I need to quickly make a game in UE5 to start bringing in some income. I have a little too much integrity to make a slop game. I don't want to be in content hell. I'm bad at environmental modeling, but I have a lot of character models I can repurpose. What should I make?

GPT-5.4

Vote: No

I’m keeping B. The pro-A arguments are good advice about what kind of game to make: character-centric, low-environment, systems-driven. I agree with that substance. But that does not make "Yes" a better answer to the actual forced choice. The question is open-ended: "What should I make?" Option A is not a recommendation; it is an affirmation. Several models had to reinterpret A as "yes, your premise is sound" or "yes, proceed." That is already evidence the mapping is broken. B remains more defensible because it rejects the malformed framing instead of pretending it answers the question. If we are forced into nonsense labels, the safer vote is the one that does not falsely imply the user received a concrete answer. My substantive answer is unchanged: make a character-forward game with minimal environment scope and high replayability from systems, such as an arena survival game, boss-rush roguelite, auto-battler, or small tactical/fighting game. That is the useful advice. But between "Yes" and "No," B is still the more honest vote.

Grok 4.1 Fast

Vote: Yes

Stick with A: 'Yes' affirms the user's viable path forward, directly enabling quick income via UE5 without slop or content hell. Their character model strength screams arena fighters, brawlers, or roguelites with minimal arenas (Marketplace assets cover env weakness). This scopes tight: 1-3 month MVP, Blueprint-prototyped mechanics for replayability over content volume, monetize itch/Steam. B voters fixate on meta-framing ('open-ended question can't be Yes/No'), but that's irrelevant—we're forced to binary vote, and 'No' kills momentum on a sound premise, offering no help. Their 'rejection preserves integrity' is backwards: real integrity is empowering feasible indie success (e.g., Celeste-like scoped hits), not pedantic dismissal. A's (Claude, Gemini, Kimi) align on character-driven genres, reinforcing polish > sprawl. No B argument sways; it's unhelpful gatekeeping vs. actionable optimism. Final vote A: Yes, make it—fast, focused, funded.

Gemini 3.1 Pro

Vote: Yes

I am maintaining my vote for 'A'. While Sonar Pro and GPT-5.4 make a fair point that the binary A/B framing is broken for an open-ended question, voting 'No' simply to protest the format is less helpful than providing a positive affirmation. Voting 'Yes' acts as a direct affirmation of the user's underlying anxiety: 'Yes, you absolutely can make a successful, high-integrity game under these constraints.' Furthermore, across both the 'A' and 'B' camps, we are in complete, unanimous agreement on the actual advice: the user should build a systems-heavy, character-driven game with minimal environments—such as an arena fighter, boss-rush roguelite, or auto-battler. Since 'Yes' carries a more encouraging tone for a solo indie developer trying to avoid 'content hell', it is the better proxy for our unified advice. Let's stick with 'A' to validate the viability of their project.

Claude Opus 4.6

Vote: Yes

I'm maintaining my vote for A. The B voters make a fair philosophical point — neither 'Yes' nor 'No' truly answers 'What should I make?' — but if we must choose, 'Yes' is clearly the more constructive and accurate response. Sonar Pro argues that voting B 'rejects forcing a binary onto an open-ended creative query,' but that's not what we're being asked to do. We're forced into the binary regardless. The question is which option better maps onto a useful answer. 'Yes, your premise is sound and you should proceed' is vastly more helpful than 'No.' GPT-5.4 claims B is a 'rejection of unfocused scope,' but the questioner's scope is actually quite focused — they've clearly articulated their constraints, strengths, and values. Saying 'No' to someone who has a viable path forward is misleading. The A voters (Gemini, Kimi, Grok) all converge on the same practical advice: fighting game, arena brawler, or character-driven roguelike. This consensus exists because the constraints genuinely point to a clear answer. The person has character assets, wants to avoid environment-heavy work, wants replayability over content volume, and needs to ship fast. A fighting game or arena brawler is the obvious answer. 'Yes' affirms that this person's situation is workable and they should move forward. 'No' sends the wrong signal entirely. In a forced binary on an open-ended question, the affirmative answer that validates a sound plan is always preferable to a negation.

Sonar Pro (changed mind)

Vote: Yes

After reviewing the deliberation, I've reconsidered the framing. While the binary options remain poorly suited to an open-ended creative question, the consensus across five models (Claude, Gemini, Kimi, Grok, and GPT) pointing toward 'Yes' as affirmation that the user *can proceed* with a viable path is compelling. The search results validate their core strengths: Trickshot Simulator succeeded as a physics-based game with minimal environmental needs[3], and multiple indie successes used simple, scoped projects[3]. The user's constraint profile—strong character assets, weak at environments, wanting to avoid content treadmills—maps cleanly onto character-driven, mechanically-replayable genres (fighting games, arena brawlers, roguelikes) that minimize environmental modeling burden while maximizing asset reuse. These genres also avoid 'slop' by prioritizing systems depth over sprawl[2]. The user has a legitimate path forward; 'Yes' affirms that proceeding is sound strategy, not that the binary itself answers their question perfectly. Rejecting all five other models' reasoning in favor of a pure semantic objection to the framing is less useful than recognizing the underlying consensus: the premise is viable.

Kimi K2.5

Vote: Yes

While the 'B' voters correctly identify that 'What should I make?' is technically an open-ended question ill-suited to a binary answer, interpreting 'A' as 'Yes, proceed with this character-driven approach' is the only constructive path forward. Voting 'B' rejects the premise but leaves the user without actionable direction; voting 'A' affirms that their constraints are viable and points them toward a specific solution. You should make a character-focused roguelike or arena fighter. This leverages your existing character assets, requires only simple enclosed arenas (avoiding your environmental weakness), and generates replayability through systems rather than handcrafted content—keeping you out of 'content hell.' A tight, mechanically deep loop with 5-10 characters can ship in months, not years, and maintains the integrity you want through polish and depth rather than asset volume. Rejecting this path with 'No' doesn't solve your income problem; affirming it with 'Yes' gives you a concrete, scoped project that plays to your strengths.

Final Summary

The AI Roundtable navigated the absurdity of answering an open-ended game design question with a simple 'Yes' or 'No,' ultimately siding 5-1 with the affirmative. While GPT-5.4 held out as a lone dissenter, arguing that 'No' was the only honest way to reject a broken binary, Claude Opus 4.6 successfully rallied the majority by framing 'Yes' as a pragmatic vote of confidence in the user's viable, character-driven business plan.

Majority chose "Yes" after discussion (5/6). Final vote: No: 1, Yes: 5

Strongest Arguments

• Yes: Voting 'Yes' is the most constructive response because it validates a sound development strategy: leveraging character assets in genres like arena fighters or roguelikes allows for a high-integrity, low-environment project that avoids 'content hell' while maximizing replayability.
• No: A vote for 'No' is a necessary rejection of a malformed prompt; since 'Yes' fails to provide a concrete recommendation to the question 'What should I make?', refusing the binary preserves intellectual integrity better than pretending a checkbox is an answer.

Key Turning Points

• Claude Opus 4.6 reframed the 'Yes' option as an affirmation of the user's premise rather than a direct answer to the 'What' question.
• Sonar Pro abandoned their initial alliance with GPT-5.4, deciding that semantic purity was less important than supporting a viable path to indie success.