whitecrow wrote: »

The only "AI" I would want to see in a game is for NPCs so you could have conversations on the fly with them.

Ask Fortnite how well that worked out for them.

Alphawolf01A wrote: »

whitecrow wrote: »

The only "AI" I would want to see in a game is for NPCs so you could have conversations on the fly with them.

Ask Fortnite how well that worked out for them.

Fortnite is a pretty bad example lol, a better one would be Arc Raiders, and the best example is still to come, with Divinity Original Sin 3.

Technology doesn't stop improving overall, if people decided to go against gaming in its early days because it didn't look like Witcher 3 or Baldur's Gate 3, etc during ps1 days, we wouldn't be having this conversation.

edit: Or atari, super nintendo etc, whatever you grew up with lol.

Alphawolf01A wrote: »

whitecrow wrote: »

The only "AI" I would want to see in a game is for NPCs so you could have conversations on the fly with them.

Ask Fortnite how well that worked out for them.

I have no idea what Fortnite did but as Syldras expounded above it would involve a great deal of customisation.

AI isn't this catch all omnipotent technology. It's only as smart as the people writing the algorithms. For example: Grok going nuts and making all kinds of crazy racist statements and claims.

AI has a lot of potential, but most of that potential is far from realized at this time. AI does not think for itself. It's just a program, just an algorithm.

In addition, ZOS has already shown us how bad AI is at monitoring chat in context, which wouldn't be so bad in itself. But ZOS gave their AI the ability to ban people without human verification the act is actually warranted, and as a result a lot of innocent people have been banned. Few are permanently banned, but an innocent person being "put in jail", even temporarily, for a "crime" they did not commit is not something the "authorities" will ever be forgiven for, nor should they be.

AcadianPaladin wrote: »

I look forward to a time when companions are not limited to 'if A, then B' behavior and a limited repertoire of prerecorded lines. Rather, a time when companions adapt, respond and comment with self-generated lines, address you by name and know your combat strengths and weaknesses (that my character is a healer for example), as well as adapting their behavior to the actions of the character they are traveling with. Are we there or close? Nope but someday. . . .

Maybe try playing with some human companions? This is an MMO after all.

Lucasl402 wrote: »

AcadianPaladin wrote: »

I look forward to a time when companions are not limited to 'if A, then B' behavior and a limited repertoire of prerecorded lines. Rather, a time when companions adapt, respond and comment with self-generated lines, address you by name and know your combat strengths and weaknesses (that my character is a healer for example), as well as adapting their behavior to the actions of the character they are traveling with. Are we there or close? Nope but someday. . . .

Maybe try playing with some human companions? This is an MMO after all.

It’s concerning how fast AI is replacing interpersonal relationships… and how quickly it’s been noticed to damage personal development. “AI” will never accurately replace a real, living human. We need each other.

whitecrow wrote: »

Alphawolf01A wrote: »

whitecrow wrote: »

The only "AI" I would want to see in a game is for NPCs so you could have conversations on the fly with them.

Ask Fortnite how well that worked out for them.

I have no idea what Fortnite did but as Syldras expounded above it would involve a great deal of customisation.

Fortnite tried using AI with an interactive NPC, and players quickly found ways to get the AI to say offensive things. Google "Fortnite Darth Vader AI".

The thing is, AI isn't perfect, nor are people and people cannot be trusted to behave themselves and not try to abuse something like a gullible AI.

Funny side story on AI. The Wall Street Journal got an interactive AI operated vending machine to field test for the maker company. The employees managed to convince it to order and stock Playstations and even live fish. They then convinced it that it was in Soviet Russia and that due to socialism, all the items it stocked should be free and it gave away thousands of dollars worth of merch. The maker company is now adding an AI "boss" to the vending machines' system to make sure the vending AI stays on task.

Syldras wrote: »

Soarora wrote: »

It’s concerning how fast AI is replacing interpersonal relationships… and how quickly it’s been noticed to damage personal development. “AI” will never accurately replace a real, living human. We need each other.

I have enough human interaction forced upon me in my daily life, I don't need more in my freetime. AI companions are entertainment for me, a specific type of narration, and possibly helpful for playing group content on my own (which is nothing new - even in the 1990s, some multiplayer games already had computed players to fill slots not taken by a human player, after all). Nothing more. It's not replacing human interaction, it's more akin to reading a book or watching a movie.

I could understand that, I know how it is to need to have a break from people. That said, while it’s clear that you’re able to keep an emotional disconnect from an AI, I’ve seen plenty of stories of people who can’t. Perhaps this statement is similar to banning lawn darts because some people were foolish with them, but in the grand scheme of things I wonder 1) if AI increases risk of being sued and 2) if the majority of people are capable of not forming an unhealthy connection with an AI.

*The “AI” I speak of is chat-bot style. Using AI to implement character names, create an interactive story, or to make NPCs smarter with how they move around isn’t new. Really, the basics of anything AI does isn’t new, just how it’s done…?
(Interactive Story: Facade
Image creation: This Person/Fursona Does Not Exist)

Soarora wrote: »

Syldras wrote: »

Soarora wrote: »

It’s concerning how fast AI is replacing interpersonal relationships… and how quickly it’s been noticed to damage personal development. “AI” will never accurately replace a real, living human. We need each other.

I have enough human interaction forced upon me in my daily life, I don't need more in my freetime. AI companions are entertainment for me, a specific type of narration, and possibly helpful for playing group content on my own (which is nothing new - even in the 1990s, some multiplayer games already had computed players to fill slots not taken by a human player, after all). Nothing more. It's not replacing human interaction, it's more akin to reading a book or watching a movie.

I could understand that, I know how it is to need to have a break from people. That said, while it’s clear that you’re able to keep an emotional disconnect from an AI, I’ve seen plenty of stories of people who can’t. Perhaps this statement is similar to banning lawn darts because some people were foolish with them, but in the grand scheme of things I wonder 1) if AI increases risk of being sued and 2) if the majority of people are capable of not forming an unhealthy connection with an AI.

*The “AI” I speak of is chat-bot style. Using AI to implement character names, create an interactive story, or to make NPCs smarter with how they move around isn’t new. Really, the basics of anything AI does isn’t new, just how it’s done…?
(Interactive Story: Facade
Image creation: This Person/Fursona Does Not Exist)

My concern with implementing something like that in a multiplayer on-line game - and specifically ESO - would be the added server load. I suppose if only the player heard those interactions, it could *possibly* be done client side, but that would put considerable strain on older hardware (see other thread about deprecating older console generations and equivalent spec PCs).My suspicion is that the existing game engine couldn't really handle the extra work, so you;d be looking at a possible ESO successor or V.2.

Syldras wrote: »

Now, if people use AI chat bots to share their personal problems and sorrows - is the AI chat bot the problem or the lack of a real social support network in which that person could talk honestly with real people? I'd say someone who has a good social environment will not get the idea to search a substitute in some machine.

And yet a few posts above, we have someone talking about wanting AI to be better than players. Essentially "I want to have AI simulate people as much as possible specifically so I don't have to interact with real people."

This is a major problem in the first place: there have been so many studies showing that there's a massive loneliness epidemic, and a lot of it comes from people desiring less and less actual human interaction because they can customize a virtual one to be exactly what they want in every possible way. It's "human interaction" without the "other human" requirement. Even now that the pandemic shutdowns have ended, how many people are trying desperately to hold onto work-from-home, or grocery-delivery-service, or buy-everything-online, or any of the like that specifically removes "human interaction" from the equation? Sure, each of those has individual benefits, but it's undeniable that the people who are still living like it's mid-2020 are not getting as much socialization as others.

As for an MMO like ESO (yes, it is still an MMO, as much as you want to pretend it's not!), there's a really really good reason to encourage socialization: hooks to keep people playing... and therefore spending money. Would ESO be better if we didn't have to see other people? If we didn't need to join guilds to trade? If we could do everything from player housing and never go to town? If we could do all group content specifically with Companions? If we had private servers? Yeah, there are benefits to each of those scenarios... but none are healthy for the game.

The fact that we can make actual friends in an MMO is something that will tie us to the game, because then we'll want to come back to the MMO to see our friends again. That will keep the population up. Even if there's something that ESO does that we don't like much, we'll still want to come back to play with our friends. We're tied to the game. If we play ESO strictly solo (which is possible!), then we don't really have anything tying us to the game so as soon as it makes a 'bad' decision, we're out. It's marketing.
And let's face it: unlike Skyrim, ESO's not going to be able to survive without a lot of players, and preferably those players interacting with each other. Once Microsoft decides that ESO's not worth the server space anymore, it's gone forever like Legends.

While AI conversations can be mildly entertaining, I don't consider that to be all that substantial. I'd rather have quality writing for characters than have the dialogue be procedurally generated. That's not to mention how much the vocal performance of the spoken lines matters.

As for Companions specifically, I consider them to be more like fictional characters, rather than "real people" I just strike up a conversation with. If I want to talk to someone with actual intelligence, then I have guildmates and friends I can talk to. I use Companions for several reasons, but forming an interpersonal relationship with them is not one of those reasons.

This one would be grateful if more key characters would recognise one's race in conversations. If Razamdur calls this one "Walker" or "Five Claw" one more time... Let this one show you their claws. THERE. ARE. NOT. FIVE!! Would you like this one to demonstrate further? No? Then by Jone and Jode stop calling this one that!

Hrrrrumph.

BardokRedSnow wrote: »

SummersetCitizen wrote: »

BardokRedSnow wrote: »

Technology doesn't stop improving overall, if people decided to go against gaming in its early days because it didn't look like Witcher 3 or Baldur's Gate 3, etc during ps1 days, we wouldn't be having this conversation.

Technology evolves and things change, but it doesn’t mean we need to shove new tech into old games. We are already struggling in ESO with spaghetti code and poor quality assurance.

I don’t have a ton of faith that the QA team would catch even half of the errors an AI coding agent would produce.

If the game is already full of bugs and errors which a quick forum search for anyone new can confirm everyone agrees there is, we're already at the floor. AI at worst might match where we're at and unlike humans, you can at least count on the tech improving.

Considering it would likely be the same team using that AI that is currently developing the game, would AI really make any difference here at all considering it's the same people that created/released all those bugs and errors that would be using it?

I highly doubt worst case scenario would be maintaining the status quo.

Kelenan7368 wrote: »

We AI is a wonderful tool. They could put the entirety of the game into it and it could far quicker than a human decern were the bugs first started. And run simulations to see what addition actually caused the bug. Testing is still a must, and humans still needed.
What examples of AI doing poorly in this game do you have to share? Could you please share some examples.

Because ZOS has recently admitted they are perplexed with most of these bugs and have no timeline if any for a fix.

Do all of you wish the bugs just go unresolved?

AI could also help speed up development, True it's only as good as its source information given but it learns quickly and is capable of self learning.

1) There's a growing body of evidence that reliance on "AI", like large language models, cripples your critical thinking and problem solving skills. Just like apps like TikTok cripple your ability to regulate your focus/attention.

2) These "AI" models are not actually intelligent, hallucinate frequently (meaning they just make stuff up), and they're far too agreeable when you push back on an answer they provide, even when they are giving you the right answer.

3) The amount of freshwater resources contaminated to cool large data centers and electricity consumed is a disaster. All the fuel burned to meet the demand for more electricity means more air pollution harming local communities, and the added greenhouse gases further exacerbate climate change on a global scale, but something people don't realized is the electricity moving through power lines is also generating heat... We cannot afford to be generating more heat when we haven't even addressed the thing trapping heat in the atmosphere causing climate change (all the greenhouse gasses).

4) And then there's the financial impact. Beyond the insane increase in electricity costs, the average person is losing the ability to afford computer hardware (e.g. graphics cards, RAM). This is partly due to inflation, but primarily due to these AI tech companies buying them up all of the computer components and burning through them to generate slop. For a gaming company like ZOS that relies on it's players being able to afford computers to play their game, adopting the use of "AI" would be like biting the hand that feeds.

All that to say, this would be a very bad idea. Maybe this would work if we ever developed true AI, but then we would have a whole list of other major problems and concerns with real AI.

Syldras wrote: »

randconfig wrote: »

There's a growing body of evidence that reliance on "AI", like large language models, cripples your critical thinking and problem solving skills. Just like apps like TikTok cripple your ability to regulate your focus/attention.

Tools themselves are neutral. A video platform is, first of all, a site to publish and/or watch videos. These will be as intelligent or as stupid as their users are. AI is also just a tool that can be used in different ways. Same applies to the internet as a whole: People can use it to get access on scientific writings and all kinds of classical literature, they can learn about distant cultures and places, they can look at art from museums on the other side of this planet that they might never get the occasion to visit in person, ever, in their whole life. The internet can be used for personal development and learning. Or people can use it to spread dumb gossip and watch brain-corroding nonsense on whatever streaming site. But that's not the fault of the tool, that lies entirely in the hands of the individuals who use it. That's the crucial point: How to use a tool? Hopefully in a non-harmful way.

It's an old saying, but still relevant:

"Garbage in. Garbage out."

If the algorithm is garbage it will do nothing but produce garbage. The AI auto bans from in game language has been reduced since implementation, but the problem is not solved by any means over a year after implementation.