Divinity
- Thread starter Izo
- Start date
Hekotat
FoH nuclear response team
- 12,936
- 13,115
- 27,859
- 43,518
Yeah I liked 1st one more as well. Dont get me wrong they were both good, but 1st one was better IMO.
- 2
Arbitrary
Tranny Chaser
- 32,876
- 95,854
Talos
Golden Baronet of the Realm
- 4,510
- 14,055
Pick up OS1 on Thursday. Should be heavily discounted. It's very much worth playing.
Both are exceptional RPGs. I would recommend starting 1. Not due to story, they are independent except some background/world info, but because the combat system is a bit more evolved (and QoL features) in 2. They play different than BG3 due to it being their own system so I personally haven't had the issue of going from BG3 to trying DoS2 again, but DoS2 to DoS1 feels more noticeable since it is the same system.
That said, they are excellent on tactician (but hard, especially in the beginning). Similar to how they build BG3, DoS1 and 2 have all hand crafted encounters, no random ones. So when you encounter something which seems too hard, you can explore and find other encounters to level up and go back later.
I found DoS2 easier to "exploit" than DoS1. In DoS2 it is all about knockdowns. Get a 10 points in a summoner 10 asap (massive boost to pet), and regardless of build get the points in warrior (forget the skill names) for the rush knockdown and the AE hit which knocks down. Focus on physical damage (easiest). You and enemies have a grey (armor/physical) and blue (magic) bar for your shield. Once that is depleted, you can chain knockdown enemies. Once all your characters have those skills, the overall game becomes much, much easier even on tactician. Downside of that is that by early Act2, most encounters are won the same way. I've tried respeccing and going magic or whatever, but anything other than that feels like I am just trying a personal challenge mode to make it more difficult for myself. There is this whole system around combining elements for damage (especially on the ground), but above method bypasses many encounters where it is obvious that you're supposed to implement/use that. Rangers are also great, buy and loot every single knockdown arrow you can find.
PS: they've patched it so skills transfer to the group. So things like lucky find procs on everyone as long as one person has it. I think when it launched you had to make sure to use the character that had the skill. You absolutely want a character with points into that, once you do, try to open every box,chest, whatever. If 15 crates are stacked on a dock, click them all. Odds are one will proc and you can end up with an item with better stats than a boss drop. It's broken (unless they patched that too).
- 27,498
- 37,024
Larian interview - Turn based, which seems like a no brainer. The AI thing I really don't care but apparently a lot of people do.
Larian is trying to find ways to cut down on development time and aims to finish Divinity in less time than Baldur’s Gate 3, which took six years to make because of its scale and Covid-19 disruptions.
Under Vincke, Larian has been pushing hard on generative AI, although the CEO says the technology hasn’t led to big gains in efficiency. He says there won’t be any AI-generated content in Divinity — “everything is human actors; we’re writing everything ourselves” — but the creators often use AI tools to explore ideas, flesh out PowerPoint presentations, develop concept art and write placeholder text.
Larian is trying to find ways to cut down on development time and aims to finish Divinity in less time than Baldur’s Gate 3, which took six years to make because of its scale and Covid-19 disruptions.
Under Vincke, Larian has been pushing hard on generative AI, although the CEO says the technology hasn’t led to big gains in efficiency. He says there won’t be any AI-generated content in Divinity — “everything is human actors; we’re writing everything ourselves” — but the creators often use AI tools to explore ideas, flesh out PowerPoint presentations, develop concept art and write placeholder text.
Last edited:
- 1
Cybsled
Naxxramas 1.0 Raider
- 18,166
- 14,697
Generative AI chat bots for random NPCs might work, like Where Winds Meet. You still have handcrafted stuff, but if players want to try to squeeze some conversations in with a character outside of that you can
There was a texting mod for cyberpunk that did that also - you could have conversations with Judy/Panam and the responses would mostly be in-character
There was a texting mod for cyberpunk that did that also - you could have conversations with Judy/Panam and the responses would mostly be in-character
- 1
Arden
Vyemm Raider
- 3,069
- 2,361
Yeah I'll be contrary here (since a lot of people automatically react negatively to AI) and say that after using a ton of GPT I think modern AI is going to be an absolute revolution for games, ESPECIALLY rpgs.
Yes, some lazy games will try to use AI to generate all of their content (and those games will suck).
But the sweet spot will be games that use humans to create all of the environments, storylines, NPCs, etc. But then imbue the NPCs with advanced AI capabilities and just let things go where they will.
Im talking all the NPCs here, not just side characters.
I mean you could completely kill the whole idea of conversation trees and pre-ordained responses from the player. The player could just say whatever they want to say and the NPC will react uniquely using NPC memory and advanced AI.
It's really easy to imagine a situation where you have games going in directions that the developers didn't even contemplate.
My guess we are in the final cycle of games without major AI contributions to the dev cycle. First will be "hidden", as in the devs try to keep it as under wraps as possible. Then it'll be of course it is. Indie devs (or studios that still own themselves to the degree of Larian) will be the only ones able not to do it. In the above article they mention that TenCent, while on the board of directors, only own a "significant minority" of the studio shares.
However, once AI really works without being that noticable for players, you think any company like TenCent or similar are going to allow pitches for games with budgets and development time that could be cut in half or more?
I would not be shocked at all if in the future you have companies with AI stables of people similar to any regular talent agency (like that Megan model or whatever her name was). AI databases of "characters". Then all you'd need would be to adjust game engines to easily slot the AI characters into them. Dev team "scenario writers" who writes out the commands press generate and the next day you have a few cutscenes, major dialogue and tons of in game scenes ready. No HR, salary/overtime, only the restrictions of contracts with the AI firm of "this is what you can do/can't do with our character(s)".
Cybsled
Naxxramas 1.0 Raider
- 18,166
- 14,697
I always think back to Everquest 1 - you could talk to NPCs, but some really only reacted to specific key words
Using an AI system, you could have more dynamic quests to provide some backend support. The quest will eventually get you to the same relative spot, but everyone will get there a bit differently depending on who they talk to and what questions they ask. That is a ton of work to do by hand, but an AI system could help automatically generate a lot of that to fill in the blanks
Using an AI system, you could have more dynamic quests to provide some backend support. The quest will eventually get you to the same relative spot, but everyone will get there a bit differently depending on who they talk to and what questions they ask. That is a ton of work to do by hand, but an AI system could help automatically generate a lot of that to fill in the blanks
goishen
Macho Ma'am
- 3,755
- 15,109
This, basically. Mortismal and Fextralife both have love for games that are melee heavy, and this one is no exception. Warfare is the easiest. Summoning is the second easiest.
Aerothurg? Meh, if you have ability points and wanna spend them.
I forget what the thieving ability is, but that. Yeh, sure, go 2 points deep, no deeper.
The archery ability? I forget, but only go like 4 or maybe 5 points into that, if you're gonna go that way.
There are tons of little things about the game that you just have to learn about playing the game.
Grabbit Allworth
Confirmed J6 Insurrectionist
- 2,009
- 8,218
I'm not going to bother typing out all my thoughts on AI because it would be a massive wall of paragraphs, but it's something I have spent a HUGE amount of time learning about and I'll summarize my thoughts.
Like it or not, AI becoming omnipresent is inevitable and the rate at which companies are iterating to release models that are orders of magnitude better than the last is already at a breakneck speed with no signs of slowing down.
In the short term, AI is going to cause a lot of people a lot pain as they're replaced, but no matter how loud they are, it's not going to matter. Fortunately, in the long term, humanity is going to greatly benefit (assuming we don't go down the Skynet timeline).
I have almost no empathy for game 'developers' because the VAST majority of them have absolutely no creative talent and/or input. Not to mention, a huge number of them are dangerhair Leftists pukes (fuck 'em) doing their damndest to inject their warped social activism into the games. The bulk of people at those 500+ employee studios are comprised of human drones that do nothing but use in-studio tools (that they really don't even understand) to implement the work of the tiny group of actually talented people. The majority of the jobs that the drones do can already be done by AI, but studios are still learning to utilize it in an efficient way. That, and social stigma/public outcry have slowed down adoption.
However, in 5 years, 75% of gaming 'devs' are either going to be producing AI slop on their own or putting the fries in the bag.
Like it or not, AI becoming omnipresent is inevitable and the rate at which companies are iterating to release models that are orders of magnitude better than the last is already at a breakneck speed with no signs of slowing down.
In the short term, AI is going to cause a lot of people a lot pain as they're replaced, but no matter how loud they are, it's not going to matter. Fortunately, in the long term, humanity is going to greatly benefit (assuming we don't go down the Skynet timeline).
I have almost no empathy for game 'developers' because the VAST majority of them have absolutely no creative talent and/or input. Not to mention, a huge number of them are dangerhair Leftists pukes (fuck 'em) doing their damndest to inject their warped social activism into the games. The bulk of people at those 500+ employee studios are comprised of human drones that do nothing but use in-studio tools (that they really don't even understand) to implement the work of the tiny group of actually talented people. The majority of the jobs that the drones do can already be done by AI, but studios are still learning to utilize it in an efficient way. That, and social stigma/public outcry have slowed down adoption.
However, in 5 years, 75% of gaming 'devs' are either going to be producing AI slop on their own or putting the fries in the bag.
Last edited:
Grabbit Allworth
Confirmed J6 Insurrectionist
- 2,009
- 8,218
That's a really bad strategy because the tools to determine its use are extremely reliable and companies that lie to their customers tend to get wrecked.
Furry
Email Loading Please Wait
- 26,630
- 38,541
Artistic lefties/furries HATE AI. And there's a fairly strong consensus in the fag-aligned community that supports them. Normies may not be able to tell when art is AI, but it can't even come close to fooling artists yet, so its dangerous to present it as a final product, though I can see why companies are tempted by the idea.