A.I. | The Future and ESO

Blood_again · March 8

Please stop taking AI like some kind of magic that just improves everything by one touch.
Using AI in gaming area for improving the gameplay process requires good understanding of how AI works and strong specific IT research skills. Unfortunately reading and citating of IT-pop articles can't replace it.

One of the basic element of using AI in your area is building an algorythm that fits the subject area. You can't just take one of the existing graphic AI and feed your gameplay data to it. It won't work that way. You need to do a research.
Who would do it, you think? The same people who create current gameplay? May be you?
I doubt it. It requires special skills, takes time and costs a lot.

Another part of applying AI in your area is a pure learning process. Pop-IT don't like sharing that part, yes. You need to have a bunch of data that is adequate to your algorythm. You need to have criteria to filter acceptable results. And you need to tune your algorythm until having a good results.
Do you know any good criteria? I have a hint. Players.
Guess who will shout that devs don't listen to players' feedback while testing all that AI generated gameplay trash?

Honestly, I'm using AI tools every day due to work and also developed one.
AI is currently good as another step in data aggregation and statistics but nothing else yet. Everything else always require human to fix the result.
If you're trying to offer an AI aggregation tool to ZoS, why do you think that they don't use one already? Should they always tell us which diameter are screwdrivers they use?
If you're trying to offer more, you actually try to be a victim of Hi-Tech hype.

kyatos_binarini · March 8

Integrating AI into ESO seems as possible as attaching a warp drive to the Flintstones' car.

ZOS_Icy · March 8

Greetings,

We have recently removed some unnecessary back and forth from this thread. This is a reminder to keep the discussion civil and constructive. Please keep our Community Rules in mind moving forward.

Thank you for your understanding.

Danikat · March 8

Cazador wrote: »

Ok I'm curious OP, are you simply interested in what AI can do or do you work for one of these companies that market it? This thread honestly just feels like a sales pitch.

I'm starting to think that might be the case, considering they've ignored all mentions of interesting things AI can do for players (like Oblivion) to focus on more efficient analytics.

If that's the case though they're going about it completely backwards. That's a useless thing to advertise to end users because we'd never see it or any direct results of it being implemented. For all we know ZOS are doing it already and just haven't felt the need to tell us, just like they don't tell us what project management or file management software they use (I assume they're at least encouraged to use Office 365 whenever possible).

Blood_again · March 8

StihlReign wrote: »

Can AI analyze player data to uncover patterns, preferences, and pain points? Can this information improve design decisions, helping developers create more engaging and enjoyable gameplay experiences? Will AI-driven analytics tools provide valuable insights into player demographics, playtime, and in-game behaviors?

I suspect the answer to each of these questions is yes.

I'm sad, you didn't read my message to the end.
No problem, I will play your quiz game.

Can your AI solution provide better quality of analysis and improvement than tools that company is currently using? Why?
Can this solution provide data in form that is as appliable to the current dev processes as the current tools? How many changes should be done if not?
How much investments are required to integrate this AI solution in current process while keeping continious service delivery? What is the conflict solving and negotiation plan when old and new analysis results are different up to opposition?

What do you suspect?

StihlReign · March 8

Cazador wrote: »

Ok I'm curious OP, are you simply interested in what AI can do

Yes. Especially when it comes to helping solve problems by allowing employees to focus on more valuable and complex work.
Where AI systems can help prevent overworking by reducing the workload through automation, it is invaluable to a company's bottom line, end product, and revenue.

Through predictive analytics, machine learning can improve an organization's methods of decision making thereby optimizing the decision-making process.

In a recent poll by Forbes, 63% of executives said that machine learning resulted in measurable revenue growth.

A forum post points out ESO has reached an estimated 6,000,000 armor combinations while wondering why more combinations are not used, and why the community limits themselves to 25 or so commonly used sets (paraphrasing and going from memory... been a while since I read the post).

Is the question valuable to the company? Can AI(ML) read the forum (allowing employees to focus on more valuable and complex work) and find questions to present to decision-makers faster than a human?

Who would have a better answer to this question - the users who responded, or Zos who has access to the actual data and the benefit of an algorithm? Would I prefer the company make changes based on that data, or the community's responses, or some combination of both?

StihlReign · March 8

Blood_again wrote: »

StihlReign wrote: »

Can AI analyze player data to uncover patterns, preferences, and pain points? Can this information improve design decisions, helping developers create more engaging and enjoyable gameplay experiences? Will AI-driven analytics tools provide valuable insights into player demographics, playtime, and in-game behaviors?

I suspect the answer to each of these questions is yes.

I'm sad, you didn't read my message to the end.
No problem, I will play your quiz game.

Can your AI solution provide better quality of analysis and improvement than tools that company is currently using? Why?

It could. Because the machine is better and faster than you when performing its intended purpose. You cannot read the data and compile it faster. The cost-benefit analysis is internal to Zos, no comment there.

Can this solution provide data in form that is as appliable to the current dev processes as the current tools? How many changes should be done if not?
How much investments are required to integrate this AI solution in current process while keeping continious service delivery? What is the conflict solving and negotiation plan when old and new analysis results are different up to opposition?

What do you suspect?

I did read your message.

Yes. There's no reason to design, purchase, or use a tool (in-house or from a vendor) that doesn't meet your intended purpose. Society rejects old processes and tools and requests new tools daily.

Blood_again · March 8

StihlReign wrote: »

Blood_again wrote: »

Can your AI solution provide better quality of analysis and improvement than tools that company is currently using? Why?

It could. Because the machine is better and faster than you when performing its intended purpose. You cannot read the data and compile it faster. The cost-benefit analysis is internal to Zos, no comment there.

Well, you start thinking in realistic direction.
How can you evaluate which is better? Don't you think that current tools of ZoS are internal too?

StihlReign · March 8

Blood_again wrote: »

StihlReign wrote: »

Blood_again wrote: »

Can your AI solution provide better quality of analysis and improvement than tools that company is currently using? Why?

It could. Because the machine is better and faster than you when performing its intended purpose. You cannot read the data and compile it faster. The cost-benefit analysis is internal to Zos, no comment there.

Well, you start thinking in realistic direction.
This is a poorly formed opinion.
How can you evaluate which is better?
I don't need to.

Don't you think that current tools of ZoS are internal too?

If you're asking if I understand who and what the Parent Company is - the answer is yes.

Blood_again · March 8

StihlReign wrote: »

If you're asking if I understand who and what the Parent Company is - the answer is yes.

I guess there is some misunderstanding between us.
The question

wrote:

Can your AI solution provide better quality of analysis and improvement than tools that company is currently using? Why?

supposed some kind of comparision. The current tools of analysis and the new AI one.
Your answer was so determined, that I started guessing if you have some ZoS insides

nathamarath · March 8

In theory this sounds great. What I wonder: How would an A.I. which is determined to find an optimised way be helpful in error seeking in player's gameplay. Someone planning a game school? We have ESO University. Landscape/occlusion/collision calculation is not A.I. and a good player (optimised play) can not comprehend a bad player's behavior unless he is able to play not well. Do we want to play worse? Utilising human choice patterns? For? Making the best quest decision? In praxis A.I. is manmade and manlearned and as all human means and tools usually dumbed down to the mainstream of sizes, abilities and characteristics, not speaking of the toning of A.I. which is a hot topic on the web.

Regarding fighting with a high level ally: will make the fight harder? I would not understand why getting help from stronger ones should make a task more difficult. It should be easier. I'd prefer having leveled zones (back) to look forward to encounters.

Regarding chaos trials: I like the idea of random-mech trials as long as you can chose that variation ahead like with a checkbox for "chaos-trial" for those who love certain instances and their bosses. Me: I love the handwritten old fashioned enemy and don't feel appealed by enemies who randomly cast abilities from an ability pool. It would be a great idea for a target dummy series, tho: practice for mechanics, let's say physical special attack and ulti dummies, side effect dummies and so on. That would be awesome.

Regarding removing battle spirit: would be the last thing to change in PvP.

edit:
corrected some spelling errors. sry, it's not my native luggage.

valenwood_vegan · March 8

It's a shame that instead of being approached as a discussion of what recent advances in technology might offer for ESO, we got what essentially reads like a generic marketing pitch for AI in gaming (made to the wrong customers, no less), backed by sources with an obvious agenda and a lot of cool buzzwords... and then increasingly adversarial responses to any attempts to point out there may actually be pitfalls as well as advantages.

So I mean, I don't have really anything useful to add since this doesn't seem like a real discussion but rather that we're being preached to and scolded for offering any alternate points of view, except to say that imo it is highly unlikely that we'll see "AI" elements incorporated directly into a ten-year-old game that doesn't seem to have the level of development resources devoted to it that it once did. Perhaps that's something we'll see in TES 6 or Zenimax's new mmo.

[EDIT for typo].

StihlReign · March 8

valenwood_vegan wrote: »

It's a shame that instead of being approached as a discussion of what recent advances in technology might offer for ESO, we got what essentially reads like a generic marketing pitch for AI in gaming (made to the wrong customers, no less), backed by sources with an obvious agenda and a lot of cool buzzwords... and then increasingly adversarial responses to any attempts to point out there may actually be pitfalls as well as advantages.

So I mean, I don't have really anything useful to add since this doesn't seem like a real discussion but rather that we're being preached to and scolded for offering any alternate points of view, except to say that imo it is highly unlikely that we'll see "AI" elements incorporated directly into a ten-year-old game that doesn't seem to have the level of development resources devoted to it that it once did. Perhaps that's something we'll see in TES 6 or Zenimax's new mmo.

[EDIT for typo].

In the future AI will impact ESO in a ton of areas. I detailed how I'd like to see the current game changed in future iterations.

At a recent presentation held for the press, the development team unveiled Necrom, though they actually started with news on the game's healthy player base.

Whereas 20 million players were confirmed last year, two million accounts were created since High Isle, bringing the new total to 22 million players and counting (there's a free play event and discount sale running right now, so a bunch of newcomers are bound to be moving their first steps in Tamriel during these days).

The scope of Elder Scrolls Online is simply huge, as the game sports 9 million words, 200K lines of spoken dialogue per language, and over 26 million lines of code.

SOURCE

This is an incredible amount of data. It would misguided to think humans can't benefit from tools to analyze this amount of data to create better products. I know what I'd like to see from a future product, as I look at and use the current product, and I can view it through the lens of current and future AI product offerings.

Feel free to tell us how you'd like to see it changed with-in the scope of what AI (or Machine learning or whatever your understanding of AI is) can and will be able to do to facilitate that vision.

Scaletho · March 8

AI could create dialogues on spot, related to ES lore and the quest player is following at the time. Even reproduce the NPC's voice.

Cazador · March 8

valenwood_vegan wrote: »

So I mean, I don't have really anything useful to add since this doesn't seem like a real discussion but rather that we're being preached to and scolded

[EDIT for typo].

Not to mention it feels like certain parts of questions are being selectively ignored. The weirdest part of the thread for me was someone with a lot of experience in this field being dismissed as closed minded for not being all in on AI.

Alpheu5 · March 8

kyatos_binarini wrote: »

Integrating AI into ESO seems as possible as attaching a warp drive to the Flintstones' car.

Possible to do but will disintegrate the original product when it's turned on?

StihlReign · March 8

nathamarath wrote: »

Regarding fighting with a high level ally: will make the fight harder? I would not understand why getting help from stronger ones should make a task more difficult. It should be easier. I'd prefer having leveled zones (back) to look forward to encounters.

Maybe more challenging. At my CP level the NPCs in the World are typically boring, ez to fight, and run through at will. The sense of danger is almost non-existent. I'm hoping in the future this can and will change because of the advances in tech. I like the open world and would like it more if intelligent scaling of npc enemies were in place. Every fight need not be 10/10 difficult...but the vast majority shouldn't feel like fighting or farming mudcrabs either. Scaling rewards, as a result, would be awesome.

Regarding chaos trials: I like the idea of random-mech trials as long as you can chose that variation ahead like with a checkbox for "chaos-trial" for those who love certain instances and their bosses. Me: I love the handwritten old fashioned enemy and don't feel appealed by enemies who randomly cast abilities from an ability pool. It would be a great idea for a target dummy series, tho: practice for mechanics, let's say physical special attack and ulti dummies, side effect dummies and so on. That would be awesome.

Regarding removing battle spirit: would be the last thing to change in PvP.

edit:
corrected some spelling errors. sry, it's not my native luggage.

In the game's most difficult PvE environment, I think it would be nice if the DPS felt more pressure because he's being targeted for his specific weaknesses. Sounds like a challenge for the team, "our vampire DPS is weak to Fire", vs The Boss is a Fire boss in the next section, I saw it on the Youtube vid so I'm prepared. Some predictability yes, but - Prepare to fight the boss and be challenged because your playstyle has been analyzed by the boss...your challenge could be ice, self healing etc etc.

Everything you do in-game is generating data.

Dropping your player model into a simulation and running it a million times against the boss, setting the thresholds, and then handing it to you to playtest doesn't sound like a bad way to go.

Thank you for your responses

Warhawke_80 · March 8

kyatos_binarini wrote: »

Integrating AI into ESO seems as possible as attaching a warp drive to the Flintstones' car.

Actually that would probably be easier....

Remember at last report if more than a couple players wear cloaks the entire server crashes....somehow I just don't see a deep learning AI working with a modified version of the hero engine, but it would be cool if one of the devs could tell us if I'm wrong...

Sepultura_13 · March 8

danno8 wrote: »

The pessimistic (or perhaps realistic) side of me tells me that LLMs (I dislike the term AI for nearly all that the term is currently being used for) will be used to primarily drive down costs for companies by replacing as many employees as possible with inferior LLM systems while hoping most people don't notice or care enough to notice.

Agreed, 1000%

Anyone who is having a hard time finding a job, or was recently laid off from one, would / could probably attest to this.

Sepultura_13 · March 8

runa_gate wrote: »

Imagine if AI could argue in zone chat for all the people who no one will talk to IRL so they need a captive audience! They could act out their inability to achieve parental approval without bothering with typing their misspelled almost-sentences

pardon me... I was doing a survey in Auridon a little bit ago but I'm doing better now

Or, what if AI could be set up to respond to / interact with the crown sellers, political conspiracy-theorists, ERPers, bigots, and other time-wasters like RL beggars, who interfere with pleasant, relaxing game-play?

Imagine how zone chat might be, then... you could actually say something meaningful and it would be read and responded to by a person who understood you and was on your same wavelength.

DigiAngel · March 8

I'm honestly completely tired of seeing AI try and get embedded into everything...it's bandwagony and ridiculous. That said, the simple truth is surprisingly easy: IF AI improves human lives...wonderful! IF AI replaces human jobs (which it has already) then...it's not something anyone should really embrace. Please keep AI out of my games.

xclassgaming · March 8

Ew, no thanks. AI is pretty gross. AI WILL replace human jobs, so it needs to be fought back and banned in all honesty. And hopefully it is in the very near future.

wilykcat · March 8

I agree with an AI to remove toxicity.

It can be a great tool for the moderators and developers to use without replacing them.

nathamarath · March 8

StihlReign wrote: »

nathamarath wrote: »

Regarding fighting with a high level ally: will make the fight harder? I would not understand why getting help from stronger ones should make a task more difficult. It should be easier. I'd prefer having leveled zones (back) to look forward to encounters.

Maybe more challenging. At my CP level the NPCs in the World are typically boring, ez to fight, and run through at will. The sense of danger is almost non-existent. I'm hoping in the future this can and will change because of the advances in tech. I like the open world and would like it more if intelligent scaling of npc enemies were in place. Every fight need not be 10/10 difficult...but the vast majority shouldn't feel like fighting or farming mudcrabs either. Scaling rewards, as a result, would be awesome.

nathamarath wrote: »

Regarding chaos trials: I like the idea of random-mech trials as long as you can chose that variation ahead like with a checkbox for "chaos-trial" for those who love certain instances and their bosses. Me: I love the handwritten old fashioned enemy and don't feel appealed by enemies who randomly cast abilities from an ability pool. It would be a great idea for a target dummy series, tho: practice for mechanics, let's say physical special attack and ulti dummies, side effect dummies and so on. That would be awesome.

Regarding removing battle spirit: would be the last thing to change in PvP.

In the game's most difficult PvE environment, I think it would be nice if the DPS felt more pressure because he's being targeted for his specific weaknesses. Sounds like a challenge for the team, "our vampire DPS is weak to Fire", vs The Boss is a Fire boss in the next section, I saw it on the Youtube vid so I'm prepared. Some predictability yes, but - Prepare to fight the boss and be challenged because your playstyle has been analyzed by the boss...your challenge could be ice, self healing etc etc.

Everything you do in-game is generating data.

Dropping your player model into a simulation and running it a million times against the boss, setting the thresholds, and then handing it to you to playtest doesn't sound like a bad way to go.

Thank you for your responses

Regarding bordom: you might be new to the game. High level zones, leveled zones in general, difficult routines and content, comprehensible elemental weaknesses of enemies - you could fight a fire spirit with ice, today machines take damage from poison or bleed damage-, we had it. It was nerfed due to players requests into what we have now. ZOS made great effords to adapt the game to the players wishes - Kudos. ESO-now reminds me somehow of that car Homer Simpson was allowed to design. We are Homer.

The idea of an intelligent enemy sounds great. Try PvP and reconsider someone using your weaknesses without having the appropriate means of defense, not speaking of attack. Not everybody is a Warden. I tried that immaculate approach in PvE and think it's nearly impossible to fortify my char against all threats because I ended either having strong defenses against statuses, elements, physical and magical attacks aka 'The Walking Brick' or a decent damage.

Regarding player models: models are unified, otherwise they would not be models. What would you need players for if you have models - the game could play itself, an idea I like much. It deserves it after all it suffered.

ArchangelIsraphel · March 8

xclassgaming wrote: »

Ew, no thanks. AI is pretty gross. AI WILL replace human jobs, so it needs to be fought back and banned in all honesty. And hopefully it is in the very near future.

DigiAngel wrote: »

I'm honestly completely tired of seeing AI try and get embedded into everything...it's bandwagony and ridiculous. That said, the simple truth is surprisingly easy: IF AI improves human lives...wonderful! IF AI replaces human jobs (which it has already) then...it's not something anyone should really embrace. Please keep AI out of my games.

This. This right here.

I've had it with AI, and soulless, AI generated garbage. Not to mention the fact that AI is being used by every kind of scammer imaginable to fleece people of their cash, while hurting legitimate businesses. Don't even get me started on the moral and ethical implications of its use, art theft, copyright infringement, and the murky waters of human rights.

While we have implemented algorithmic tools and different types of generators for many years, AI as it stands now is a whole different beast from the type of AI that say, for example, dictates the behavior of your companions in game. At least the people who wrote, scripted, and created the companions were real people with real thoughts and a job to do. AI has none of that. (And no, sorry, I don't care what people say- AI isn't any more sentient than my keyboard is. It's a program, not a person.)

And in all honesty: if you don't want living, breathing, people to write, test, and give input on your game, do you really care at all? Why should I care about your game if it wasn't worth actual human effort to produce?

Keep AI out of my favorite game, thank you very much. I have more respect for the creators of this game than that, and I don't want to see them replaced by AI.

AI could create dialogues on spot, related to ES lore and the quest player is following at the time. Even reproduce the NPC's voice.

You do realize voice actors need their jobs, right? That it is completely unethical to steal someones voice through AI reproduction and sell it without the consent of the VA? The very thought that anyone could be okay with robbing a VA of their living by stealing their voice is disturbing, to say the least.

FluffyBird · March 9

AI is not a "FUTURE MUST HAVE BUY NOW" tool, but it's also not "ew, gross" by itself. It's a tool. Like a hammer. You can hit a nail with it, or a priceless piece of art, or someone's head.
And imo the #1 problem that must be solved is how to ensure that this tool is used for helping humans instead of replacing them.
Using AI for picking companions reactions and smarter combat moves? Yes, please! (though, I suspect that it doesn't need whole AI for that)
Using AI to create companions quests and skills and dialogues and steal someone's voice to boot? Now that's gross.

colossalvoids · March 9

Not sure if I was reading a topic or a forum-add lmao. Glad that there were knowledgeable responses already, as those topics are mostly a buzz than something substantial to begin with.

PrincessOfThieves · March 9

danno8 wrote: »

The pessimistic (or perhaps realistic) side of me tells me that LLMs (I dislike the term AI for nearly all that the term is currently being used for) will be used to primarily drive down costs for companies by replacing as many employees as possible with inferior LLM systems while hoping most people don't notice or care enough to notice.

This. I don't understand people's optimism about this when gaming (and not just gaming) companies are already using everything they can to cut corners.
Sure, AI is theoretically capable of cool things, but in the world we live in it will just be used to cut costs and displace workers. The way AI systems are trained proves it, they are build on the data that is collected and used without people's consent and sometimes even used to actively harm their livelihoods (especially in case of artists, voice actors etc). In my opinion, cool auto-generated npc dialogue isn't worth it when it means that the actors whose voices were used to train the AI will be out of work.
Yes, AI itself is a tool. But so far there are already examples of this tool being developed and used in hideously unethical ways, and I honestly have no reasons to think that it will get better.

danno8 · March 10

ArchangelIsraphel wrote: »

And in all honesty: if you don't want living, breathing, people to write, test, and give input on your game, do you really care at all? Why should I care about your game if it wasn't worth actual human effort to produce?

Keep AI out of my favorite game, thank you very much. I have more respect for the creators of this game than that, and I don't want to see them replaced by AI.

Part of the fun of playing video games is the idea that an actual person is trying to convey something to you, either a message or a meaning through the dialogue or action of an NPC. With "AI" it is simply spitting out some averaged result of an algorithm based on your input in the hopes that mediocre response is something that is actually relevant to the situation.

I personally wouldn't want to play a game where "AI" was the creator of all the dialogue or story, except perhaps to try and break the algorithm for kicks.

The only thing that "AI" could be good for is for things that simply take too long or require too much resources for humans to accomplish. Like having different sized character models and getting equipment to fit on the models properly for example. "AI" could potentially find patterns that it could use to stretch and fit equipment well enough to look good on different sized models. In these cases "averaged" outputs are usually adequate.

ZOS_Icy · March 10

Greetings,

We have removed some insulting back and forth that was disruptive. Please ensure you are treating others with respect on the forums even when they have views that differ from your own. With that being said, while we understand that you may have interest in other topics, we ask that threads remain focused on ESO. As we want to keep the discussion on the forums focused on this game, we've closed this thread.

Thank you for your understanding.