How far are we from AI recreating EQ99?

[DeathsSilkyMist]

Originally Posted by BradZax [You must be logged in to view images. Log in or Register.]

My point would be that you don’t disagree with me and yet here you are seven pages later still arguing with me.

Let's go back to OP's question:

Originally Posted by Dulu [You must be logged in to view images. Log in or Register.]

https://www.youtube.com/watch?v=MWl2qQRyutc

Watched this video, where the guy basically recreates Doom in an hour.

Keep in mind, Doom is a single-player game, and came out in 1993.

But... the video is 3 months old and was probably recorded some time before that.

So the question is, how far are we from making a game like P99 in AI?

The question OP asked is: How far away are we from AI being a significant developement partner when making a complex game like Everquest? The context comes from OP's amazement at how far the youtuber got in one hour using AI.

The answer is we are years away from AI being a significant development partner in the creation of complex video games like Everquest. AI will be coding assistants watched over by experienced programmers at best in the next few years, unless we get a new breakthrough in AI.

AI will not be able to level up newbies to the point where they can make a game at the quality of Everquest in 100 hours.

Originally Posted by BradZax [You must be logged in to view images. Log in or Register.]

You asked where are the indi mmos made with YouTube tutorials, I gave them to you.

To be honest, those four games look terrible. Three of them are not released yet, so we don't know if they will collapse before they release, or collapse soon after release.

The one game that is released has 1 review. These are not good examples.

[Kich867]

Originally Posted by BradZax [You must be logged in to view images. Log in or Register.]

My point would be that you don’t disagree with me and yet here you are seven pages later still arguing with me.

I'll be real dawg, its because you're a layman and you aren't being specific enough with what you're talking about, so we have to draw inferences on what we think you're trying to ask, to which you repeatedly just said "that's not what I mean" to an extent that I can barely wrap my head around what you're even doing here.

You made up your mind zero seconds in and are engaging with social media at this point just to talk to people. Which is fine sort of, just don't get all haughty over it.

If your point was "can someone with no programming experience talk to AI a bunch to learn how to do this?" I'd say probably, yeah. The debatable part beyond that would be scrutinizing the project and dealing with the hard parts of game development, especially multi-player persistent world.

It would be really challenging, particularly due to the lack of idempotency--which basically means it doesn't produce a consistent result. If you ask it A, it does not always respond with B. Because fundamentally that's just not really how they work, right.

And so the biggest hurdles will be anytime you need to revisit things, or append things, or expand on a thing you've already built, it's not necessarily going to "know" that and give you back what it thinks you want and how that not really being specifically what you're trying to do, and then trying to figure out how to get it to do what you actually wanted.

Honestly what I would expect to see happen, what I think is far more likely, is that AI would get someone like that far enough through the project to have that person begin to understand what it is they need to actually do and then rely less and less on the AI as they just simply aren't that helpful.

But yeah, at the end of the day, building an MMO isn't a secret, it's a pretty well known thing how to do it's just a huge pain in the ass. So like if you asked AI about it it should be able to tell you these concepts and what is generally considered best practice (although even that is suspect, remember, AI can't think, it isn't intelligent, it doesn't really understand things, so it's just going to say whatever it sees enough people saying is a best practice).

[BradZax]

Originally Posted by Kich867 [You must be logged in to view images. Log in or Register.]

You are engaging with social media at this point just to talk to people.

This is an internet forum.

Originally Posted by Kich867 [You must be logged in to view images. Log in or Register.]

But yeah, at the end of the day, building an MMO isn't a secret, it's a pretty well known thing how to do it's just a huge pain in the ass. So like if you asked AI about it it should be able to tell you these concepts and what is generally considered best practice.

Yep! And it will hold your hand as you accomplish each step with zero prior experience!

So now, it's possible now.

Honestly what I would expect to see happen, what I think is far more likely, is that AI would get someone like that far enough through the project to have that person begin to understand what it is they need to actually do.

Yes!

[DeathsSilkyMist]

Originally Posted by BradZax [You must be logged in to view images. Log in or Register.]

This is an internet forum.

Yep! And it will hold your hand as you accomplish each step with zero prior experience!

So now, it's possible now.

Something being possible does not mean it is probable.

Is it possible for monkeys banging on typewriters to generate Shakespeare? Yes. Is it probable? No.

The games examples you provided all look terrible, and are a farcry from the quality and scope of Everquest.

If your point is simply that AI can generate slop that nobody wants to play, then I agree with you.

[loramin]

It's possible because Brad declares it to be so, based on no programming background and (poor) inferences from a few YouTube videos.

Nevermind that reality disagrees ... Brad has declared it to be true!

Do I have the substance of your argument more or less correct?

[BradZax]

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

Something being possible does not mean it is probable.

Is it possible for monkeys banging on typewriters to generate Shakespeare? Yes. Is it probable? No.

Yeah but these like, philisophical arguments though compelling, are not quite based on actual contextual data of today.

Like I could say, its possible the world is flat because I haven't seen the world, and that would not be probable because of the amount of other people who have seen or navigated the world.

Just like learning how to make 1990s video games there are many examples of indi projects and developers doing it successfully, countless hobby projects, open source unity versions of everquest, downloadable VR mods, emulation servers etc, that it would indicate that what I am saying, IS more probable than not.

And the bet isn't: "what are the odds that someone does this in 3 years?" it's, "how long until someone CAN do it."

The answer is right now.

[DeathsSilkyMist]

Originally Posted by BradZax [You must be logged in to view images. Log in or Register.]

Yeah but these like, philisophical arguments though compelling, are not quite based on actual contextual data of today.

Like I could say, its possible the world is flat because I haven't seen the world, and that would not be probable because of the amount of other people who have seen or navigated the world.

Just like learning how to make 1990s video games there are many examples of indi projects and developers doing it successfully, countless hobby projects, open source unity versions of everquest, downloadable VR mods, emulation servers etc, that it would indicate that what I am saying, IS more probable than not.

And the bet isn't: "what are the odds that someone does this in 3 years?" it's, "how long until someone CAN do it."

We keep going in circles. Let me ask you directly.

If a single newbie developer were to create Everquest from scratch using the current AI tools, what percentage of the work would be done by AI, and what percentage of the work would be done by the developer?

Is it 80% AI and 20% 1 newbie developer?
Is it 50% AI and 50% 1 newbie developer?
Some other percentage?

[BradZax]

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

We keep going in circles. Let me ask you directly.

If a single newbie developer were to create Everquest from scratch using the current AI tools, what percentage of the work would be done by AI, and what percentage of the work would be done by the developer?

Is it 80% AI and 20% 1 newbie developer?
Is it 50% AI and 50% 1 newbie developer?
Some other percentage?

That's a tough question to answer because it's like, "If AI taught you how to write and mail a letter, how much of the work is done by AI and how much of it is done by you?"

Idk, that is a weird question. So i asked chat gpt your question and it said:

The honest answer is: AI can probably do 40–70% of the raw production output, but you (the developer) will still do 60–30% of the total work, because the “total work” is mostly decision-making, iteration, debugging, and taste.

For further clarification I asked, "of the work I have to do, how much of it can you teach me how to do, step by step?"

It said:

All of it.

[DeathsSilkyMist]

Originally Posted by BradZax [You must be logged in to view images. Log in or Register.]

According to chat GPT (because there is no way I would give you an accurate answer to such a strange question, because like we're talking, "if AI taught you how to write and mail a letter, how much of the work is done by AI and how much of it is done by you?"

Idk, that is a weird question. But i asked chat gpt and it said:

Chat GPT is wrong. There is no scenario in which current AI tools could even do 30% of the work on an Everquest-style MMO. If you did that, the game would be buggy and easy to cheat in. It wouldn't be a game people would want to play. This is not going to change in the next few years, unless we get a new breakthrough in AI.

[BradZax]

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

Chat GPT is wrong. There is no scenario in which current AI tools could even do 30% of the work on an Everquest-style MMO. If you did that, the game would be buggy and easy to cheat in. It wouldn't he a game people would want to play. This is not going to change in the next few years, unless we get a new breakthrough in AI.

I don't think you can make that claim.