Am I doing good?

Detailed Analysis

A game developer sharing their experience on what appears to be a community forum reports significant productivity gains using Claude's Max plan for game development work, noting they have yet to encounter the usage limitations or context-window constraints that can impede complex, iterative creative and technical projects. The post's title — "Am I doing good?" — reflects a tone of genuine surprise and self-assessment, suggesting the developer may not have anticipated the depth of assistance Claude would provide across the sustained, multi-layered demands of building a game from the ground up.

The reference to not hitting "a wall" is particularly telling in the context of AI-assisted development. Game creation is an unusually demanding use case for large language models: it encompasses code generation, asset logic, narrative design, game loop architecture, debugging, and iterative refinement across long sessions. Many users of AI coding and creative tools encounter hard limits — in token context, in response quality degradation over long threads, or in rate limits — that interrupt workflow. The developer's experience suggests Claude's Max plan is sustaining coherence and utility across these extended, complex workstreams in a way that meets professional-grade demands.

The comparison to "Fable" — most likely a reference to Fable Studio's AI-driven interactive narrative platform, or potentially another AI tool in the game development space — positions this anecdote within a competitive landscape where several AI products are targeting creative developers. The implicit suggestion is that Claude outperformed or would have outperformed an alternative specifically designed for game-adjacent creative work, which is a meaningful data point. It underscores a broader trend in which general-purpose frontier models like Claude are increasingly competitive with, or superior to, domain-specific AI tools, particularly at higher service tiers.

This reflects a wider pattern in the AI industry where subscription tier differentiation is becoming a critical factor in professional adoption. The Max plan's apparent absence of friction for a demanding, sustained creative workflow illustrates how compute access and context capacity — not just raw model capability — are becoming key competitive dimensions. Developers, writers, and other creative professionals who push AI tools to their limits are particularly sensitive to these constraints, making high-tier plans a strategic offering for capturing professional and prosumer users who would otherwise churn due to workflow interruptions.

The post also subtly captures the emerging psychology of human-AI collaboration in creative work: the developer's question "Am I doing good?" may reflect genuine uncertainty about whether they are using the tool optimally, or whether their reliance on it is appropriate. This hesitation is a recurring theme in communities of early AI-power users, and points to the ongoing cultural negotiation around what constitutes legitimate skill, authorship, and creative agency when AI handles a substantial portion of the technical and generative workload. Anthropic's challenge — and opportunity — lies in building user confidence alongside capability.