by Huxiu

July 21, 2026

By Chen Yifan and Liu Xuanqi; Edited by Miao Zhengqing

"Feiniu" sits in a wheelchair, wearing a black headband embedded with electrodes.

He lifts his hands completely off the control panel. The wheelchair glides forward smoothly—no joystick, no remote control, driven solely by the electrical impulses read through his scalp.

Five years ago, a diagnosis of a spinal cord tumor led to two surgeries that left him paralyzed twice. Later, over a 48-hour hackathon hosted by lifestyle platform Xiaohongshu, he and his wife built a mind-controlled wheelchair from scratch, securing first prize in the hardware division.

By SOLOMOAT Editorial Team

With zero prior programming background, he taught himself embedded systems, circuit soldering, 3D printing, and AI algorithm deployment. Today, he showcases his work in the H4 Future Tech Hall at the 2026 World Artificial Intelligence Conference (WAIC), developing affordable assistive devices for China’s three million spinal cord injury patients.

This was the first scene encountered at WAIC this year, serving as a microcosm of Hall H4.

Over the past four years, WAIC’s headline acts have shifted predictably: 2023 was dominated by foundation model parameter races; 2024 was marked as the debut of embodied AI, with humanoid robots commanding center stage; and 2025 spotlighted multi-agent orchestration. By 2026, the true protagonist of the underground hall is no longer a specific technical architecture, but an entirely new organizational species—a decentralized mode of enterprise where founders are neither elite academics nor former big-tech partners, but ordinary individuals.

Feiniu’s case answers a recurring question: when AI radically lowers the barrier to execution, who enters the market? The answer is everyone, particularly those historically excluded from traditional venture playbooks.

In July, Shanghai’s asphalt softens under 40-degree Celsius heat. The physical layout of WAIC acts as an architectural metaphor for the industry’s power structure. The ground floor belongs to Big Tech, model vendors, and state-backed enterprises, marked by expansive, brightly lit pavilions. Yet this year, executives on the main floor—from compute providers running token factories to foundation model developers—repeatedly raised the same concept: the One-Person Company (OPC).

Incumbents track the OPC not out of an aspiration to downsize, but because their next cohorts of enterprise clients, developers, and content suppliers are pivoting from traditional corporate entities to autonomous individuals.

An OPC is defined by structural design rather than headcount: devoid of corporate hierarchies or rigid job descriptions, AI handles operational execution while the human exercises strategic judgment. Whether operating solo or within a lean unit of three to five people, the organizational signature remains identical: an ultra-lean structure, deep AI integration across workflows, and an operating model where the founder’s discernment constitutes the enterprise itself.

Down in the basement-level Future Tech zone, changes emerging from the economic periphery are gaining momentum, threatening to reconfigure the enterprise landscape.

Inside the exhibition hall, lighting illuminates functional zones categorized by neighborhood concepts, featuring bug-reporting desks and impromptu open-mic stages. Compact booths, spanning barely four to five square meters, feature laptops and hand-drawn system architectures. Founders pitch their products directly to passing attendees. Most operate in teams of roughly ten; many consist of fewer than three; several run entirely as solo ventures.

The founder profile here diverges sharply from conventional tech pedigree: rather than primary authors of peer-reviewed papers or serial founders armed with institutional backing, these are individuals driven by personal operational bottlenecks, assembling custom AI workflows to solve their own problems before scaling them to external demand.

01. The Proliferation of the One-Person Company

On Douyin’s micro-drama platform Hongguo, five of the top six titles are AI-generated animated shorts. Synthetic short-form content has shifted from an emerging concept to commercial deployment, frequently produced by one- or two-person teams operating outside formal studio environments.

This year, WAIC inaugurated its first nationwide independent developer competition, the Future Tech OPC Pioneer Challenge, drawing over 600 projects across eight regional tracks.

The macroeconomic data supports this shift. According to the China OPC Development Trend Report released by the Zhongguancun Talent Association, registered sole-proprietor corporate entities exceeded 16 million by the first quarter of 2026, accounting for 27.4% of total domestic enterprises. New registrations climbed 47% year-on-year in 2025, expanding at an average rate of 15,700 entities per day across 426 specialized incubation hubs. While legal registration does not automatically equate to a fully automated solo venture, the metrics reflect a structural normalization of single-operator enterprises.

Conghua Research, which captured the gold award in the startup track, demonstrates an unconventional development path. Founded by Xu Chong without a formal finance pedigree, technical co-founders, or established institutional networks, the venture originated in real estate asset operations. After teaching herself financial modeling and basic software development, Xu deployed a proprietary multi-agent architecture bridging real estate operations, quantitative equity research, and AI workflows to address China’s niche Real Estate Investment Trusts (REITs) market. Operating without outbound marketing spend, the venture onboarded 42 institutional enterprise clients over two years. In her analysis, the operational addressable market for OPCs lies embedded within deep industrial sub-segments.

Lumiverse founder Chang Nan previously led core rendering engine development in North America before managing technical teams at Tencent. He now runs a three-person studio producing interactive entertainment products via generative pipelines, matching the output of traditional teams over six- to twelve-month development cycles.

Their pipeline establishes end-to-end automation: moving from initial concepts to 2D art, 3D asset generation, and final build delivery through orchestrated commercial APIs. Workloads that historically required a dozen specialists across two quarters are compressed into three operators over eight weeks. In Chang’s assessment, their execution benchmarks in 3D production workflows rival specialized business units within major tech conglomerates.

Similar models populate the hall. A serial founder demonstrated a wearable hardware node developed for personal memory logging, targeted directly at managing ADHD-induced context switching. The project, operated by a single-digit team, prepares for its commercial release in North America, highlighting an OPC principle: developing an initial product to resolve the founder’s immediate operational friction. Nearby, another founder showcased a hardware unit that synthesizes personal activity streams into conceptual, AI-illustrated print formats.

Media production represents the densest operational arena for OPCs. OiiOii, a developer tooling studio in the pavilion, provides a streamlined video creation system: users upload sample video assets to train an automated production profile, allowing the engine to generate fully rendered, style-consistent video assets from text inputs across news reporting, educational explainers, and market analysis. Their core client base spans individual creators and micro-drama production houses, with solo operators generating individual assets reaching four million engagement metrics—a workflow that previously required a dedicated director, editor, and channel manager.

Current OPC trajectories broadly follow three operational archetypes:

The Deep-Vertical Solo Specialist (exemplified by Conghua Research): An individual operator penetrates a niche industry vertical, deploying specialized agent systems to replace operational overhead and relying on cognitive domain advantage and proprietary internal data.

The Elite Technical Unit (exemplified by Lumiverse): High-caliber technical operators maximize collaboration density through generative workflows, matching the productive output of corporate engineering units.

The Demand-Led Operator: Founders lacking deep technical or domain moats who maintain highly specific customer insights or distinct aesthetic intuition that resist algorithmic standardization.

02. "Licensing a Fully Formed Enterprise"

Beyond lean production, the OPC movement presents a structural evolution in organizational design.

Ark InsightX, an active agent and wearable hardware startup of four full-time staff and three interns, runs internal operations entirely around agent orchestration. The team defines a core strategic objective, which a primary orchestration agent decomposes into discrete work packages. Rather than relying on traditional middle management for allocation, team members claim tasks dynamically. Operational boundaries remain fluid across product management, engineering, enterprise sales, and pipeline maintenance. Project contributions are evaluated algorithmically by agents tracking computational complexity, token utilization, deliverable quality, and market visibility.

Here, operating hierarchy is generated dynamically around deliverables rather than static titles. Human capital moves fluidly within an agent-managed workflow, lowering the switching cost between domains. A non-technical team member can fine-tune specialized models via structured internal repositories, while a product designer simultaneously handles enterprise outreach.

This shift has fostered a secondary market of infrastructure providers building tools specifically for solo operators.

While OiiOii packages creative capabilities into exportable micro-skills, Outstory operates further upstream in interactive narrative production. The platform supports creators from initial concept generation through interactive branching, asset rendering, global distribution, and post-launch telemetry feedback. At its recent creator summit in Shanghai, which drew 4,000 international developers, top-tier independent creators on the platform reached eight-figure annual revenue distributions in RMB. Outstory absorbs operational friction—including overseas distribution compliance, platform onboarding, and narrative logic design—through automated toolchains.

OpenCSG approaches the market via open-source model repositories and compute assets. Serving over a million users, the platform enables independent developers and OPCs to structure proprietary datasets and share validated model weights. As founder Chen Ran notes, the primary operational deficit for most solo ventures is not ideation, but access to verified production benchmarks and differentiated proprietary data.

Neofield focuses on structural enterprise replacement. Reasoning that solo founders struggle most with simultaneous execution across product architecture, UI/UX, engineering, deployment, and customer success, Neofield provides a collaborative workspace populated by persistent, specialized agents—acting as dedicated product managers, technical writers, and designers. Unlike basic task-completion agents, these instances maintain long-term institutional memory, internalize operating procedures, and accumulate domain assets.

These structured workspaces consolidate internal documentation, proprietary data, and custom automated skills. Over time, these compiled virtual work environments can be packaged, licensed, or acquired directly. Rather than purchasing discrete software tools, operators can effectively license pre-configured, operational enterprise units.

03. The Strategic Deficit: Solitude, Execution, and the Real Odds

At the center of Hall H4, Spark-Innopower operates as an investment and incubation firm focused on the OPC ecosystem. The firm itself runs as a three-person venture comprising the founder, a CTO, and an operations lead. The core thesis: genuine OPCs are defined by deep AI-driven execution leverage; absent automated leverage, the model reduces to traditional freelance labor.

Securing institutional venture backing for solo ventures remains counter-intuitive. As Hu Xuewen, Vice President at Haochen Capital, observed, OPC operating economics diverge sharply from the capital-intensive scaling playbooks of the consumer internet and deep-tech eras. Traditional venture diligence frameworks frequently misprice these assets, presenting an asymmetric investment opportunity for funds willing to underwrite non-standard organizational leverage.

Yet the operational reality of solo ventures is demanding. Spark-Innopower prioritizes immediate commercial validation and cash generation. One portfolio investment—founded by a first-year undergraduate at Shanghai University of Finance and Economics—monetized an event aggregation platform within two months via performance-based commission structures. Because computational inferencing and token expenditures scale directly with platform traffic, solo ventures require viable unit economics from inception to sustain operations.

AI lowers technical execution barriers, but it steepens market competition. In markets like AI micro-dramas, hit rates often sit below 0.01%.

OpenCSG’s Chen Ran describes a steep funnel across the open-source developer base: broad experimental participation, limited commercialization, and minimal sustained enterprise survival.

Haochen Capital’s Hu identifies three primary failure modes common to the model:

Resource Exhaustion: A solo operator pilots a minimum viable product, but abandons development when go-to-market validation fails to materialize within two to three months.

The Community-to-Commerce Gap: A product generates traction within developer communities, but fails to cross the threshold into enterprise procurement or paid utility.

Organizational Fracture: Ventures that successfully achieve product-market fit disintegrate when expanding into multi-person teams due to misaligned equity structures, conflicting operational expectations, and governance friction.

Solo founders fail less often from technical deficiencies than from the compound burden of maintaining capital discipline, driving commercial conversion, and managing organizational scaling simultaneously.

04. Strategic Discernment as the Core Value Anchor

When the marginal cost of execution approaches zero, strategic discernment becomes the scarce economic asset.

As Lumiverse’s Chang Nan observes, the operational bottleneck has migrated rapidly: from software engineering to asset production, and ultimately to product discernment. When execution is automated, human intervention concentrates on structural selection—determining what to build and what to discard.

In traditional enterprise models, operational execution served as a primary moat. Today, developer environments like Cursor, generative image platforms like Midjourney, and video synthesis engines like Sora commoditize production. The strategic focus shifts from mechanical capability to problem selection.

Kun Peng of Hugging Touch draws a parallel to cinema: while post-production line roles will increasingly be absorbed by automated workflows, core above-the-line roles—directors, screenwriters, and producers—command greater leverage. The defensible capabilities within the OPC framework reduce to two vectors: demand discovery (user insight) and editorial taste (unique product positioning).

Economic history suggests that as a resource shifts from scarcity to abundance, its complementary assets capture outsized value. Information proliferation elevated the value of human attention; similarly, commoditized computational production elevates the value of strategic judgment. The ultimate capital asset of an OPC is the founder’s discernment—a quality that resists automated replication and operational scaling.

In content markets, solo operators maintain an advantage by acting as structural outsiders to foundation models. As models optimize for mainstream statistical averages, the resulting outputs converge toward standardized median formats. Breakout success, however, stems from directional variance. An individual operator’s personal constraints, distinct thesis, and specific operational experience generate non-standard outputs that algorithmic models cannot predict.

This structural dynamic requires OPC operators to calibrate their exposure relative to foundation model labs rather than traditional enterprise competitors. Because inferencing costs vary and model tiers evolve rapidly, solo ventures must actively avoid building in areas that align directly with model capabilities.

Products designed around deterministic problem spaces—such as standard software routines, structured mathematics, and baseline diagnostics—lie on the direct optimization path of foundation model vendors. Conversely, subjective domains devoid of deterministic loss functions—including non-linear narratives, specific artistic taste, and complex creative direction—remain structural blind spots for current model architectures. Sustainable venture design requires anticipating foundation model roadmaps and engineering around their direct trajectory.

This delineates two defensible operating paths for modern solo ventures:

Content-Driven Ventures: Relying on taste, variance, and non-standard perspectives to deliver outputs that foundation models cannot anticipate.

Vertical Domain Ventures: Anchored in specialized industry know-how, proprietary workflow integration, and private operational data.

The moat for the former is variance; for the latter, specialized exclusivity. Both models avoid direct competition with general-purpose frontier models.

As OpenCSG’s Chen Ran summarizes: foundation models undergo constant revision, but proprietary data compounding retains durable value. In an operating environment where all market participants access identical model APIs and commoditized tooling, competitive advantage migrates from the toolchain to the proprietary inputs fed into the system. Infrastructure remains a shared commodity; cognitive discernment and proprietary data remain private assets.

These lean enterprises represent an active experiment in organizational structure. While project-based micro-settlements and agent-evaluated performance frameworks remain early in their development cycle, their structural implications are already influencing broader corporate strategy.

Spark-Innopower advocates for internal project-based compensation over fixed monthly salaries. Under traditional industrial models, hourly compensation functioned because physical labor scaled linearly with time spent on the line. In generative workflows, productivity decouples entirely from temporal input: an operator leveraging tailored agent pipelines can achieve in two hours what manual execution requires two weeks to produce. Under these conditions, time-based compensation models penalize high-leverage execution. Project-based compensation prices the deliverable directly, indifferent to operational duration, tooling selection, or the proportion of synthetic execution involved.

Under this framework, collaborators operate as micro-vendors rather than conventional employees, shifting internal corporate governance from administrative management to commercial transaction. Organizational models transition from centralized time ownership to decentralized capacity trading, wherein every participant within the network functions as an autonomous operating unit.

Historically, corporations expanded headcounts to manage coordination complexity, establishing management hierarchies, project managers, and internal review systems to mitigate information loss across functional silos. As AI drives the marginal cost of coordination and synthesis toward zero, the economic rationale for large administrative hierarchies diminishes. Corporate organization points toward decentralized, sovereign units operating across personal, enterprise, and institutional layers. The OPC serves as the functional operational unit of this decentralized model.

Hu Xuewen notes that modern AI-native operations emerge organically from the bottom up rather than via traditional top-down corporate planning. Operational nodes assemble dynamically around opportunities, forming decentralized task networks coordinated by agent layers.

Kun Peng notes that technical teams entering the market today approach development without reliance on legacy corporate assets, deploying architectures from zero to one through modern AI frameworks. For this cohort of operators, capital and organizational assets are outputs of successful deployment rather than prerequisites. Enterprise adaptation increasingly relies on supporting bottom-up autonomous development and integrating cleanly with external solo operators.

05. Structural Reorganization from the Ground Up

Outside the subterranean hall, WAIC’s primary exhibition levels remain defined by corporate scale: enterprise pavilions, continuous promotional displays, and institutional delegations negotiating high-value financing rounds.

Yet beneath this established exterior, operating fundamentals are adjusting. The attention given to the OPC model by enterprise leaders indicates an awareness of this structural transition.

In the lower-level exhibition halls, early-stage, loosely structured ventures are gradually challenging the standard conventions of modern corporate enterprise—fixed working hours, monthly payroll, multi-tiered approvals, rigid job descriptions, and annual budgeting cycles. The legacy infrastructure of the firm faces pressure from decentralized, lean operational models built directly on autonomous workflows.

These new organizational forms are navigating structural shifts in the market, scaling outward from deep within the industry's operational framework. The legacy tier above has noted the development; the remaining question is whether they will integrate these decentralized networks or concede market share to them.

Frequently Asked Questions

How can a solo operator compete with a large corporation?

Start where institutional complexity produces slow decisions, generic service, or ignored customer constraints.

What makes an entry point strategic?

A strong wedge gives access to real demand, creates learning faster than incumbents can respond, and can expand into an owned asset.