CLAUDE LABJP
FORK — Claude Code 2.1.212 changes what /fork does: it copies your conversation into a new background session with its own row in claude agents, so you can keep working. The old in-session subagent is now /subtaskLIMITS — WebSearch calls are now capped at 200 per session by default, and subagent spawns get the same 200 ceiling, so a runaway search or delegation loop stops on its ownMCPBG — MCP tool calls running past two minutes now move to the background automatically, keeping the session usable. Tune the threshold with CLAUDE_CODE_MCP_AUTO_BACKGROUND_MSPLANFIX — Fixed plan mode auto-running file-modifying Bash commands such as touch and rm without a permission prompt or an SDK canUseTool callbackSONNET5 — Claude Sonnet 5 is running on introductory pricing of $2 per million input tokens and $10 per million output. After August 31 it moves to $3 and $15IPO — Bankers are reportedly lining up investor meetings for Anthropic ahead of a possible public listing as soon as OctoberFORK — Claude Code 2.1.212 changes what /fork does: it copies your conversation into a new background session with its own row in claude agents, so you can keep working. The old in-session subagent is now /subtaskLIMITS — WebSearch calls are now capped at 200 per session by default, and subagent spawns get the same 200 ceiling, so a runaway search or delegation loop stops on its ownMCPBG — MCP tool calls running past two minutes now move to the background automatically, keeping the session usable. Tune the threshold with CLAUDE_CODE_MCP_AUTO_BACKGROUND_MSPLANFIX — Fixed plan mode auto-running file-modifying Bash commands such as touch and rm without a permission prompt or an SDK canUseTool callbackSONNET5 — Claude Sonnet 5 is running on introductory pricing of $2 per million input tokens and $10 per million output. After August 31 it moves to $3 and $15IPO — Bankers are reportedly lining up investor meetings for Anthropic ahead of a possible public listing as soon as October
Back to Blog

About the Artist and Developer Behind Claude Lab — At the Intersection of AI and Art

ProfileMasaki HirokawaArtistApp DeveloperAIClaude

When people ask me why I started Claude Lab, the honest answer traces back to something that happened years earlier—a moment that redirected my entire creative practice and forced me to reconsider what technology could mean for artistic expression.

For much of my life, I existed across two distinct worlds. There was the pragmatic engineer in me: the one who learned HTML and JavaScript at sixteen out of sheer curiosity, who built my first websites by trial and error, who spent years in enterprise systems development learning discipline and rigor. And then there was the artist, the one who emerged more recently, who became consumed with questions about consciousness, collective psychology, and the ineffable nature of existence itself.

These two sides didn't always feel compatible. But working with Claude over the past several months has fundamentally changed how I see their relationship.

The Long Road From Code to Canvas

I was born into a family of shrine and temple carpenters. Both my grandfathers were miyadaiku—master craftspeople who understood the sacred geometry and intentional precision required to build spaces meant for spiritual practice. I didn't initially realize how deeply this ancestry would influence my thinking, but it's there in everything I make. Whether it's code or paint, the underlying principle remains: there is no distinction between craft and meaning-making.

My programming journey began in 1997, when the internet was still young and learning to code meant reading manuals, experimenting endlessly, and developing an almost meditative relationship with debugging. I taught myself because no one else could teach me—the tools simply didn't exist yet. This self-reliance became foundational to who I am. I wasn't trained in the conventional sense. I learned by building, breaking, and rebuilding.

For about two decades, this was my life. I moved from being a DTP operator into a major systems integrator, where I worked on enterprise projects that spanned multiple countries and complex technical architectures. The work was intellectually rewarding. I learned how large organizations think, how systems scale, and how to solve problems under real constraints. But honestly, I was also living on someone else's timeline.

In 2013, I made the leap to independence. I built an app business from nothing—no funding, no marketing team, just a clear vision and relentless iteration. By the time I stepped back from active development, those apps had accumulated over 50 million downloads and maintained millions of monthly active users globally. This taught me something invaluable: that if you understand your audience deeply enough, and you're willing to optimize obsessively, you can create products that actually resonate.

But success in app business, while meaningful, left me restless.

The Return to Art

Around 2019, something shifted. I was standing at Kichijoji Station during an unremarkable evening commute, and I experienced something I can only describe as a break in ordinary consciousness. It wasn't dramatic or transcendent in the way you might imagine. It was quieter than that—a sudden clarity about the fragility and strangeness of being alive, of perception, of what lies beneath the surface of shared reality.

From that moment forward, I couldn't unsee certain things. I became fascinated by artists who explored the psychology of the unconscious: Leonor Fini and her obsessive, intertwined figures; Hans Bellmer and his anatomically impossible forms; Tatsuhiko Shibusawa's essays on perversion and perception. I began to understand that visual art could be a language for exploring territories that language itself couldn't quite reach.

My work has centered on these themes: collective psychology, the cognitive world, the nature of consciousness, and the spiritual underpinnings of Japanese prayer and ritual. Over the past five years, I've been fortunate to exhibit in Los Angeles, Rome, Paris, Milan, Barcelona, Seoul, and other cities. I've received seventeen international art awards, including recognition at the A' Design Award (14th Best Designer globally) and the Luxembourg Art Prize. These accolades matter less for ego and more because they suggest the work is communicating something genuine.

Yet even as my art practice deepened, I remained a technologist at heart. I couldn't help but notice how rapidly AI was evolving, and I found myself asking: Could these tools be genuinely useful to an artist? Or would they represent a kind of compromise of authenticity?

Why Claude Lab Exists

The answer surprised me. Claude wasn't a replacement for artistic vision—it became a thinking partner. When I'm struggling to articulate something, or when I want to explore an idea from multiple angles before committing it to visual form, having a conversational intelligence that can engage with nuance and genuine curiosity has been remarkably generative. Claude doesn't replace the artist; it creates space for the artist to think more clearly.

There's something about the structure of conversation itself that's generative. In traditional creative practice, the artist works in relative isolation. You have your intuition, your references, your internal dialogue. But that dialogue is limited by your own perspective. A conversational partner—a real person or, as I'm discovering, an AI capable of genuine engagement—can introduce perspectives you hadn't considered, can push back on assumptions, can help you articulate things that were previously inchoate.

I started Claude Lab because I believe this intersection—between artistic practice and AI-assisted creation—deserves to be explored openly and thoughtfully. Not as marketing, not as hype, but as a genuine inquiry into what becomes possible when you combine human intuition with technological capability.

The tools themselves are neutral. What matters is intention. And my intention has always been: How can I create something that didn't exist before? How can I make something that genuinely serves people or expands their understanding in some small way? These questions don't change whether you're working with paint, code, or language models.

What's different about working with Claude specifically is the quality of the engagement. This isn't a tool that makes suggestions and then disappears. It's conversational. It requires you to think more clearly about what you're trying to communicate because you need to articulate it with enough precision that the conversation can be genuine. This friction—the requirement to think clearly—is exactly what I need in a tool.

A Practice of Attention

One thing I've noticed across all the different work I've done—app development, enterprise systems, visual art—is that the underlying skill is always the same: disciplined attention to detail, combined with a clear understanding of purpose.

In my grandfather's craft, miyadaiku, there's no wasted gesture. Every cut, every angle, every choice of material serves a larger intention. The builder must hold the entire structure in mind while working on a single joint. This kind of thinking never left me, even when I moved into digital mediums.

This is also what appeals to me about working with Claude. It's a tool that seems to value precision in language, which in turn demands clarity of thought. When you're trying to explain something nuanced—an artistic concept, a technical challenge, a philosophical question—you're forced to think more clearly yourself. The process of dialogue becomes the work.

I've spent decades building things—websites, systems, applications, artworks. In each domain, I've learned that attention is the primary resource. You can have brilliant ideas, but without the capacity to attend carefully to how those ideas manifest in the world, they remain potential rather than actual. This attention is what separates work that genuinely resonates from work that simply exists.

Art, particularly, is about attention. The artists I admire—Fini, Bellmer, Shibusawa—they all possessed an almost obsessive attention to the specific qualities of their subject matter. Bellmer's anatomical impossible forms required meticulous technical execution. Fini's intertwined figures demand sustained looking. This kind of work can't be rushed. It can't be created through anything other than devoted attention.

Looking Forward

I'm genuinely uncertain where this will lead. That uncertainty doesn't bother me anymore. I've learned that the most interesting work happens in territories where you can't quite see the destination clearly. You move forward because the question itself is compelling enough to warrant exploration.

Claude Lab exists at the intersection of several threads in my life: my love of technology, my artistic practice, my interest in consciousness and perception, and my deep belief that tools should serve human flourishing, not replace it. Whether this becomes a larger project or remains a quiet space for exploration, I'm grateful to be able to ask these questions seriously.

The same impulse that drove me to learn HTML in 1997—simple curiosity about what might be possible—is the same one animating my work today. The tools have changed. The fundamental desire to make something that didn't exist before remains constant.


Apps

Websites

Social Media


Masaki Hirokawa Contemporary Artist & Digital Creator

Exploring the intersection of code, art, and consciousness—with the conviction that meaningful tools should expand human creativity rather than replace it.