Kingdom Come: Deliverance II Creator Praises The Witcher 4, Calls for AI Integration

Daniel Vávra, the esteemed creator of the Kingdom Come: Deliverance series, is closely following the development of The Witcher 4. He firmly believes that developers could gain significant advantages by more actively utilizing artificial intelligence.

The new material from 'The Witcher' looks incredible! I am impressed! Now imagine… What if the much-maligned AI didn't take jobs from programmers, but instead allowed many more people to create games like The Witcher 4? What if, thanks to AI, CD Projekt could make a game many times larger and more complex with the same number of staff?

According to Vávra, the intelligent application of AI has the potential to dramatically enhance gaming experiences.

A game that is translated and fully voiced in all languages immediately after release, featuring a much more realistic world, significantly more quests, and NPCs with far more actions and animations, because the old, tedious motion capture will be a thing of the past. Instead of sluggish, weak games with AI, we will get more games like Crimson Desert, GTAVI, or W4, but without waiting 10 years for each. And all of this without crunch and without developers destroying themselves in the process (I know what I'm talking about).

Vávra added that the game industry has undergone considerable changes: budgets are increasing, and projects are becoming more complex. Consequently, developers are actively seeking new tools to realize their ambitious ideas. He did not rule out the possibility that Warhorse Studios might themselves employ AI in future projects.

And all of this for a fraction of the budget of modern AAA games. Because that's inevitably what will happen. In a world where anyone can create an AA-level game, the advantage will go to those who can make large games with previously unthinkable quality. Does that still sound scary? Not to me. I have several ideas that were previously impossible to implement. Games bigger and more ambitious than KCD2. And riskier in terms of commercial success. Now, it's possible.