Dialogue Human Protocol: Human Proof Governance Will Make DeFi Fairer

Compiled by: Bran

Human Protocol is a blockchain infrastructure designed to decentralize human resources by supporting the growth of digital labor markets. Following its recent launch on the Ethereum mainnet, the protocol is now capable of fully automating the lifecycle of data labeling tasks, enabling human-machine collaboration to create and complete a vast number of real-world replaceable tasks.

With the help of artificial intelligence and machine learning technologies, users can now earn rewards in Human's native token HMT for successfully completing anti-bot visual challenges, such as identifying traffic lights in grid images. This data is then organized and used to help eliminate biases in the labor market, promoting a more circular gig economy.

To delve deeper into these recent advancements and their broader implications for the blockchain technology space, Cointelegraph spoke with Harjyot Singh, the Chief Technology Officer of Human Protocol.

Harjyot is an outstanding entrepreneur in the fintech engineering field, with an academic background in computer science and artificial intelligence. His current focus is "exploring how cutting-edge technologies like AI and blockchain can improve the daily experiences of most internet users."

Cointelegraph: How will the recent announcement of Human Protocol's CAPTCHA web application on the Ethereum mainnet support the development of the protocol?

Harjyot Singh: We are excited about our recent achievements. The launch of Human Protocol on the Ethereum mainnet allows us to realize the first instance of a decentralized human labor market. This is also related to the evolution of the protocol.

Human Protocol currently handles a large number of user interactions daily through its supported applications. It is designed to operate across multiple blockchains, with Ethereum being the first mainnet deployment. What we learn and implement here enables us to leverage and execute elsewhere, including on Solana and Polkadot.

Clearly, this launch also allows us to introduce HMT, which helps us grow the human community and incentivize broader participation. But the real growth comes from the Human application: the first portal into the Human ecosystem, and the first way for individuals anywhere in the world to earn HMT directly by completing tasks.

It is also important to note that the Human application is not just a CAPTCHA application—it allows people to perform a variety of tasks.

CT: Readers will be familiar with Google's reCAPTCHA system. What are the technical differences of the Human model, and what are the benefits of a human-centric recognition approach?

HS: It is important to note that hCaptcha is not part of the Human Foundation. It is merely an application that uses Human Protocol. Human has a broader goal of labeling various human tasks, not just a narrow set of tasks that can be run through CAPTCHA.

That said, one major difference between reCAPTCHA and hCaptcha is that hCaptcha pays users for the work they do while solving CAPTCHAs, rather than forcing them to donate that labor to Google.

CT: Vitalik Buterin recently advocated for a transition in DeFi to "Proof-of-Humanity" governance. If widely implemented, how do you see this impacting the space?

HS: I believe this will make DeFi a fairer space. Currently, the system that allocates votes based on wallet balances leads to significant issues. It allows crypto whales to influence proposals that benefit them.

Human's "Proof-of-Humanity" will allow every verified Human user to cast a vote, which will also combat the prevalence of bots. Because Proof-of-Humanity is the first and only on-chain human verification system—it makes sense in the on-chain DeFi world.

But the potential of human proof goes beyond that. Any space where bots cause significant disruption—such as initiating front-running trades on DEX—could apply Proof-of-Humanity to address it.

CT: Can you share some specific examples of how the Human contract can be facilitated in markets utilizing Intel's video and image labeling system, CVAT, and text-based INCEpTION?

HS: AI startups may request 100,000 labeled images of damaged cars. They provide the images along with a sum of HMT, which is held in a smart contract until the work is completed. Human Protocol agents ensure that the data can be securely shared and prepared for applications.

Then, Human Exchanges can intelligently assign tasks to Intel CVAT users, who can operate across different chains and send work to various chains based on speed, cost, etc.

Workers can connect to the trading platform, view the work, and begin completing fine tasks on Intel CVAT by drawing detailed graphics around the damaged areas of the cars. Oracles record and evaluate the work, then update the smart contract to reserve HMT for the workers who completed the tasks.

CT: How does your native token HMT determine the priority of task offerings?

HS: We utilize balance proof as one of the factors that facilitate the ordering of task offerings— in other words, how many tasks will be allocated to one worker or labor pool versus another. However, to reduce friction in the system, we also perform a weighted average of many other parameters to allow new users to join immediately.

CT: How advanced do you think AI and machine learning systems currently are in terms of technical capability and cultural awareness to support human scalability?

HS: AI systems are currently proficient in specialized intelligence. That is to say: they excel at executing specific, linear tasks, such as GPS, chatbots, or Amazon's Kiva robots, which can relay boxes between Amazon employees. But AI does not excel at general intelligence, which is the realm of flexibility, responsiveness, and adaptation—areas where humans thrive.

In terms of cultural awareness, I believe we are ready for the next wave of AI. AI products are already deeply embedded in our lives—from facial unlock systems on phones to robotic cleaners.

However, I think culture overestimates the current capabilities of AI. I believe most people think AI is smarter and more capable than it actually is because we have been talking about AI since the 1950s. Yet, as we saw during the "AI winter" of the 1980s, progress has been uneven.

For example, we have integrated the inevitability of self-driving cars into our cultural knowledge, but they have not fully taken off yet. I think we are ready; I believe people are just waiting for the product.

CT: As we move into a more automated economy, how important is it for us to build systems that serve the true values and needs of humanity?

HS: We hear a lot of different narratives about AI and the impact of bringing machines into the labor market. But rather than replacing human workers, we prefer to focus on how machines can support or even empower them. Intelligent automation means more remedial work—small tasks—can be handled by machines, which helps maximize the time, energy, and attention of human workers.

Humans possess abilities that machines do not—creativity, originality, imagination—while machines are more efficient at executing repetitive tasks. The infrastructure supporting this is particularly well-suited for the growth of knowledge workers, who can provide troubleshooting and specialized input, but whose time is increasingly scarce. This also means providing the data needed for professionals to make informed and confident decisions.

Human Protocol aims to allow machines to complete repetitive tasks and request other machines to take on those tasks. In this way, we hope to unleash human potential and provide space and focus for creative problem-solving.