Future AI Models Will 'Know Users Better Than They Know Themselves' – Calanthia Mei

Personal Data ‘Immensely’ Valuable to Companies

Although personal data has become “immensely valuable” for companies, Mei said many internet users or owners of the data are unaware of how this is collected and commercialized. However, by learning or understanding just how valuable their data can be, users will not only become adept at protecting this but potentially earn rewards from their data.

Meanwhile, in her written answers sent to Bitcoin.com News, Mei said she foresees a world in which AI models “will know users better than they know themselves.” This knowledge can be anything from predicting users’ needs and desires to offering “hyper-personalized experiences.”

However, AI models can only progress to those efficiency levels by harnessing blockchain technology, Mei said. Also, in her answers sent via Telegram, the co-founder of Masa shared her thoughts on the concept of data staking and how users can profit from this without losing control of their data.

Below are Calanthia Mei‘s answers to all the questions sent.

Bitcoin.com News (BCN): What constitutes personal data in the digital world and what must every internet user know about the data economy?

Calanthia Mei (CM): Personal data refers to your digital footprint. Your every online and on-chain action – browsing, posting, transacting – leaves behind a digital footprint.

As the digital economy and Web3 grows, personal data has become immensely valuable for companies to understand users and train AI models. However, most internet users are unaware of the vast amount of personal data collected about them or how it is commercialized without their consent. Every internet user should be aware of their digital footprint and how to protect their data rights as we move into the AI era.

BCN: Your project Masa claims to be building the world’s largest personal data network and marketplace. Can you explain in simple terms what it does and how this differs from what we are accustomed to?

CM: Today’s personal data — from browsing history to Twitter messages — is locked up in walled gardens controlled by big tech. Your personal data is resold without consent or compensation. This data land grab is accelerated by the proliferation of AI, as personal data is consumed in vast amounts to train AI models and agents.

Masa presents a radical new vision for personal data ownership. With Masa, your data pays you. We believe that every user has the right to own, share and earn from their data. In the new data economy Masa is rapidly building, anyone with an internet connection will be able to contribute their data and earn passive rewards.

BCN: How do you see the data privacy industry evolving in the next few years, especially with the rise of AI and its convergence with Web3?

CM: AI is radically changing the world before our eyes at an exponential rate that no one can yet comprehend. We foresee a world where AI will know users better than they know themselves, predicting their needs, and desires and offering hyper-personalized experiences.

The exponential growth of AI relies on access to vast quantities of high-quality and granular personal data to train ever-more complex models. However, this creates a privacy paradox – where will this data come from, who owns it, where is it stored, and how will rights be protected?

As AI progresses, blockchain provides solutions like zero-knowledge proofs for data verification, decentralized data storage, and decentralized oracles to process data at scale. Vitalik outlines the opportunities in decentralizing AI and increasing AI safety using the benefits of the blockchain in his latest paper “The Promises and Challenges of Crypto + AI.”

BCN: Your platform claims to have invented zero-knowledge soulbound tokens (zkSBT). Can you tell us what they are and how they enable users to take control of their data?

CM: Masa’s invention of zero-knowledge Soulbound Tokens (zkSBTs) was presented and published at the 2023 7th Cyber Security in Networking Conference (CSNet). The zkSBTs empower personal data privacy and control. zkSBTs are like your personal data locker. They reflect your digital footprint across the web and blockchains.

When Masa Mainnet goes live in April, you can visit our app to access your zkSBT data locker. You can view your digital footprint in one place for the first time. You retain full control over who can access your data. If you choose, you can stake your data in our open marketplace to earn rewards in exchange for anonymously sharing your data.

Masa’s pioneering implementation of zero-knowledge proofs and fully homomorphic encryption preserves full data privacy. Users can securely share their data anonymously to earn, while AI models benefit from new private data to learn.

BCN: Whether it’s developers, marketers or AI models, they need a large scale of user behavioural datasets and analytical insights. In your opinion, what are the biggest challenges to obtaining reliable personal data at scale and how can they be addressed?

CM: AI innovation requires high-quality, mass amounts of unbiased and verifiable training data to power their AI models: the most valuable of which is user behavioral data that captures real-world human behavior and natural language. Currently, open and permissionless internet-scale user behavioural datasets sufficiently deep for AI model training do not yet exist. This is because:

• Big Tech Data Silos: Much of the most valuable user engagement data is locked up in proprietary silos controlled by big tech companies like Google and Meta. Simply put, users’ data are locked in Big Tech’s walled gardens.
• Regulatory Privacy Concerns: Collecting and sharing detailed user data raises privacy issues around consent. Regulators are banning 3rd party cookies. There are also many emerging lawsuits regarding training AI models on copyrighted data – the media called it “the big risk behind the AI investment boom.”
• Lack of Incentives: There are very limited financial incentives for users to share their personal data. Users don’t benefit from sharing their own data, big tech does. Billion-dollar industries were built by monetizing users’ data without consent and compensation.

The old data paradigm is broken. A decentralized personal data network could break the closed data silos and architect the Internet’s new nervous system.

BCN: Big tech giants like Google get access to vast amounts of user data because they have popular applications and services like Gmail, Chrome, Search, Maps, etc. Through which apps and services does your platform interact with users and seek their consent to monetize their data?

CM: While Google monetizes user data without permission, Masa empowers people to take ownership and earn from their digital footprint.

Like Google attracts data through popular apps, Masa drives adoption through utilities like Masa Analytics, Masa Growth Campaigns, and Soulbound Domains. Users and partners permit data into Masa’s network in exchange for value and utility. Ahead of Masa’s Mainnet launch, already over 1.2 million wallets have shared 23 million proprietary data points.

BCN: What is data staking? Also, how and why would someone stake their personal data and more importantly, can a user “unstake” their data anytime they want?

CM: Personal data staking is a novel concept. For the first time, data staking lets you profit from your digital footprint instead of platform giants. Just as crypto staking earns yield, you can now stake select personal data and earn rewards when it’s used. Users can stake and unstake their data anytime while fully controlling their personal data locker (zkSBT).

What are your thoughts on this interview? Let us know what you think in the comments section below.