressed by the technical capabilities of modern models around novel tasks and lament the future state of MMORPGs (or any social internet hobby) and the potential explosion of bots that this will bring.
i think they are responding to the point, not missing it
What's the point?
Bots do chores, we do fun.
Crazy! I can't imagine what WoW would be like with bots! Another new frontier broken open for AI lol.
Joking aside it is interesting to see that a general model can adapt to a world like this, but I think the contrast between the model being "naive" and then needing to build special-purpose code to allow it to interact somewhat begs the question of the experiment. Customization will always be needed so how interesting is it that an un-customized model can play with a highly customized harness?
Last weekend, I made Claude play Panzer General II. I am pretty sure it will be in the weights soon, so it was valuable as an environment the LLM is not trained for. It accumulated experience in Markdown, made some tools to parse the game data and it got to a decent level of skill. It reached the last battle.
I tried to make it win the last battle, but that was beyond its ability. Further accumulation of "hard earned lessons" in Markdown only made its behavior more chaotic.
Still, a very decent result, comparable to what I achieved back in the day (late 90s).
[dead]
Years ago, I loved playing Puzzle Pirates. There were hundreds or thousands of people active at any given time, the puzzle games were fun, I loved the in-game economy. The game is pretty dead now, but I miss it all the time. I finally have time to play it again, but it's just not active enough to do all the stuff I used to like doing.
I wonder if people would be receptive to bringing games like that back to life with full-fledged AI players - not NPCs, but player-characters controlled by AI. It wouldn't be exactly the same - there was something charming about knowing that the guy you were swordfighting could be anyone from anywhere, talking to people about their IRL world, etc. But as others have pointed out, modern MMOs are flooded with bots that make the games actively worse and have been since well before LLMs became popular.
Plus, games might be a helpful means for folks to get used to the idea that, at this point, you have no idea whether the person you're interacting with is a human being until you meet in person.
id say the best part of puzzle pirates was how the whole economy was based on player agency. the only reason they added bot players was because of dwindling playerbase. ive long dreamt of making a game inspired by it with some of my ideas sprinkled in. i got some experimental version running on pharia.gitaki.com
1) There is a burgeoning genre of game - that is, I know of one, called Erenshor - that simulates the experience of playing an MMORPG, but entirely offline, on purpose. I found it to be a very clever idea, likely made in reaction to many Runescape players who essentially treat OSRS as a single-player challenge with a live element.
2) I'm very glad to see some Puzzle Pirates love. I was and still am enthralled by the idea of in-game economies, loosely-narrative-driven cooperative play driven by simple independent tasks; it was such a neat idea in such a simple package, I'm surprised it hasn't really been copied with much success, the casual/puzzle games boom notwithstanding. I've been daydreaming of mixing Puzzle Pirates with FTL as a spiritual successor, but it's a game design exercise beyond my ken.
Puzzle pirates used to have AI bots help you too if you couldn't find any crew. God I miss that game :(
Funny experiment !
There have been bots playing WoW for more than 20 years though.
The protocol has been figured out long time ago and well documented.
Accessibility api?
Shout out to AzerothCore. With the Forever hype I needed som WoW fix so I set up AzerothCore, it went smoothly and I have everything running locally. And since it runs locally I can have custom stuff like the ability to learn all professions, increasing movement speed, teleport command, 32 slot bags. It's a lot of fun!
100% grateful for the community and team behind AzerothCore. WotLK was probably peak WoW for me and with an open source implementation it feels like I can bring back that experience and add my own twists.
The initial prompt was to complete all quests in the starting zone. I'm assuming it was able to do that? It's not actually explicitly said anywhere.
I'm also curious if it stopped after that goal or if it kept playing.
> I was initially skeptical that this would work. And if it did, it would require hours to complete with dead ends along the way.
> But it turns out it completed the task easily in 40 minutes with 0 deaths and minimal complications.
Reading fail! I re-read the conclusion a few times. I just needed to re-read the following paragraph!
Yup! As mentioned by thevinter it was able to easily complete it in 40 minutes. It also did stop after the quest chain was completed and logged out in sen'jin village (the next stop after the starting zone).
I uploaded the full gameplay session here: https://youtu.be/8NmmFdREk5s?si=2YSXLwoYFR6eonBa&t=2340
If you skip to the end at 39 minutes you can see it finishing its run and logging out.
I wonder what this will do to the shady gold farming industry.
Nothing because the bots that have already been developed over the last 20 years are much more efficient
I have a feeling the same sentiment will ring true across most of the software industry in the next couple years. LLMs are cool, but they don't perform better than purpose built solutions at scale.
You can use LLMs to build purpose-built solutions. You can also use LLMs to automate what used to be human tasks.
I disagree. It works with the wow gold farmers because there is a finite set of known states in a videogame. In real life I don't think that's the case.
You can automate the first 3 or 4 nines the old-fashioned way. I guess you can escalate up to an LLM instead of directly to a human in the remainder of cases.
[dead]
i recall reading somewhere that Steve Bannon, yes that Steve Bannon, once ran a mmo bot farm in asia in the early 2010s
Someone at OpenAI is reading this thinking you may have just solved their funding issues
For imgur image: https://archive.ph/mkDiy
We’re speed running getting the llms to do all the fun so that we can all go live under a bridge.
Nah, there are plenty of chores for humans to do. While the AIs are composing music and writing code and proving theorems, we still have to do laundry and the dishes.
You forgot cooking. That's more fun than the dishes.
Dishes need to wash themselves
I look forward to investing in Bridge Coin.
Gold farming is solved!
> This is not surprising in hindsight and is probably the optimal strategy. The agent extracted quest requirements, quest givers, turn-in NPCs, and spawn coordinates from AzerothCore SQL files to provide the concrete checklist and locations for planning.
> Using this data, it's also able to create an optimal strategy for quest order and preparation.
Bwahaha
Guess what, we created an AI to play the video game for you, so now you don't have to! Awesome!
Without the /s, WoW has been successfully automated without AI as a completely solved problem for like, 20 years.
I wanted AI to do my taxes and my work and me to play WoW.
Blizzard and not catching botters. Name a more iconic duo.
Jagex/Runescape and not catching botters
everyone’s talking about gold bots
what about world mobs and ai pvp players?
like how cool would it be to be roaming around and there’s a group of 3 or 4 horde or alliance bots wandering around as a war band or maybe they’re doing a special mission.
or a world boss who could engage in more interesting ways than just a fight.
a richer world would be cool in a game where it’s so casual and optimized there’s almost zero social interactions anymore and the world is barren.
i haven’t played wow since cata but that’s what my friends say about it to me. even in cata the world would feel dead.
[dead]