Live data from Hacker News

The Real Challenge Is Teaching AI to Play D&D

wired.com

61–70 of 83 posts

Re: The Real Challenge Is Teaching AI to Play D&D

#61
post #29

Earlier quoted context omitted.

IMHO one of the hardest parts of DMing is building tension and a sense that there is real danger, without just bringing out enemies that can completely roll over the players. Ideally the group should survive, but constantly feel threatened and like they "just barely" made it through a story's climax. In other words those victories need to be earned, which is a fine balancing act since dice rolls are random and bad th…

I'll usually fudge at least a few dice rolls as a DM depending on how the game or encounter's going as the situation requires. If I roll a bunch of crits in a row for example, i'll maybe just call it one instead a bunch, maybe that wasn't a one hit kill when it should have been, stuff like that. If it's too random it usually stops being fun for everyone. It's a really hard balance though you don't want the players to…

I remember one looking through the rule book for a Buffy the Vampire Slayer RPG which had an interesting concept: players had a certain number of "plot points" (I forget exactly what they called them) which allowed them to overturn an unfortunate die roll. It was based on the idea that in the TV show, the main characters would often manage to get out of a tight spot through well-timed luck.

Re: The Real Challenge Is Teaching AI to Play D&D

#63
post #8

Of course the article is actually about writing a digital DM not a player. But isn't the main problem with this defining win conditions? Pretty easy to define win condition for chess or go. Now, I haven't really played DnD (much to my sadness), but how do you win? edit: I know the answer but I'm making a point with my question

I think one can argue that people have been playing role playing games GM'd by computers for going on 40 years now. Going back to the gold box games in the literal sense, but almost all video games are essentially dungeons and dragons under the hood -- hitpoints, encounters, npcs, etc.

Re: The Real Challenge Is Teaching AI to Play D&D

#64
Not Dungeons and Dragons, but Doug Lenat's Eurisko entered fleets for the role-playing game Traveller Adventure 5: Trillion Credit Squadron. It won the 1981 and 1982 championships.

Eurisko was GOFAI, written in RLL, a frame-based language implemented in Lisp.

https://aliciapatterson.org/stories/eurisko-computer-mind-it...

Re: The Real Challenge Is Teaching AI to Play D&D

#65
post #39
post #30

Earlier quoted context omitted.

True, but I can imagine a world where AI DMs are in vogue, and are curated not for their one-size-fits-all perfection, but for their aesthetic or functional qualities. The hardest thing to capture will be truly surprising off-screen storytelling moments that role playing games enable. Otherwise, you can probably optimize for 'challenging combat,' 'fun puzzles,' 'social interaction' etc... Have a hard-nosed, Old Schoo…

interesting, i didnt know about that setting in rimworld. the game has been high on my list due to the fact ive had much FUN with DF, so im gonna bring rimworld up on that list!

Now is the perfect time! Rimworld just got a gigantic (free) 1.1 update and a smaller (paid) DLC to go with it.

Rimworld is like 2d dwarf fortress, but way more accessible. The controls make sense, and the game teaches you as you play. Also Rimworld has about 3 and a half metric tons of mods.

I first saw RimWorld while chilling in a hotel room watching YouTube a while back, and I bought it back while it was still in early access. Tynan has done a phenomenal job making the game go from good to fantastic.

Re: The Real Challenge Is Teaching AI to Play D&D

#66
post #8

Of course the article is actually about writing a digital DM not a player. But isn't the main problem with this defining win conditions? Pretty easy to define win condition for chess or go. Now, I haven't really played DnD (much to my sadness), but how do you win? edit: I know the answer but I'm making a point with my question

I think one can argue that people have been playing role playing games GM'd by computers for going on 40 years now. Going back to the gold box games in the literal sense, but almost all video games are essentially dungeons and dragons under the hood -- hitpoints, encounters, npcs, etc.

> hitpoints, encounters, npcs, etc.

Those are the rules, not the game.

Re: The Real Challenge Is Teaching AI to Play D&D

#67
post #25

> Forget X, the real challenge for AI is Y. As soon as Ys become Xs over time, and we come up with new Ys: we’re making progress.

Deep Blue was 24 years ago.

Your point? If you're saying we haven't made progress since: I beg to differ.

Re: The Real Challenge Is Teaching AI to Play D&D

#68
post #58
post #25

> Forget X, the real challenge for AI is Y. As soon as Ys become Xs over time, and we come up with new Ys: we’re making progress.

You forget the step where when we get comfortable enough with something we stop calling X AI at all because obviously it' just Z. NB: I'm not convinced this is the wrong thing to do, many of the things we've historically labeled "AI" are very obviously not, at least in retrospect. Perhaps the cycle ends only when an AI can explain to us how we are getting it wrong.

We just don't have a very good definition of what "intelligence" actually is. Hence we're trying to set the bar somewhere. "Surely a machine that can play chess is intelligent". Nope, that's not it. "But surely a machine that recognizes objects in images is intelligent". Nah, it's just good at recognizing patterns of pixels. And so on and so forth. It's a noisy process, but we're slowly moving in the right direction.

Re: The Real Challenge Is Teaching AI to Play D&D

#69

A huge part of the charm of TTRPGs is playing a non-digital game face to face with other humans. Even if we had AI that was as entertaining as a skilled human DM (unlikely to be possible IMHO), the human might end up being preferable since hanging out with humans is one of the goals of this particular pursuit. Plus, we already have (weak) AI DMs. They're called video game RPGs. You're the player (as are many others i…

There are ways to play traditional pen and paper rpg over internet now (roll20 for example), and it's actually better than playing them physically in many respects - the system draws the map, allows moving the tokens zooming and scrolling, calculates the distances, shows the fog of war, throws dices and includes bonuses in the rolls. After initial setup it makes it much easier for GM and players because you don't need to remember to include all the bonuses every time you attack with your sword.

You still use voice (usually over discord) and DM decides what happens on the fly.

But most importantly you don't have to play with people that live close to you, makes it much easier to find someone to play with, it's why I started playing p&p rpg again after over a decade.

I could see sites like roll20 recording the sessions with metadata (at this point they said this and GM told "roll perception" and they rolled d20 + 3) as a way to train a virtual DM. Of course it would need crazy amount of games.

Re: The Real Challenge Is Teaching AI to Play D&D

#70

A huge part of the charm of TTRPGs is playing a non-digital game face to face with other humans. Even if we had AI that was as entertaining as a skilled human DM (unlikely to be possible IMHO), the human might end up being preferable since hanging out with humans is one of the goals of this particular pursuit. Plus, we already have (weak) AI DMs. They're called video game RPGs. You're the player (as are many others i…

There's no end of people making video games that are direct attempts to convert D&D to a digital format. Take the new Baldur's Gate for example[0]. All the same spells present, digital dice rolling, etc. Personally, I say let things fall to the medium that best suits them. If D&D is sufficiently complicated rules wise that it is actually more fun to have software take care of that so you can just focus on the tactica…

CRPGs are no competition, there's no comparison, really. When you do anything outside the box CRPGs just break, while a good game master will run with it and that's how the best stories happen. We found a boss we didn't wanted to fight. Our spellcasters decided sending a construct with 2 bags of holding to put one into another near the boss will destroy it. It worked, but it also created portal to another dimension, and when we tried to recover the hostages 2 player characters got thrown into that dimension. The rest of the party had to pay a powerful mage to recover them with our best magical weapon. Half a session (several hours) was spent arguing over who owns the magical weapon and if it's good idea to give it away :) They almost fought in game :) Meanwhile my character stranded in another dimension decided to become a priest (I was a barbarian before).

These kind of stuff never happens in CRPGs - it couldn't because there's infinite number of possible open-ended solutions to any problem. If AI can improvise them I'd argue it's as good as a Turing Test.

BTW complicated rules aren't that big of a problem - there is a compromise between CRPGs and TTRPGs - and it's computer-assisted TTRPGs played over internet. The most important advantage is that you can play with people from other continent, so it's much easier to find players. But the computer-assisted part is also making the rules much less of a problem. You set up your character (all the items, feats, skills, abilities) before the game, and when DM wants he asks you to roll the attack, you click and the system rolls the dices for everybody to see, adds the needed modifiers, calculates the damage (including critical hit if needed etc. ) and shows how much hitpoints you lost.

It's faster than doing all this stuff physically and beyond the initial setup and occasional level-up it makes complicated rules more bearable.

Post reply on HN