From Mighty Bear to SecretSauce Labs: Game Development Studio Pivots Entirely to AI Content Technology

After three years of navigating the complex intersection of game development and generative artificial intelligence, wearemighty is officially leaving the gaming industry behind to reinvent itself as SecretSauce Labs. The company, which operated as Mighty Bear Games before adopting the wearemighty moniker in January 2026, built its reputation crafting titles such as Disney Melee Mania, Mighty Action Heroes, and Butter Royale. Moving forward, the studio is shifting its focus entirely toward AI-driven content technology, aiming to help global brands produce high-volume media at scale while strictly maintaining visual consistency across advertising, social media platforms, and product portfolios.

The strategic pivot to SecretSauce Labs is not a sudden reaction to industry trends, but rather the result of a very specific operational hurdle encountered while the team was still deeply entrenched in game development. According to CEO and co-founder Simon Davis, the seeds of the new enterprise were sown in late 2022. At the time, Davis and co-founder Ben were working on a game project that demanded an extraordinary creative output: approximately 25 million distinct, completely on-brand playable avatars.

Attempting to generate that staggering volume of content manually was entirely unrealistic from a production standpoint. Naturally, the team turned to the early iterations of generative AI tools available at the time, hoping to automate the heavy lifting. Instead, they quickly ran into a brick wall of technical inconsistencies.

"We would ask for the same character twice and get a different face or outfit, or sometimes an extra finger," Davis recalls, highlighting the erratic behavior of early machine learning models.

That frustrating limitation sparked a three-year technical endeavor. Rather than abandoning the technology, the team began building the necessary infrastructure to bridge the gap between what generative AI could theoretically produce and what creative production pipelines actually required to function in a professional environment. For Davis, whose career spans more than two decades of building games across nearly every major genre—including puzzles, role-playing games, shooters, and racing titles—the stagnation of traditional creative pipelines made the pursuit feel urgent.

"The production process has barely changed," Davis says, noting that while the software used by creatives has steadily improved over the years, the fundamental workflow remains stubbornly outdated. "The pipeline is still the same slow process of send a brief, make the creative, review it, and then ship."

Davis’s fascination with artificial intelligence predates the current wave of generative tools by a decade, having experimented with machine learning as early as 2012. However, the immense avatar challenge transformed a long-held academic and technical interest into a pragmatic commercial problem. The core issue, he discovered, lay in the inherent architecture of early models.

"They were fast and creative, but they had no memory," Davis explains.

This missing cognitive component—the inability to retain contextual continuity across multiple prompts—created a massive chasm between raw technological capability and practical studio application.

The Problem with Generative AI

The limitations of early generative systems became glaringly apparent when the team began experimenting with architectures that allowed them to upload comprehensive style guides and existing brand assets. While these advanced tools could theoretically access the reference material, ingesting data did not guarantee that the resulting creative output would actually adhere to established brand parameters. For a studio striving to produce content at scale, introducing off-the-shelf AI simply multiplied the volume of unusable variations rather than solving the underlying output problem.

"We learned the hard way that looking good and being on-brand are two different problems, and off-the-shelf tools were only solving the first challenge," Davis says.

This distinction is particularly critical within the gaming sector, where creative assets must maintain absolute continuity across diverse touchpoints, ranging from high-end cinematic advertising and fast-paced social media campaigns to the interactive assets embedded within the game itself. Brand consistency, Davis emphasizes, is no longer about simple digital storage.

"Plenty of tools let you upload logos, past assets, and a style guide," he notes. "What’s mostly unsolved is checking every new generation against that material and catching a miss before a human ever sees it."

To address this, SecretSauce Labs developed a proprietary infrastructure that wraps around underlying generative models. This system is designed to evaluate incoming requests, determine precise creative requirements, and dynamically apply the exact elements of a brand’s historical material needed to ensure total fidelity.

For Davis, solving this specific vulnerability proved to be a far greater engineering hurdle than cutting production expenses. While high-quality AI video generation—complete with tight brand guidance and proper audio synchronization—remains relatively expensive, he views financial cost as a straightforward budgetary obstacle rather than a fundamental scientific barrier.

"Of the problems we set out to solve—scale, brand, and cost—brand drift was the hardest by some distance," Davis says. "Cost looks like the hard one on paper, because good AI video with tight brand guidance and proper sync still isn’t free, but ultimately cost is a budgeting problem. Brand drift is a research problem, and it took us three years of production work with creative experts from the world’s leading gaming companies to get right."

Creative Control

The promise of artificial intelligence as a force multiplier for smaller creative teams is one of the most widely debated topics in modern technology. However, Davis argues that the common narrative surrounding time savings is often misleading because it focuses exclusively on initial generation speed rather than the entire pipeline.

"Most of the ‘AI saves time’ story ignores that generating 10 assets and being able to use two of them isn’t actually faster," Davis points out. "Once you count the editing and the brand fixes, it can cost more than doing it properly the first time."

When deployed correctly, however, the efficiency gains can be transformative. Davis notes that tasks which historically required a dedicated art team of six people working continuously for three months can now be executed by a single individual within a single day—provided the resulting work is immediately usable without extensive revisions.

"The meaningful gains come from not having to redo the work afterwards," he says.

This efficiency touches on a sensitive industry debate surrounding creative control and artistic direction. Critics of generative AI frequently voice concerns that automated tools compromise artistic integrity or force studios to sacrifice their unique visual identity. Davis acknowledges these valid concerns, noting that current AI models are entirely capable of producing technically polished assets that remain fundamentally wrong for a specific game’s artistic vision, making stringent quality control an absolute necessity.

At the same time, he warns that much of the software currently marketed to game studios is built on a false premise.

"Most of what gets called ‘creative control’ in GenAI game tooling right now is theatre," Davis says bluntly. "You get a canvas full of boxes and arrows, a dropdown to pick your AI model, and 20 extra steps between you and the output. This may feel like control, but I’d say it’s just the illusion of control."

Instead of adding complex, multi-layered interfaces, Davis believes the primary focus should be on whether a system can genuinely comprehend the overarching intent of the creative team. This technological challenge also intersects with internal cultural hurdles within studios. Beyond technical roadblocks, the adoption of AI frequently encounters emotional and professional resistance from creative teams worried about the future of their jobs.

"I’ve seen studios where the intent is there at the top and the team quietly slow-walks it, or sandbags results outright, because people are scared that if the tools work too well, they become redundant," Davis observes.

Addressing this fear directly is, in his view, a prerequisite for successful integration. "It’s worth saying plainly: this isn’t about cutting headcount. We’re building a team’s capability, where they keep their own workflows and standards while getting trained up on how to actually use the tools."

What the Industry Is Getting Wrong

Reflecting on three years of intensive development, Davis believes the broader tech and gaming industries remain overly fixated on the raw mechanics of AI models rather than the practical outcomes they are meant to deliver to creative professionals.

"The industry keeps mistaking more steps for more control," Davis says. "If a tool needs 20 clicks to hand back what you asked for in one line, then it hasn’t solved anything. It’s just made the process feel more serious."

Ultimately, the proliferation of available models has fundamentally redefined where the true competitive advantage lies. As foundational AI models become increasingly commoditized, the scarcity has shifted away from raw generation capabilities and toward architectural memory and automated judgment.

"My biggest learning is that the hard part was never generation, it was memory and judgment," Davis concludes. "Every platform has more models than anyone needs now, but what’s scarce is a system that remembers what your brand is and checks its own output against it before anything reaches a human."

With SecretSauce Labs now fully operational as an AI content technology firm, Davis and his team are pivoting entirely away from building entertainment software to focus on delivering verified creative outcomes for brands worldwide—leaving their history as game developers behind to solve the systemic challenges they once faced in their own studios.