Cryptle 여섯 번 안에 오늘의 크립토 단어를 맞혀 보세요.플레이
0% read

Sapien

Sapien is an AI quality company building Proof of Quality, or PoQ, an evaluation system for AI work. PoQ measures data, reasoning, outputs, decisions, and actions against the customer’s own standard, so teams can systematically evaluate model behavior and target improvements.[9]​[1]​

개요

Sapien is an open, decentralized data protocol that connects human intelligence with AI systems by organizing and verifying human contributions for AI training. It addresses the challenge of large-scale quality assurance by aligning financial incentives with data accuracy and reliability. Under a framework called Proof of Quality, customers define an evaluation rubric, items are reviewed by a panel of independent experts, and their assessments are combined into a signed, auditable score that can be used to verify labels, AI answers, and expert reviews instead of relying on ad-hoc spot checks.[12] The system operates through , validation, reputation, and incentive mechanisms that record contributor performance and task outcomes on-chain. Through this structure, Sapien replaces centralized quality control with an on-chain framework that enforces accountability, verifies data through peer review, and rewards consistent, high-quality contributions.[2][3][10]​

As of June 2025, Sapien reported operating a global network of over 1.2 million registered contributors who had completed more than 100 million tasks, and reported having onboarded 27+ enterprise customers in under eight months across verticals such as autonomous driving, robotics, education, healthcare, and generative media as of that date.[10] Its contributor base spans at least 70 countries and includes domain experts such as doctors, lawyers, and software engineers, who perform specialized data-creation and validation work for enterprise AI models.[9] Enterprises using Sapien’s network include advanced AI and automotive firms such as Amazon, Zoox, Toyota, GAC, MidJourney, Alibaba, and Lenovo, which rely on high-quality human data for applications ranging from self-driving vehicles and robotics to generative media and voice assistants.[10]​

Contributor quality and rewards are governed by the Proof of Quality system, which combines , hierarchical peer validation, and on-chain reputation to determine access to tasks, the routing of work to qualified reviewers, and potential penalties. Higher-reputation contributors who stake more $SAPIEN gain access to more complex workflows and higher-value opportunities, while low-quality or fraudulent work can result in slashing or removal from the network.[5] Growth of the network is supported by ecosystem and education partners such as Coins.ph, , and , which help onboard and train contributors through workshops and onchain community programs.[10] Sapien was founded by a team led by CEO , co-creator of ’s , and CSO , founder of Polymath and author of the real-world-asset standard, who are using mechanisms to build a decentralized network of human AI workers.[11][9]​

Proof of Quality

Proof of Quality is Sapien’s core system for structuring human judgment in AI workflows. It replaces ad-hoc checks with a standardized process in which items are routed to qualified reviewers based on expertise and reputation, each reviewer independently evaluates the item against a customer-defined rubric, and their assessments are aggregated into a consensus score. This score, together with metadata about who reviewed the item and how they scored it, is recorded as a signed, auditable artifact that can be used to verify training labels, grade AI model outputs, or document expert findings.[12]​

Within this framework, and collateral-backed validation support reliability rather than serving as an end in themselves. Before completing advanced tasks, users lock $SAPIEN tokens as that stands behind their validation work, earning rewards for contributions that match panel consensus and risking reductions in stake for low-quality or dishonest output.[5] Quality control is maintained through hierarchical peer validation, where higher-ranked contributors review the work of those below them. Accurate are rewarded in proportion to the difficulty and volume of items they correctly assess, while approving poor submissions can lead to penalties or slashing. Task rewards are influenced by three factors: complexity, performance ranking, and configuration, with larger and longer-lived stakes enabling reviewers to take on more responsibility and earn a greater share of rewards for validated work. Conversely, repeated errors or misconduct can result in retraining, a reduction in stake, or permanent removal. Through this model, Sapien uses PoQ to build a scalable, trust-based system that maintains data integrity via transparent, performance-driven incentives and verifiable review records.[5]​

SAPIEN

The $SAPIEN token functions as an that underpins the protocol’s incentive and validation infrastructure.[8] It is used within Sapien’s on-chain systems for , collateral-backed validation, and emerging agentic mechanisms, rather than to determine how enterprise customers pay for or access Proof of Quality and related services.

Contributors stake $SAPIEN to qualify for work, with locked value backing the quality of their validation and other higher-responsibility tasks. They earn tokens based on task quality and consistency and can reinvest their earnings to access more complex workflows that require greater . Reputation and staked holdings together influence both earning potential and participation in protocol governance, creating a feedback loop that ties contributor performance to the robustness of the validation layer over time.[8]​

In practice, contributors are typically paid primarily in , with a default split of 80% and 20% $SAPIEN, and can choose to receive up to 100% of their rewards in $SAPIEN.[1] The token is used within the Proof of Quality system as for validation work and as a unit in reward calculations tied to on-chain reputation, so that higher-quality contributors who maintain accurate, consensus-aligned reviews can unlock larger task volumes and a greater share of protocol rewards.[5]​

토크노믹스

SAPIEN has a total supply of 1B tokens and has the following distribution:[8]​

  • Supporters/Investors: 26.82%
  • Team & Advisors: 20.18%
  • Contributor Rewards: 15%
  • Community Treasury: 13%
  • Seasonal : 13%
  • Liquidity Incentives: 7%
  • Incentives: 5%

Sapien operates a SAPIEN Staking Rewards Program under which token holders can deposit SAPIEN into the Sapien Vault on and receive vSAPIEN, a token representing their vault position.[13] Rewards are streamed into the vault from an existing 50 million SAPIEN allocation, with up to 10 million distributed per year, so that each vSAPIEN redeems for more SAPIEN over time.[13] The program’s rate is capped, opening at up to 20% annually and declining as the vault grows and as the program matures, and rewards come from this fixed allocation rather than new token creation.[13] Deposits in the vault itself are not subject to slashing; slashing risk applies only to vSAPIEN that contributors explicitly lock as within the protocol’s validation systems.[13]​

자금 조달

Sapien raised a total of $10.5 million in a single seed funding round led by Variant, with participation from Primitive Ventures, , , HF0, and angel investors including YGG founder and SoftBank Vision Fund alum Kevin Jiang.[9][11] The funding supports the scaling of Sapien’s decentralized human-powered data foundry, the development of tooling to better match enterprise AI pipelines with suitable AI workers, and the creation of AI-20, an interoperability standard for decentralized training data developed in collaboration with OpenLedger.[9]​

By the time of the seed round, Sapien’s network included contributors across 70 countries, with participation from professionals in fields such as medicine, law, and software engineering.[9] The company reported working with 17 global enterprise clients, including Fortune 500 organizations, across industries such as autonomous driving, education, medical research, and institutional DeFi.[9] As part of its long-term roadmap, Sapien uses this capital to deepen its decentralized worker network, enhance on-chain incentive mechanisms, and further integrate its protocol with enterprise AI workflows through standards like AI-20.[9]​

잘못된 내용이 있나요?

참고 문헌 (13 출처)

홈카테고리위키 MC이벤트용어집