• We are currently performing site maintenance, parts of civfanatics are currently offline, but will come back online in the coming days. For more updates please see here.

Player stats, sales, and reception speculation thread

Some pretty smart people are saying that AI will be writing the majority of code by next year. It seems plausible that Firaxis and other developers will be retooling their staff to push gaming forward quickly. Which might make it child's play to rework Civ 7 to be ageless and to add another era. Or two.
From my experience of coding with LLMs, this is probably true. The majority of code will probably be written by AI by next year. However, the majority of code written by AI doesn't work. And the way to get the best results from it is to target it at specific questions or problems, get it to do a first draft for you then edit from there. It can reduce time a competent coder takes to get to good working code, but it is not an independent thinking and problem solving replacement for a coder. You might need less coders as a consequence, but it isn't a fundamentally different way to produce a game - it's an efficiency.

I think like others say, main impact we'll see is in indies and modding, as it reduces the order of magnitudes between potential output of a small team and a AAA developer. We could see more of a competent and quality cottage industry emerge outside of the main publishers.
 
If the hype about AI is true, then all of our thinking is moot.

Some pretty smart people are saying that AI will be writing the majority of code by next year. It seems plausible that Firaxis and other developers will be retooling their staff to push gaming forward quickly. Which might make it child's play to rework Civ 7 to be ageless and to add another era. Or two.

On the other hand, AI might just create new games that instantly make all that has come before irrelevant. If the transition happens then the calendar itself may change to 2025, 2026, AGI 0, AGI 1, AGI 2...

If the transition doesn't happen, well, a lot of very rich people have spent enormous sums to make it happen. So, we should see some pretty entertaining consequences either way.

Which basically means we probably ought not worry so much about one iteration of one computer game.
In my experience, using AI to write boilerplate is one of the few practical use cases. Re-writing, re-factoring, or adding to an existing codebase of meaningful size is much more haphazard. In those cases, if it's a matter of reducing resources spent, you're likely going to instead spend more cleaning up and testing AI-made changes than doing it the old fashioned way.
 
In my experience, using AI to write boilerplate is one of the few practical use cases. Re-writing, re-factoring, or adding to an existing codebase of meaningful size is much more haphazard. In those cases, if it's a matter of reducing resources spent, you're likely going to instead spend more cleaning up and testing AI-made changes than doing it the old fashioned way.
And even then it has to be boilerplate with clear documentation, that matches the versions of the software that you're using, and so on.

It has very limited actual application, in my experience. It's usable for prototyping, it's helpful for things that are flat-out obtuse to write (like regex), but the moment you have a complex, internal codebase that it can't be trained on, the utility dips massively. The same goes for older versions of architecture which lack modern documentation (or the documentation available is only for later versions).

I think anyone who thinks a game engine can be written with it, knowing what little I know about the complexities of engine programming specifically, should really not be confident about positive outcomes. Be curious, sure. But neutral at best is what I'd recommend (and to be clear: I'm negative on it).

And that's before we get into the IP side of things. It's true indie studios could use this to accelerate output, but indie studios will also not have the armies of lawyers that the big tech proponents like Microsoft, etc, have to handle any disputes that might arrive from "accidents" in the training datasets.
 
  • Like
Reactions: j51
Some of the stuff that ChatGPT or similar can do is pretty great, but I sincerely doubt it will come close to refactoring a true AAA game. Maybe there's a level of using it to come up with some game balance, or certainly it could potentially be useful for creating more narrative events.

AI generated code might be the "majority" of code when you count probably the 5000 platformer/tower defense games, or like those terrible games you see all the time in youtube ads, where it can just shove random boxes on the ground, give you a random gun, and have you shoot random guys that come. But it's not going to play any major part in any major studio anytime soon.
 
  • Like
Reactions: j51
...
 

Attachments

  • Screenshot_20250606_123646_Facebook.jpg
    Screenshot_20250606_123646_Facebook.jpg
    89.3 KB · Views: 55
Could we maybe just have 4 Ages like in Civ 3? A Medieval age between Antiquity and Exploration that focuses on establishing religion and reinvigorating scientific thought.

Maybe we could also quickly just pump out a few more civs and leaders so that every antiquity age civ (and every current leader) has a clear and logical path of progression. Keep the mix/matching if they want to.
 
What you guys are saying sounds more likely than the predictions of instant transformation. But again, a lot of smart, rich people are adamant that transformation is right around the corner. So, we shall see.

For me, I see the transformation on the battlefields in Ukraine which makes me believe that we are already in another world. Autonomous drones are surely coming.
 
What you guys are saying sounds more likely than the predictions of instant transformation. But again, a lot of smart, rich people are adamant that transformation is right around the corner. So, we shall see.

For me, I see the transformation on the battlefields in Ukraine which makes me believe that we are already in another world. Autonomous drones are surely coming.
Dont know if autonymous drones have anything to do with AI really. Its simple to measure distance. I am in a land surveying business myself. I think as simple is to measure heat signals if done right with adequate equipment. Its quite really simple to trigger mechanical force of a weapon to scanning information beforehand. It doesnt really need any more of an thought process to execute the command.
 
In this thread: Harumph; Civ VII is yet another installment where the AI opponents can't play the game well enough to provide a meaningful challenge. :sad:
Also in this thread: AI is on the verge of taking over the world!:run:
 
Last edited:
What you guys are saying sounds more likely than the predictions of instant transformation. But again, a lot of smart, rich people are adamant that transformation is right around the corner. So, we shall see.

For me, I see the transformation on the battlefields in Ukraine which makes me believe that we are already in another world. Autonomous drones are surely coming.
A lot of smart, rich people are invested in convincing everyone that transformation is around the corner. Of course it's what they're saying. They've also said we'd never have to drive again, and that in-person work was going away due to the Metaverse, and that the store of the future involved no checkouts and simply walking out with a bag of groceries.
 
I think some of the claims about what is being marketed as AI are greatly exaggerated, but there's no denying the LLMs have advanced to an impressive degree. As a developer, I am not really too worried that they will replace professional coders, not from what I've seen so far anyway. For professional level software, you still need someone who actually understands the code. Prod the LLMs a bit, and it is revealed that while they can produce some very convincing, and often very useful output, there isn't really any real comprehension or intelligence behind it. So they will not be the ones understanding the code. Nor will the "prompt engineer" who is getting AI to produce things they couldn't have written themselves.

That is not to say that these things are not useful, far from it. I have started using them myself, and this is going to transform parts of how I work, especially when I'm working on something I am less proficient with. Instead of spending a day or more reading through library documentation, studying unfamiliar syntax, searching the web for examples of something close to what I am trying to do, I can just ask the chatbot to give me an example of that thing, and it will generally be able to get quite close. I can ask it to tweak the example to specific requirements. I can ask it to explain parts of the example to me. I will ultimately write and test the actual code myself, but by using the LLM I was able to obtain knowledge I needed to do that much faster than I would have otherwise. I see it as an very powerful tool, a great leap forward in how I can search for and process information. It doesn't replace me, it helps me do my job better and more efficiently. It's similar perhaps, to a carpenter getting access to a nail gun instead of a hammer. Houses will go up quicker, but you still need the carpenter.

For non-professional use, I think the LLMs will make coding accessible to more people. Just like I, a man with very modest artistic abilities, can now ask a chatbot to "make me a picture of Darth Vader, but he's a duck" and get it in seconds, a non-coder or novice coder may be able to have a chatbot give them the means to make a simple program they want. I think that's pretty cool.
 
I've just seen Ursaryan has a video up on YouTube with his top 5 recommendations what what needs to change in the game - looking beyond bugs and UI specifically to gameplay. I think he highlights specialists, treasure fleets, percentage yields modifiers, legacy path & victory conditions, and city states and suzerain bonuses.

I found it really interesting because for me, and based on reactions to the game so far from many others, all of this is lipstick on a pig and doesn't begin to shift the dial on whether I might even think about buying the game, let alone playing it. It's all just more of what we've seen in patches having no effect on long term sustained play time.

Best case scenario, I think whatever firaxis do without a fundamental overhaul of ages and Civ switching at this point caps their audience at a max of who has already bought it coming back to play more regularly. I could see them very successfully running a relaunch where they introduce mechanics that cater to the core audience they've alienated grabbing them a bunch of extra revenue.
 
Personally, I think the biggest issue with the game is the Exploration age. Its mechanics suck and are mostly contrivances. It is sadly also the age that lasts the longest.

So in that sense, UrsaRyan is absolutely correct: Specialists are flavourless, Missionary Spam is boring and braindead, Treasure Fleets are 100% luck dependent. All three tie into Exploration.

Exploration Religion is a stain on this game, and should be treated the same way Ideology is treated in Modern - as a diplomatic (not cultural) mechanic that helps bring players together or further apart, so that Exploration diplomacy isn't a pure extention of Antiquity relations. Ideology itself could be improved (by for instance making the Ideologies themselves have friendships or rivalries, rather than a strict system where everyone hates anyone on a different side - Democracy and Communism were able to work together to tackle Fascism, remember.)

Replace the Relic system with Fine Art, which is made by Great Artists that you attract by entering Celebrations. Or something like that.

Suzerain bonuses that give you Free techs upon becoming the Suze (another thing Ursa correctly points out as 'not balanced') are so potent you rarely initate endeavours early into the Age, even if you get a bonus towards Endeavours. In my current game, I've rolled a Layafette + Greece start, and used my unique Endeavour twice. Twice. Across all three ages. Suzing City states, improving Trade Relations and acceping Research Collaborations from the scientific leaders have all been way more important than just +1 social policy slot (which the AI also gets if they accept, so never ever offer this one to Himiko).

Aside from Exploration I'd say the worst mechanic in the game are the Crises which add nothing positive to the game. No stakes, just annoyance. Praise the heavens you can switch those off.

I WILL say though, Antiquity and Modern are both good, and very fun, having played through the game a few times now. The victory conditions aren't fully balanced - Cultural and Economic are much easier than Militaristic, which is significantly easier than Scientific. It needs a bit of balance tweaking, but that's fine. The AI is absolutely able to win Cultural or Economic if you aren't careful.

The Civs themselves are also excellently designed, and each of them plays very differently. In my current game I went Greece => Ming => Mughal and each of these was completely different, which made the game interesting. Going from the brainy Ming into the greedy but stupid Mughals was a surprisingly pleasant whiplash, which forced me to adapt myself to Mughal's economic strengths, and use the gold to inflate the Culture bonuses I inherited from Great Walls, Acropoli and Layafette's Attribute tree. No, it isn't a straight line. But you do have to *think* about your plan, and the game rewards you for making good strategic decisions. Which is how it should be.

The game itself feels easier but also more fun, and large part of that is the low cost that units have. Pumping out units and maintaining them is way easier in Civ7 than it's been in any other Civ game, and that keeps it dynamic. You can always invade enemies, and defend yourself by buying units in a pinch. So can the enemy. War has stakes, and the classic steamroll like you'd have in Civ5 and Civ6 simply doesn't happen in Civ7. Usually, at least one AI will be able to snowball into a huge tech lead up to the max of their Settlement limit (by my observations an OP leader like Tubman or Himiko, or whoever is lucky enough to start as the Maya) and become a threat to win.

I dunno. The game has changes, and sometimes those changes feel like net neutrals. The game itself though? Absolutely playable, and good. Underrated if anything.

It's the Heroes 4 of Civ.
 
There was some talk about content creators and why they haven't been posting Civ7 content. A few days ago, Potato McWhiskey released a video saying he wasn't doing well mentally and needed a break from the negativity and toxicity he's been receiving from the community.

I guess that explains the recent absence of one content creator.
 
Yeah I watched that too, I feel bad for the guy. He said the toxicity of his fans has been getting to him for over a year now and it’s been really tough. Wishing him all the best and hope he can change up his fans somehow.

I wish Marbozir would go back to making old civ game content, I know he doesn’t like the new game and that’s fine, but I miss his civ content and he was the main YouTuber I watched for civ.
 
It was a hard video to watch. It didn’t apply only to Civ VII, and he’s in a very dark place. He struggles with depression and thoughts of self harm for his entire life. I don’t think it’s fair to talk about his mental health struggles in the context of Civ and the reaction to the game. It sounds like some of his longtime fans have been negative recently in his comments, but his mental health is a longer struggle.

I’ll try to find a transcript of the video later, but he actually mentions Civ very little.

Wishing him all the best.
 
He talks about Civ, his videos, and the community pretty incontrovertibly at around minute 11. It goes on for at least 5 minutes.
 
Oh I am sure a elf wasn’t implying that the civ fans and reaction and his fan community (much less the broader civ community) was somehow the cause of his mental health issues, that would be terrible. Just explaining why they were missing these days.

Edit: ah ok sorry I misunderstood
 
Last edited:
Back
Top Bottom