The 2025 Video Game News Thread (I was forced by anti-innovation communists to change the title again)

This still seems to me as the sort of thing that gets completely lost in translation. Not just “the original article was in Japanese”, but also in “this is a technical topic being talked about by a Chief Executive”.

A lot of code can be understood to be “machine produced” in a way, via template expansion. This is completely normal and has been happening for 30+ years easily. I don’t see it as much of a stretch for this to be misunderstood by an executive and/or any people in the middle of the game of telephone between the developers that self-reported on “how many kilobytes of your final source did you actually type?” and the CEO making that statement.

A disturbing amount of code certainly is being produced by AI tooling right now, and I am strongly opinionated that this is industry-wide self-own of potentially disastrous proportions, but I’ll eat my hat[1] if it is 90%.


  1. I almost never wear hats. It’s coming up on cold weather here though, maybe I’ll offer to eat a beanie. I guess that would be equally unpleasant. ↩︎

8 Likes

yes, programming as a profession is trying to replace itself because it believes itself irreplacable. it’s very funny, they’re like: the actual code is secondary, we can understand requirements and work toward solutions. the requirement is making money, and it’s hard to compete against lobor costs that are measured in kW/h… maybe just let the ai do what it’ll do, design and implementation and everything.

what a boring, stupid, shitty future we’re working on.

8 Likes

Seems kind of not far off from saying “90% of code is created by Ctrl+C and Ctrl+V.” It might be sort of accurate to say that but it’s that remaining 10% that absolutely must be done by people who know what they’re doing, or the total result is going to be a disaster.

You know, you hear a lot about AI doing “coding,” but has an AI actually created an entire program?

2 Likes

As a human programmer myself, 80-90 percent just seems completely absurd. My company pays for some of these tools and I’ll use them periodically, and they’re just not there.

I feel like every couple of years there’s some new tool people are pushing as the solution to the problem of high labor costs in software development and, surprise surprise, we’re still doing things the old fashioned way for everything that actually matters. Before it was AI it was the “low code revolution.”

The hope for all these business types is that they can stop paying senior software developers massive salaries, but the reality is that these things are nowhere near being able to handle the kind of tasks that actually make up the day to day work of a software developer. They completely fall apart whenever you get off the beaten path.

I think large language models are pretty cool technology, but I think the way they’ve been deployed is wildly irresponsible. The unrealistic expectations from managers are one thing, but the social consequences are absolutely terrifying. It breaks my heart seeing people use this crap for “companionship.”

9 Likes

It’s pretty easy to invent a metric where through a combination of rebranding any automation as “AI”, double-counting, copy-paste, boundary shifting, and just generally creating a weird definition for a quanta of ‘code’.

eg, using an engine where every individual item placed in the gameworld could potentially have 30+ handler functions for different events, all created as blank stubs by default, and all but a tiny handful . you could easily claim that over 99% of a games code is generated by machine (and therefore by “AI”)

This sort of talk is pretty common at the management level.

At it’s heart it’s the same scam as most SaaS, or Shadow-IT nonsense that has been around for decades. It’s popular because if you play your cards right as an IT consultancy you can make money on the initial buy-in and cleaning up the mess as your clients GTFO.

One of the reasons I generally liked Ed Zitron’s writing about AI is that he made this connection quite frequently. Recently he seems to less willing to stray into the weeds in this way, which I find less effective.

3 Likes

Not that it entirely invalidates your comment as Hino probably doesn’t have his own hands in their recent games’ code (or maybe he does and that explains some issues), but Hino started his career in video games as a programmer and was, by all accounts, a pretty good one back in the PS1/PS2 era. So I do not think it would be fair to insinuate (or give him the excuse) that he has no idea what he is talking about. Something having gotten lost in transcription or translation is a more charitable explanation.

3 Likes

It’s been more than I year since I checked on Project Caesar, but Paradox officially announced it as Europa Universalis V… with the tagline Be Ambitious

Johan Andersson put together Paradox Tinto studio in Barcelona specifically to take over and build EUV from the ground up, and with this announcement it looks like all hands are on it. EUIV was basically a money printer for Paradox, 12 good years with big DLCs every year or two. When Tinto took over the roadmap in 2020, and their first release out of the gate, Emperor, was really well received but then they delivered some buggy sub-par feature-poor DLCs during COVID and lost a lot of the good will. I don’t think I’ve picked up EUIV since 2021-2022 personally, but I understand how impossible of a task it is to support such a gigantic simulation for so long while also being beholden to a DLC model that demands transformative but stand-alone changes. I think they were severely limited in the kind of ideas they could implement and bogged down by tons of technical debt that starting fresh is exciting. The game never really ran well in the late game, and to be honest it became a micromanagement mess in a lot of ways.

But starting over seems so daunting. They have to serve a lot of different audiences, like people who will absolutely tear apart every mechanic to find exploits or people who want a deep realistic or alt-history simulation experience. It needs to have enough depth for people who will spend thousands of hours, but also not scare away casuals who might only spend 500 hours or less. It needs to have hundreds of achievements and enough historical flavor to make every micronation in the holy roman empire feel somewhat unique. But Paradox did it recently with Crusader Kings 3, and I thought they did a great job with the launch of Victoria 3. EUIV will not have a long honeymoon and a bad first impression like Cities Skylines 2 could possibly be catastrophic for Paradox, so I can imagine they’re feeling a lot of pressure to get this right.

If EUIV was for strategy freaks, EUV’s foundation promises to be even more freakish. They pushed the start date back a hundred years to 1337 and end date forward to 1837. 500 years of gameplay is going to be like 100 hour campaigns where the world gets radically transformed. They’re replacing a lot of the board-gamey mechanics with sim mechanics. Most notably pops are replacing the dice rolls and governance points. You have to feed people and build a robust political economy to advance. I thought the estates system in EUIV was a good way to constantly have tradeoffs and manage particularist rebellions but it did feel a little tacked-on 6 years into the game. Now, it seems very fundamental to nation building. We gotta keep the polish peasants pacified.

They are also promising new ways to play ‘tall’ rather than just have the game centered around endless conquest. This makes me very happy as someone who never really bothered with stuff like trade leagues or found naval empires more trouble than they’re worth. I’m kind of fond of the convoluted way that trade worked in EUIV but the output was always just money. Here they are promising that trade has some kind of supply line aspect to it, which could get super overwhelming, or be amazing. Check out this long list of tradegoods.

All this before we even get into changes in warfare, colonization, tech trees, government types, religion, culture, diplomacy, characters, and a lot more. It’s just going to be absolutely gigantic. Be Ambitious indeed.

7 Likes

I have hundreds of hours in Europa Universalist IV and never really understood trade. Tech, conquest, diplomacy, colonialism, I understand all those really well. I even know how to protect trade routes or steer trade in nodes. I just have no idea what is optimal. I’m guessing.

If EUV can maintain that quality - of having complexity that one can play well with without fully understanding - I think I’ll enjoy it. They’ve done that with Crusader Kings III. Victoria III is close but a mite fiddly (trade is more important and I feel like I have to understand it to do well); Stellaris is close but a mite simplistic (the economy is more abstracted and easier for me to manage).

3 Likes

yeah this seems insane to me. The promised automation/delegation had better work as they say

2 Likes

Sure, why not.

9 Likes

all the private equity bullshit and in the end the site is worth so little to them that they’ll sell it to several guys ?

10 Likes

It’s possible that the investors only just realized that their company named Giant Bomb isn’t an arms manufacturer.

11 Likes

For the idiots among us, what are pops?

7 Likes

It’s a Kellogg product described as a “crunchy, sweetened, popped-up corn cereal”.

5 Likes

Good for them, but it has a very “We were fine with everything Fandom changed up until they told us to stop swearing” energy.

At least the wiki is presumably safe from Fandom.

Now they can live like kings. Damn hell ass kings.

1 Like

Oh sorry it is short for part of population! Basically every human alive in 1337 A.D. is counted toward a province, which is a modifier that affects the manpower, production income, taxes, and trade value for the state. EU IV had about 4941 provinces, and this map is going to be even bigger. Everything from Japanese Diamyos to the Incan Empire are represented.

Each population within a province has cultures, religions, and social class that dictates what resources it needs to consume in order to have its needs met. If they see a government’s actions actions and ideologies as illegitimate, they become radicalized and start rebellions. But if their needs are met, they live longer, have more children and it attracts migrants so the game simulates the growth of your nation. It means more workers, more prosperity, bigger military, more conquests. Or you might want to invest in technologies and defend yourself, because this was a turbulent time, and every nation-state has neighbors who are also militarizing and going through social upheaval during the renaissance, enlightenment, revolutions, and industrialization.

6 Likes

Oooh I see, thanks! So is this perceived as a way of simplifying the experience with less gameplay “currencies” to handle overall? Or does it make the simulation more complex because even more relies on pops?

1 Like

developer behind forum favorite Loretta is back with some sort of nightmarish strategy city builder of all things

I am absolutely cackling like the Onion Sickos cartoon. I have hundreds into EU4 and I am going to be an absolute monster for this

1 Like