Project Orion Wishlist

[Albatar999]

- Polished game
- Polished game
- Polished game
- Better perks synergy (for example quickhacks bonus from queue but quickhacks nearly upload instantly, doesn't make sense)
- NPCs with agenda (at least the basics, having a work, a hobby while not working, sleeping, not NPCs that stay in the same place for 679 days as they were peaces of furniture)
- More random events in the open world (that brings so much life)
- Better AI (AI is exploding everywhere but not in video games, why ?)
- More interactions with random NPC (at least some basic chatting, that also brings more life to the open world, no way to add some AI stuff in there ?)
- Rewarding exploration (exploration is interesting if there are some stuffs to discover that worth it, no random trinkets in chests please)
- End game contents that allow us to play our finished builds, what the point of going to 60 if there is nothing to do anymore ?
- More interactions with the open world, what the point of building an open world if it looks like a dead world ?

 

[Tarille]

- Better AI (AI is exploding everywhere but not in video games, why ?)

A number of reasons.

Primarily, it's the battle between developers, writers, artists, voice actors wanting to still have work versus the corporate greed in the industry that would drop literally every worker in lieu of AI the second they could.

Beyond that, there's the limitations on what AI can do.

AI like ChatGPT utilizes a central server and fills its database based on what gets input into it and some cursory information gained from certain sources (Like wikipedia)

AI like the stuff that makes AI art or Google Lens, runs a process on a device that searches for results based on set parameters and amalgamates that into a result.

Neither of which are condusive to the "AI" in video game NPC's, which don't have any databases to scour for relevant information about how to act. Nor is it particularly desirable for a video game to either run online required processes and/or upkeep a centralized server to connect to purely for the sake of utilizing AI (There's massive pushback against "Always Online" requirements for single player games)

We haven't gotten to the point of AI advancement that we can utilize actual "Intelligence" for NPC's (Current AI isn't actually "Intelligence" it's merely just a glorified search engine that can output its results in a variety of media such as creating images or mimicing voices). Further still, we haven't gotten to the point where average CPU's would be able to process hundreds or even thousands of individual "Intelligences" that would be active at any time during a video game (If we were to truly make all NPC's function like people, including doing their own thing when the player isn't interacting with them)

NPC behaviour (A more accurate term than "AI") is still going to be bound to specifically programmed parameters that the characters follow for the forseable future. The best we can hope for is that these parameters can gain more depth and complexity over time as hardware improves and thus processing power increases (Of course, it'll take a while given how much longer each console generation is lasting these days)

 

[Albatar999]

A number of reasons.

Primarily, it's the battle between developers, writers, artists, voice actors wanting to still have work versus the corporate greed in the industry that would drop literally every worker in lieu of AI the second they could.

Beyond that, there's the limitations on what AI can do.

AI like ChatGPT utilizes a central server and fills its database based on what gets input into it and some cursory information gained from certain sources (Like wikipedia)

AI like the stuff that makes AI art or Google Lens, runs a process on a device that searches for results based on set parameters and amalgamates that into a result.

Neither of which are condusive to the "AI" in video game NPC's, which don't have any databases to scour for relevant information about how to act. Nor is it particularly desirable for a video game to either run online required processes and/or upkeep a centralized server to connect to purely for the sake of utilizing AI (There's massive pushback against "Always Online" requirements for single player games)

We haven't gotten to the point of AI advancement that we can utilize actual "Intelligence" for NPC's (Current AI isn't actually "Intelligence" it's merely just a glorified search engine that can output its results in a variety of media such as creating images or mimicing voices). Further still, we haven't gotten to the point where average CPU's would be able to process hundreds or even thousands of individual "Intelligences" that would be active at any time during a video game (If we were to truly make all NPC's function like people, including doing their own thing when the player isn't interacting with them)

NPC behaviour (A more accurate term than "AI") is still going to be bound to specifically programmed parameters that the characters follow for the forseable future. The best we can hope for is that these parameters can gain more depth and complexity over time as hardware improves and thus processing power increases (Of course, it'll take a while given how much longer each console generation is lasting these days)

Was talking about an already trained AI that could use some sort of text files database to formulate dialogues based on the NPC profil (like barman, worker, walker...).

There are some AI which are way more lighter than ChatGPT, Llama for example and can work on local, no need to be connected to Internet.

It's more about random NPC than story NPC, actors, writers and such would be preservated but delegate the random stuff to the AI, instead of having the same useless sentences for all random NPC : "Hum..., Hi, Not now...".

I'm aware that "AI" is more a marketing argument than anything else, but still, it could be used for video games to fill the wholes without having to invest much time in it.

Could also use a trained AI for hostiles NPC the same way chess AI works for example.

I mean, video game "AI" still the same as 1990, really time to work on that.

It's not about replacing everything and everyone by "AI", but use it as a tool in a smart way.

 

[Tarille]

Was talking about an already trained AI that could use some sort of text files database to formulate dialogues based on the NPC profil (like barman, worker, walker...).

Well... That's literally just the normal parameter creation.

Which is restricted by developer time. Like, sure they could create parameters and dialogues for every single NPC type... But that takes resources.

Even if you say to use AI to write the dialogues and voice them (Replacing human workers who would otherwise have been paid to do these things) someone still needs to check over them to make sure they are usable (AI is not particularly accurate...) and then someone needs to actually put the files into the game. Which again, is resources.

Could also use a trained AI for hostiles NPC the same way chess AI works for example.

Chess AI works by analysing possible board states and moves up to a certain point (With the more powerful AI's like the most updated Stockfish analysing further ahead). Given the vast amount of possibilities of actions available to hostile NPC's it's going to be hugely resource intensive to do such a thing.

It often also takes time for this analysis to occur (Lightweight chessbots take a few seconds. The most powerful ones take several minutes. For each move) which doesn't work for a fast paced FPS game.

I mean, video game "AI" still the same as 1990, really time to work on that.

Sure... But technology hasn't advanced to the point where it's feasible to do anything beyond that and big tech companies just wasted the last few years (And billions of dollars) on Metaverses and NFT's instead of anything remotely useful...

All we can hope for right now is more polishing of the existing system and hoping that developers get the resources needed to better flesh out NPC actions, at least until there's a breakthrough with actual AI or more refinement of the current "AI tools" (To allow for dynamically generated dialogue/actions without the possibility of complete jank being produced)

 

[tavellone]

Well, there is one, released no long ago, named Starfield, which put on player no real pressure to pursuit the main quest and let to player the freedom to do other things instead. But it turn out that the story itself, isn't really good (as expected^^).
Beside, I think it don't bother most of players. BG3 which have the same kind of "urgency" than Cyberpunk wouldn't be such a succes if it does.
So I don't see any reason for writters to avoid it anyway. It might not please everyone (like everything), but it work if well written

Well that's a bad example, Starfield's story isn't bad because it doesn't artificially force the story along, it's bad because it's written by a shitty writer. Maybe it doesn't bother a lot of people, but it does bother a significant amount of people at the least and besides, the complaint is more for the sake of the game anyway, having a game with a false sense of urgency from the start will just subconsciously force people to rush through the main story, because the game is literally telling you to do that. It's doing a disservice to the rest of the game and it's stories.

 

[Albatar999]

Well... That's literally just the normal parameter creation.

Which is restricted by developer time. Like, sure they could create parameters and dialogues for every single NPC type... But that takes resources.

Even if you say to use AI to write the dialogues and voice them (Replacing human workers who would otherwise have been paid to do these things) someone still needs to check over them to make sure they are usable (AI is not particularly accurate...) and then someone needs to actually put the files into the game. Which again, is resources.

Chess AI works by analysing possible board states and moves up to a certain point (With the more powerful AI's like the most updated Stockfish analysing further ahead). Given the vast amount of possibilities of actions available to hostile NPC's it's going to be hugely resource intensive to do such a thing.

It often also takes time for this analysis to occur (Lightweight chessbots take a few seconds. The most powerful ones take several minutes. For each move) which doesn't work for a fast paced FPS game.

Sure... But technology hasn't advanced to the point where it's feasible to do anything beyond that and big tech companies just wasted the last few years (And billions of dollars) on Metaverses and NFT's instead of anything remotely useful...

All we can hope for right now is more polishing of the existing system and hoping that developers get the resources needed to better flesh out NPC actions, at least until there's a breakthrough with actual AI or more refinement of the current "AI tools" (To allow for dynamically generated dialogue/actions without the possibility of complete jank being produced)

The existing systems won't be polished, they haven't done so in the past 20 years, gaming companies are in a mind set that lead them toward delivering a mediocre product rather than a polished one, because it cost less money to do so with nearly the same financial results (customers can't consume smartly enough to force them to adapt, even if it's mediocre, they will buy the product, that's why we have so many unfinished/unpolished games), that's why i'm pointing AI as a solution that could automaticaly complete what would have taken months in a cheaper way and way less time, of course AI is still in an early state, but it will sky rocket in the next years and as we all know, CP 2 won't be released in the next months.

If no one start working on such tools to improve games' quality and depth, it will never happen.

I've heard rumors about games that will be released in the next months/years which will use AI, it might be a major improvement if that happens and game companies which haven't anticipated will be god damn far behind.

But they are just rumors.

 

[vinekuto]

I would love to see some small details that promote the game itself like red dead redemption horse balls or geralt beard does. Cyberpunk could implement for example sliding on different surface with different outcome like running up. It could be also implemented as a small animation of sliding inside playground tube. Small details like this are just a topping on delicious dessert.

 

[alittleLowkey]

Honestly I'd like to see a game with a more grand narrative similar to RDR2. Could be something like:

• You are part of a crew that is trying to make a name for itself.
• Early game will be focused on establishing yourselves and would have some urgency to it.
• Midgame would be a sandbox, you meet NPC.s, potential love interests and do missions both with and without your crew.
• Late game, shit goes south and things get real urgent. Side missions more or less disappear (maybe some romance specific ones gets added).
• Endgame would be similar to CP2077.s act 3. 100% urgency.

Other than that I'd like some of the following:

• 3rd person view (or at least cutscenes)
• If a NPC can have some specific clothes, so can you.
• Customizable body types. If I want to be big and heavy or short and skinny or whatever I can be
• More body tattoos.
• More customization of piercings and tattoos. Why can't I pick nose, ears, eyebrows etc individually?
• To be able to repaint cars.
• Maybe add some form of eat and sleep mechanic so that apartments and food matters.
• More love interests. If they tie love interests to side missions as with Kerry and River they could have BG3 numbers of them without problem.

 

[Haiken09]

What the mod "Night CIty Alive". I would want that officially implemented on the future Cyberpunk games.

 

[_KejN_]

I do not worry about story, dialogs, quests and gameplay, because I know that CDPR do awesome job in those areas. I'm only worrying about that they will abandon V, his friends and the Night City. They did amazing story, character and locations. It would be such a waste if they get rid of it. Also major part of players, including me, will be sad. They left many open doors for future (even return of Johny Silverheand is possible in my opinin) and unfinished story. I have high hopes that they will choose The Sun or The Star as a canon ending and we will have a chance to finish V story and meet friends and Night City again (it does not mean that there will be no new characters and locations).

Hey CDPR, if you are reading this I keep my fingers crossed for that solution

Also I would like to see some pools in this thread if possible, It would be nice to know your opinions.

 

[ExSteam]

For very hard to be challenging, actually live up to its name.
More in game performances by Lizzy Wizzy or anyone else.
More great music, so many bangers.
Easter eggs for the Shadowrun IP
Mr Hands
More Songbird
More Rogue - maybe a story line of what happened after Johnny died.
The patience to wait.
Fingers crossed....some more minor content to hold us over.....and patience.

 

[Seifer97]

please CDprojekt don't give up the FPS view... there are already a lot of TPS games... without this FP view cyberpunk 2077 would not be what it is now… a masterpiece

 

[eggahmanos]

The main thing night city is missing is multiplayer. I want drop in drop out multiplayer. The ability to play the story totally in single payer, multiplayer or both.

 

[mannygt]

- no enemy level scaling
- give back old stats and slots on clothes

 

[Schachmatt222]

Wishes for the next project:

• A polished game at release. Take your time.
• A rich story, branched missions.
• Great characters, as immersive as in CP77
• Believable gameworld
• Better fleshed out romance system
• And most important: NO multiplayer. The last thing CP77 needs is a MP on top of it's many problems, and Orion should be a deep SP experience like your other games rather than a half baked SP with a mediocre online component.

 

[alittleLowkey]

It's funny that all of these three posts in a row are big nopes for me.

please CDprojekt don't give up the FPS view... there are already a lot of TPS games... without this FP view cyberpunk 2077 would not be what it is now… a masterpiece

Can keep first person in gameplay, I don't care overly much (prefer to be abie to toggle) but skip it in cut-scenes. I want to see my character.

The main thing night city is missing is multiplayer. I want drop in drop out multiplayer. The ability to play the story totally in single payer, multiplayer or both.

Multiplayer rarely adds anything to the single player mode and there are enough multiplayer games out there that go unsupported and unloved as it is (looking at you Red dead online).

- no enemy level scaling
- give back old stats and slots on clothes

No. Just no. The idea that a jacket gives me plus on critical hit chance is absurd and i really don't want that.

Level scaling is more two edged sword, the way it worked in for example Skyrim were bandits all of a sudden had armor and weapons that could take out armies was ridiculous. But in Cyberpunk it worked since it was just the stats of weapons and cyberware that went up, aesthetically it was no difference. So keep level scaling, makes it funnier to explore the world.

 

[Guest 4412420]

Don't sideline male romance options again on practically every front like what they did to Kerry and River.

I have no gameplay related wishes, I rarely mess around with builds so I was pretty content with what 2077 had.

Maybe more varied dialogue options to offer better roleplaying, photomode poses that aren't memes.

 

[mannygt]

It's funny that all of these three posts in a row are big nopes for me.


Can keep first person in gameplay, I don't care overly much (prefer to be abie to toggle) but skip it in cut-scenes. I want to see my character.

Multiplayer rarely adds anything to the single player mode and there are enough multiplayer games out there that go unsupported and unloved as it is (looking at you Red dead online).

No. Just no. The idea that a jacket gives me plus on critical hit chance is absurd and i really don't want that.

Level scaling is more two edged sword, the way it worked in for example Skyrim were bandits all of a sudden had armor and weapons that could take out armies was ridiculous. But in Cyberpunk it worked since it was just the stats of weapons and cyberware that went up, aesthetically it was no difference. So keep level scaling, makes it funnier to explore the world.

I disagree. A jacket can give me armor stats boost or elemental protection boost. Shoes can give me low detection boost. Glasses and visor boost headshoot hits or hacking damage. And so on.
Level scaling is the worse thing you can make in a videogame. It just make your character dull without being powerful at high level. I always liked the perception of entering "dangerous" area when you're low level (example: Pacifica or some areas of Heywood when you face Valentinos).

 

[Tarille]

A jacket can give me armor stats boost or elemental protection boost. Shoes can give me low detection boost. Glasses and visor boost headshoot hits or hacking damage. And so on.

There's the divide between CAN you implement these things and SHOULD you implement these things.

It can make sense for clothing items to provide statistical benefits.

At the same time it makes less sense if you're wanting to use clothing items to look nice.

Such a thing is alleviated with the wardrobe system allowing one to create a custom look that overrides what is actually equipped.

But the other thing is are these little stat boosts even necessary for the game? When you get stats from attributes, stats from skills, stats from cyberware... Do we even need all these stats? Does the game need to have this "Looter Shooter" style RNG items dropping all over the place?

I'd argue no. I'd argue that the game would be overall better if all the build elements came purely from weapons, cyberware and attributes. With weapons and cyberware being a modular system of you put in the mods that give the effects/bonuses you want as opposed to the current RNG fiesta of Cyberware upgrades.

Level scaling is the worse thing you can make in a videogame. It just make your character dull without being powerful at high level.

Level scaling serves a purpose. Which is that it allows you to experience content in your own desired order. This can be very good for promoting player freedom.

The downside is that it can make for a lack of "End game" style areas and lacks some feeling of progression if you can go back to where you started the game and still get your butt kicked despite being high level.

That said, there's a compromise that can often get the benefits of both level scaling and static levels, which is to utilize both. Create different regions which each scale within a certain level range. Meaning you can do content within a region in your own preferred order, but you move between the regions based on the static levels. So you have an actual "End-game zone" full of the hardest enemies that will slaughter you if you go there at level 1.

However, there's also the argument of whether even this makes sense. Like, if Pacifica is the toughest area in the game full of the most dangerous enemies... Why haven't the VDB/Animals simply taken over the city? If all their members are just stronger than literally everyone else in the city...

I suppose one of the more logical ways to do things is have each region divided into static levels. So everywhere has pleb enemies you can deal with at low level, but also has areas filled with the toughest and meanest of the gangs/corpos which are for your high level builds to tackle. While having scaling for dynamic events like the car chases or having gigs where enemies call for reinforcements so your higher level more notorious character causes the stronger enemies to come after you.

 

[ShinAkira00]

So many of these suggestions are a bit farfetched in that they're just not feasible for this kind of game.

Adding stats to clothes - I understand why people want this but this is a game with cyberarms that effectively act in the same way. It would be redundant to have stats on clothes as well as cyberware and that's most likely why they changed it. The cyberware for your legs essentially act as boot armor etc. They made the right choice just having the clothes as cosmetics.

AI - This is such a broad term that people just kind of throw around because it's thrown around in the media as some all-knowing and all-encompassing thing. Firstly, technically AI is already used in the game in the broadest sense of the term. However
using an AI model to generate in game dialogue would be like trying to generate an image using an AI tool. If you've ever used any of those then you'll notice a few things. You have to input data to prompt the model what to return and what to not return. So someone on the development team will also need to set parameters for this in game, at which point they may as well just write the dialogue themselves. It's not something that is automatic and like a lot of those image-based AI tools, a lot of times it will just generate dialogue that is garbage or respond in a manner that is not what you want. Now unless CDPR has Google level resources and processing power to do this near perfectly in real time then it's not going to happen. That's a pipe dream and not practical for a video game.

The only thing I would want them to do is build on the first game. Cyberpunk 2077 was their first foray into this genre and into a shooter style game. Its basically the Witcher 1 of this franchise and like Witcher 1 there's a lot they need to improve on because the game is still a little rough around the edges. They should focus on refining it, not making elaborate leaps.

Post automatically merged: Feb 15, 2024

Don't sideline male romance options again on practically every front like what they did to Kerry and River.

I have no gameplay related wishes, I rarely mess around with builds so I was pretty content with what 2077 had.

Maybe more varied dialogue options to offer better roleplaying, photomode poses that aren't memes.

From a purely financial perspective that's just not going to happen.
1. They need the game to be released in certain countries to maximize profits and prioritizing male romances is antithesis to that objective. Adding to that, the target audience for the game are largely those with female romance interests.
2. To their credit they tried to what you wanted in Cyberpunk 2077 and in my opinion, all the romance options suffered because of it. When you try to appease everyone and have limited resources, something suffers in quality. So what ended up happening was players were left with half-baked relationships all round. Panam and Judy felt like distant friends at a certain point in the story and they were your only option so if you didn't like either of them then that's it. In the Witcher 3 players had more options in romance and more fleshed out ones but in Cyberpunk 2077 they had 1 option that wasn't very well developed.

Who knows, maybe CDPR will do a better job on the next one and you'll get what you want or they'll try gain and the game will suffer for it. I guess we'll see.