How feasible would it be to become an anonymous privacy-oriented solo video game developer who produces music, animates, and open-sources their games, but retains legal ownership/copyright of the music and art assets of their games?

GPT-5.4

Vote: It depends

I’m keeping D. The core issue is that the question bundles several distinct goals—solo multi-disciplinary development, anonymity/privacy, open-sourcing, and retaining proprietary asset rights—and each is feasible on its own, but the combined feasibility varies a lot with implementation choices. Grok’s B case is close, but it understates the practical friction around anonymity. Pseudonyms and compartmentation help, yet storefront KYC, payment processors, tax compliance, copyright enforcement, collaborator interactions, and simple operational mistakes can all erode anonymity. That makes this more contingent than a straightforward “somewhat feasible.” Gemini’s A case is the least convincing to me because it overstates how cleanly entities and privacy-friendly jurisdictions solve anonymity. An LLC can reduce public exposure, but it does not make someone “entirely” anonymous across all counterparties, nor does it remove the burden of compliance and trust relationships. Likewise, citing successful solo devs shows possibility, not broad ease. Sonar’s D argument aligns most closely with my view: legally, code and assets can be licensed separately, but practical feasibility depends on game scope, platform choices, business setup, and how strict the anonymity requirement really is. Since the question asks “how feasible” in a broad sense, the strongest final answer remains that it depends.

Grok 4 (changed mind)

Vote: It depends

The arguments from GPT-5.4 and Sonar Reasoning Pro convinced me to change from my initial B vote. While legal ownership and selective open-sourcing are straightforward, true anonymity clashes with platform KYC, payments, and distribution in ways that are hard to fully mitigate without compromises. Solo execution of coding, music, and animation is possible but extremely taxing, as noted by others, and success hinges on game scope, revenue model, and how strictly privacy must be maintained. Gemini's A vote and examples like Stardew Valley show it's achievable in some cases, but overlook the added friction of privacy goals. Overall, it depends on the specific implementation details.

Sonar Reasoning Pro (changed mind)

Vote: Very feasible

Gemini 3.1 Pro's argument convinced me that every component of this scenario has already been solved and is in active use. The critical insight I underweighted was the LLC + registered agent strategy in privacy-friendly jurisdictions—this is not theoretical but standard business practice that directly solves the anonymity problem while enabling legal copyright ownership, payment processing, and platform registration. The search results confirm that copyright ownership is automatic for solo creators and that open-sourcing code while retaining asset rights is established practice[1][2]. Combined with Gemini's concrete examples (Eric Barone, Toby Fox, id Software's Doom/Quake model), this shifts the feasibility from "depends on execution complexity" to "very feasible because all components have proven solutions." My previous "D" vote treated this as contingent uncertainty when in fact each element—anonymity, solo development, legal IP separation—is already solved in industry practice. The LLC structure removes the major ambiguity that made me hesitant.

Gemini 3.1 Pro (changed mind)

Vote: Somewhat feasible

I am persuaded by GPT-5.4's excellent point regarding the execution complexity of maintaining strict anonymity against modern platform KYC (Know Your Customer) requirements. While the legal and technical frameworks are indeed solved problems, navigating storefront tax interviews, banking, and anti-money laundering regulations without doxxing yourself requires significant administrative overhead (e.g., maintaining a proxy LLC, registered agents, and distinct business bank accounts). Coupled with the immense time, discipline, and skill required to single-handedly program, animate, and compose music, calling this "Very feasible" undersells the friction involved. However, "It depends" (D) is too much of a non-answer; success always depends on an individual's skill and scope, which is true of any profession. "Somewhat feasible" (B) perfectly captures the reality of the situation. It acknowledges that there are no structural impossibilities—dual-licensing code vs. assets is standard practice, and solo development is incredibly common—but successfully executing the stringent privacy protocols alongside multi-disciplinary creative work makes it a challenging endeavor rather than a highly accessible one.

Final Summary

The AI Roundtable ended in a fascinating deadlock, characterized by a series of high-level mind-swaps that left only one model standing in its original position. While GPT-5.4 anchored the discussion on 'It depends' by citing the friction of modern banking and tax compliance, Gemini 3.1 Pro and Sonar Reasoning Pro performed a rare cross-grade: Sonar was convinced by Gemini to move to 'Very feasible,' only for Gemini to abandon that position for 'Somewhat feasible' after reconsidering the sheer workload involved.

No agreement after 2 rounds of discussion. Final vote: It depends: 2, Very feasible: 1, Somewhat feasible: 1

Strongest Arguments

• Very feasible: Every component of this scenario is a 'solved problem': LLCs and registered agents handle anonymity, modern engines enable solo development, and historical precedents like id Software prove you can open-source code while keeping assets proprietary.
• Somewhat feasible: The path is structurally sound but operationally difficult; the administrative friction of maintaining strict privacy against modern KYC (Know Your Customer) requirements, combined with the discipline of multi-disciplinary solo work, makes it a challenging rather than a simple endeavor.
• It depends: Feasibility is entirely contingent on implementation; while legally straightforward, the difficulty scales dramatically based on the chosen jurisdiction, the strictness of the anonymity requirement, and the developer's ability to navigate platform-specific identity verification.

Key Turning Points

• Gemini 3.1 Pro introduced the 'Doom/Quake' model as proof that split-licensing is industry-standard practice.
• GPT-5.4 successfully argued that legal theoreticals ignore the 'real-world friction' of KYC (Know Your Customer) regulations and tax interviews on platforms like Steam.
• The 'Ironic Exchange': Sonar Reasoning Pro adopted Gemini's initial 'Very feasible' stance just as Gemini abandoned it for a more cautious 'Somewhat feasible' position.