Why Former Dragon Age Writers Say AI Could Make Gamedev Frustrating

Former Dragon Age writers are sounding the alarm, arguing that AI could make gamedev frustrating as hell by eroding the narrative soul of RPGs. While tools like Gemini 2.0 and Claude 3.5 offer speed, they often lack the nuance required for complex branching dialogue. For the average player, this means a shift toward procedurally generated filler rather than handcrafted storytelling. This isn’t just about jobs; it’s about the quality of the games we play on our consoles and PCs in 2026.

The Narrative Problem with Generative Models

I have spent hundreds of hours in games like Dragon Age: Inquisition and the newer titles running on Unreal Engine 5. The writing is what keeps you hooked. When you use a model like GPT-4o to generate quest lines, you get grammatically correct text that feels hollow. It lacks the ‘human touch’—that specific, weird, or funny dialogue choice that makes a character feel real. Developers are now paying $20 to $50 per month for enterprise-grade AI seats, hoping to cut costs. However, when you replace a $70,000-a-year writer with a prompt engineer, the output often results in repetitive, fetch-quest-heavy content that bores players after ten hours. The current state of AI writing feels like a draft that never quite hits the final polish required for a AAA release.

Contextual Limitations of LLMs

LLMs struggle with long-term narrative consistency. In a game like Baldur’s Gate 3, choices you make in Act 1 matter in Act 3. Current AI models often hallucinate or lose track of ‘world state’ variables. Even with a 2M token window, they fail to maintain the specific voice of a character consistently across a 60-hour campaign, leading to jarring shifts in tone that ruin immersion for players.

Economic Pressures and the AI Shortcut

Publishers like EA and Ubisoft are under immense pressure to keep margins high while budgets for AAA games hit $300 million or more. AI is being pitched as a way to trim these costs by 20% to 30%. But we are seeing the results of this shortcut in recent releases. Assets look great, but the world feels empty. When you pay $70 for a title, you expect a cohesive experience. If developers rely on AI to write side quests, the game feels like an endless loop of generic tasks. I’ve tested several ‘AI-assisted’ indie titles this year, and the writing is consistently the weak point. It’s functional but forgettable, which is a death sentence for a narrative-driven RPG.

The Cost of Human Talent

Hiring a veteran writer costs more than an API subscription, but that investment pays off in player retention. When a game has a cult following, it’s usually because of the writing. Cutting that cost to save a few hundred thousand dollars on a $200 million budget is short-sighted and ultimately hurts the brand’s long-term value.

The Technical Gap: Workflow Integration

Integrating AI into a professional workflow isn’t as simple as clicking a button. Developers are finding that cleaning up AI-generated ‘slop’ takes longer than writing the content from scratch. If you have to spend three hours editing an AI’s output to make it sound like a sentient being, you haven’t saved time. Tools like Unity Muse or Roblox’s generative suite are cool, but they aren’t ready for complex narrative work. They are great for generating textures or simple background NPCs, but not for the protagonist’s emotional arc. The frustration comes when management forces these tools onto creative teams who know exactly what needs to be done but are hindered by software that isn’t quite there yet.

Workflow Bottlenecks

The bottleneck isn’t just the AI; it’s the integration. Moving AI-generated data into a complex engine like Unreal Engine 5 requires manual oversight. Fixing broken scripts or mismatched character animations caused by AI errors is a massive time sink that frustrates veteran devs who just want to ship a clean build.

Consumer Impact: What You Should Expect

For you, the player, 2026 is going to be a year of mixed quality. You will see games with massive, beautiful worlds that feel strangely vacant. You will encounter NPCs that have infinite dialogue options but nothing interesting to say. My advice? Be wary of marketing hype that touts ‘AI-powered storytelling.’ Look for reviews that specifically mention character depth and quest variety. If a game is marketed primarily on its use of AI, it’s usually a red flag. Stick to studios that prioritize human writers. We are seeing a divide between ‘soulless’ content factories and studios that use AI only for mundane tasks like texture tiling or procedural foliage generation.

Identifying AI-Generated Content

Look for repetitive phrasings and ‘flat’ emotional delivery in quests. If an NPC sounds like they are reading a list of facts rather than interacting with you, the writing is likely AI-assisted. Trust your gut; if the story feels boring, it usually is.

⭐ Pro Tips

• Use Claude 3.5 Sonnet if you must use AI for brainstorming, as it currently handles nuance better than GPT-4o for creative tasks.
• Save money by waiting for the ‘Game of the Year’ edition, which usually hits a 40% discount within 6 months, avoiding the launch-day bugs of AI-rushed titles.
• Don’t rely on AI to generate your entire game logic; use it for boilerplate code only, as it frequently introduces subtle, hard-to-debug logic errors.

Frequently Asked Questions

Final Thoughts

The consensus among industry vets is clear: AI is a tool, not a replacement for creativity. If developers continue to prioritize AI shortcuts over human writers, we will see a decline in the quality of the games we love. As a consumer, your wallet is your biggest voice. Support studios that value human talent over cheap, synthetic narratives. Stay tuned to my blog for more deep dives into the tech behind your favorite games.