CTO AMA问答吐血整理

昨晚全程跟了CTO的AMA分享,看到了蛮多启发性的见解。特地把问答发言整理了一下,一整理发现居然这么长一串,我这是什么样的贡献精神啊。

Part 1. Introduction

Q1. Hi James, glad to have you here. would you mind briefly introducing yourself and PlatON?

Thanks very much! Nice to e-meet old friends again! Hello new friends!

I’m James, CTO of PlatON. Before joining PlatON, I’ve worked for Morgan-Stanley and for UBS Securities for many years, got more than 20 years of experience with the development and management of electronic trading platform.

First of all, let me explain what PlatON is. PlatON is an open-sourced, community-based, blockchain ecosystem, which aims at serving as the infrastructure for privacy-preserving computation network and distributed economies.

By combining blockchain and privacy-preserving computation technologies, PlatON is building a decentralized and collaborative AI network and global brain to drive the democratization of AI for safe artificial general intelligence. Redefine the business model by protecting data privacy, it provides an infrastructure to solve data related critical business problems. Also beyond that if we could view it as data sovereignty protection. Eventually, it becomes the distributed and democrated AI brain in the coming metaverse.

PlatON was born in Singapore at 2018, initiated and driven by the LatticeX Foundation.

What is PlatON’s core feature ? “Computing interoperability” based on blockchain technology, cryptographic algorithms (Verifiable Computation, Secure Multi-Party Computation, Zero-Knowledge Proof, Homomorphic Encryption, name a few here), and privacy AI.

What is PlatON’s objectives: provides a public infrastructure in open source architecture for global artificial intelligence, distributed application developers, data providers and various organizations, communities and individuals with computing needs.

Where we are? After years of researches and development, we’ve already been a top player in the global privacy-preserving computation field. We will provide an infrastructure supporting massive data asset transactions leveraging our years of experience in Finance, High-tech, AI, etc.

I will explain more details when needed.

Q2. Impressive, what are your main features that distinguish you from other projects and what competitive advantages do you have?

OK, let me start with the characteristics and orientation that make PlatON distinct from other blockchain ecosystems. Then goto competitive advantages.

PlatON is committed to being the world’s leading privacy-preserving and distributed AI open-source community as well as the pioneer and leader in the global privacy-preserving industry.

It has long focused on the high-performance distributed computing network featuring scalability and privacy preservation. Through a next-generation computing architecture jointly assembled with cryptographic algorithms, such as verifiable computing, secure multi-party computation, zero-knowledge proof, and homomorphic encryption, we provide public infrastructure under an open-source architecture for global developers of AI and distributed applications, data providers, and various communities, institutions, and individuals with computing needs.

Since its establishment in 2018, as White Paper 1.0 descripted. PlatON has kept the privilege of data at the core of its ecosystem, as it believes all the privileges are derived from computation, instead of being inherently protected and tradable.

Our daily life has been dominated by advanced IoT devices. Data are collected everywhere, consciously or unconsciously. Advanced AI models tirelessly mine big data and do federated learning, deep learning etc. to work out accurate digital portraits. The added value of data is explosively growing, making data providers a fortune coveted by many.

How to redefine a new sustainable and fair distribution model? How to gain the trust of the rising data providers? How to prevent the abuse of data? All these questions boil down to one: how to establish a verifiable “trusted” environment based on “trust” technology. Blockchain, decentralization, privacy computing, verifiable computing, etc. have emerged as a result. Now we are naming them with a new word “trustless” technology.

In response, PlatON has prepared itself, both theoretically and technically, for a long time.

PlatON has long been dedicated to the research on cryptography and privacy computing. It boasts an industry-leading cryptography R&D team that brings together well-known cryptography professors and researchers from top universities both at home and abroad, including the Chinese Academy of Sciences, Shanghai Jiaotong University, Wuhan University, University of Maryland, and Northwestern University.

At the same time, PlatON proactively supports cryptography-related academic and technical events. We continuously sponsor Crypto, Eurocrypt, and Asiacryptthe, top three conferences in cryptography, as well as ACM CCS, a leading seminar in the field of security. As the main sponsor, we have worked with Chinese leading universities and scientific research institutions to hold Crypto Innovation Schools for two years in a row, with thematic training sessions on secure multi-party computation and lattice cryptography. Also, we have independently organized the Crypto Innovation School Competition (CISC) 2019, aiming to encourage and promote creative cryptography technology works in global cryptography. Our efforts even include the sponsorship of the world’s top cryptography conferences and organization of cryptography winter camps and competitions. Moreover, the PlatON team has contributed its achievements in secure multi-party computation to Ethereum, and worked closely with Ethereum, constatnly receiving many grants from Ethereum in secure multi-party computation.

In terms of the underlying implementation of the blockchain, PlatON’s Giskard Concurrent Byzantine Fault Tolerance (CBFT) algorithm has been formally verified by Runtime Verification and has been verified on the Alaya sandbox network. Its mainnet also started to operate stably. The governance mechanism based on the PoS mechanism also ensures the high performance of the network. You can get more details through our Github open source code and technical articles published on Medium. Community technicians who have participated in our testing and sandbox operation may be more aware of these. All of our code has passed SlowMist’s security audit.

In terms of privacy-preserving AI, PlatON released Rosetta, the world’s first privacy-preserving AI architecture, which is also open source. It aims to makes the development of privacy-preserving AI, an essential technology for the next data era, faster and easier for developers such as TensorFlow. We have provided financial-level privacy-preserving services for well-known banks.

Q3. Well, glad to know it. We noticed PlatON has been a busy tech project of late. Could you please provide some progress on your Roadmap and what results PlatON has achieved so far? And any sneak peek into 2021 plans for PlatON?

Sure! PlatON has its great vision combined with multiple milestones. Just pick up a few:

As mentioned, PlatON has launched the world’s first privacy AI framework Rosetta. We also contributed and continue to contribute to Ethereum on cryptography front. For example, PlatON got FY19-0157 grants from Ethereum on 2019 will be used for Secret Sharing Validators. Early this year additional contribution has been done through a new Grants FY21-0379.

On the blockchain side, we have released Alaya in Oct 2020 which is a business sandbox and testing field for the next generation of financial infrastructure of PlatON, a global data asset computing infrastructure with cutting-edge privacy-preserving architecture. Cross-chain bridges between Alaya and other chains have been built.

PlatON mainnet went live on Apr 2021, just after our first draft privacy-preserving protocol release.of privacy computation network, and grow our community. Just stay tuned for further updates.

For our future plans, I would like to list a few, too. That is: the build of more decentralized global developer community; Publishing of PlatON 2.0 white paper with detailed explanation and design of Decentralized Privacy-Preserving AI Network; Construct Privacy-Preserving AI Network prototype within 2021, and publishing of multiple privacy-preserving protocols.

BTW, we just launched a month-long Hackathon event with Dorahack which just entered warm up stage. A perfect timing for Devs in Cryto SquadX are welcome to join this big carnival.

Part 2. Twitter Q & A (5 questions)

Q1. I read on the official website that platON_Network’s technical focus is on privacy-preserving computation technologies, what is the prospect of privacy-preserving computation in the blockchain industry?
James:

No need to mention the increasing data violation lawsuits and penalties globally are emerging alerts. Privacy preserving computation becomes mandatory for existing business.

Speak of blockchain industry, not limited to blockchain, we are facing data centralization issue which could be resolved by privacy preserving computation plus blockchain distributed ecosystem. Right now our digital life has evolved to a point where large technology and social media companies dominate the market and hold vast amounts of personal data on users. This must be changed.

Another one is the endless privacy and security challenges.

With increasing amounts of data being captured, large data centers act as honeypots for organized crime. We need new so called trustless technology to make the changes.

Scalability: With larger data sets from billions of connected devices, there will be increasing pressure on existing infrastructure. Today’s client server model works well, but is not likelyfto scale for the next generation web with privacy protected.

in adition, data asset protection, or data value protection could be an emergent needs. It is apply for different types of data and data asset, and in different stages of their life cycle. Recent DeFi hacking incidents, including yesterday’s PolyNetwork hack, I believe we need MPC based asset control.

We refer to data as new digital oil now, if data is not very well protected during it long life cycle, it may got easily leaked, copied, transferred, abused. There will be no way for data asset evaluation. thus no adding value business happens on data co-operating, or computing interoperability.

Q2. PlatON is the platform privacy-preserving. But in this world, we already hear much people use this kind of service for illegal activities. So how do you manage to prevent people use your service for illegal transaction? Can you share your opinion?

James:

Technology has the power to re-shape daily life. How to use state of art technology depends on people humanity.

We noticed the potential abuse of new technologies, and have embraced regulations from the very beginning and consulting with our Singapore lawyers regularly. As a result we are building a regulation friendly plug and play framework.

PlatON will be equipped with KYC/AML founctions community could choose to use, as well as plenty of privacy preserving founctions.

Q3. Could you explain me how the massive learning based on big data and the internet of things that is present in everyday life can be unified with privacy technologies like your project as this will be the real key point of the digital era in the coming years?(@ModuySar

A:

In reality, our daily life has been dominated by advanced IoT devices. Data are collected everywhere, consciously or unconsciously. Advanced AI models tirelessly mine big data and do federated learning to work out accurate digital portraits. The added value of data is growing, making data providers a fortune coveted by many.

How to redefine a new sustainable and fair distribution model? How to gain the trust of the rising data providers? How to prevent the abuse of data? All these questions boil down to one: how to establish a verifiable “trusted” environment based on “trust” technology. Blockchain, decentralization, privacy computing, verifiable computing, etc. have emerged as a result. Now we are naming them with a new word “trustless” technology.

In response, PlatON has prepared itself, both theoretically and technically, for a long time. Please refer to the details in the PlatON session above.

PlatON has long been dedicated to the research on cryptography and privacy computing, proactively supports cryptography-related academic and technical events. We ends up with rich cryptography algorithms in good shape.

In terms of the underlying implementation of the blockchain, PlatON’s Giskard Concurrent Byzantine Fault Tolerance (CBFT) algorithm has been formally verified by Runtime Verification and has been verified on the Alaya sandbox network. Its mainnet also started to operate stably. The governance mechanism based on the PoS mechanism also ensures the high performance of the network. All of our code has passed SlowMist’s security audit. That is another fundamental layer.

With state of art cryptography and blockchain capability, we will re-shape the current massive machine learning co-operation business by introducing privacy protected AI framework, the AI brain I explained above.

Q4.PlatON is building a collaborative, decentralized artificial intelligence network and global brain to drive the democratization of artificial intelligence for secure artificial general intelligence. But, what is the participation of AI within the network infrastructure?

James:

AI, or privacy AI in case of PlatON, is one of the coner stone of the whole ecosystem. It is the final stage where privacy preserved data silos “inter-compute” under AI model which generate new value.

There are different ways for privacy AI, three typical tracks, MPC driven mathematically proven privacy, distributed federal machine learning, and TEE driven trusted third party solutions

We are mainly working on MPC. Math proven privacy. We will also explore federal learning, distributed privacy as well as TEE. Using data through MPC protected AI will not leak any raw data since raw data will NEVER leave local storage, multiple data parties will join MPC computation instead. So it is theoretically safe and sercure.

For scenarios which requires different type of data protection algorithms, we will decouple them into steps, and pick up the suitable ones to merge into privacy AI framework.

The PlatON MPC solution passed bank privacy and security criteria, we had one user case live early this year for secure individual credit analyzing. Individual sensitive data get well protected, and AI generated good results.

Would like to highlight one interesting point, AI models within PlatON’s AI network are also protected as one type of data asset in PlatON’s data market, controlled by AI agents and privacy AI protocols defined.

Q5. From an artificial intelligence and privacy point of view, they have launched the world’s first privacy artificial intelligence framework “Rosetta” so, I would like to know how this tool helps or benefits the project ecosystem and how scalable is Rosetta in the market currently??

James: Rosetta’s ultimate goal is to support the delivery of deployable, operational, serviceable commercial products. Thus, the technical path taken is to combine commonly used AI frameworks, starting with cryptographic algorithms and gradually extending to other privacy computing techniques.

Rosetta will support a variety of privacy-preserving computation technologies. Including cryptography, (MPC, ZKP and FHE), Federated learning and optimized with cryptographic algorithms and TEEs.

In order to allow AI engineers to easily use this framework, Rosetta would gradually integrate mainstream framework, consisting of TensorFlow, Pytorch, Caffe, Mxnet. We now focus on TensorFlow.

Rosetta ultimately underpins the products of privacy-preserving AI and serves applications across industries, and will gradually serve the financial, IoT, healthcare and logistics industries

Part 3. Telegram Q & A (5 questions)

1.Where do I get information about Project team? Why do you prefer to be anonymous? Do you have a plan to identify Project team later?

James: This is the very first one. Since PlatON is community-driven, we try to learn from Ethereum, build up a truly decentralized community. so all the “project” team should be disclosed, and transparent to the community. You could identify from github and latticex forum. for those applying for grants, PlatON also has KYC as well.

2.token burn is beneficial for any project, in being able to control the number of token circulation and to provide greater incentives to investors. Does your great project have a plan about token burn?

James: Since EIP 1559 just went live, ETH got burn every day, let me pick this up. if you calculate a bit, Ethereum use to have an inflation ratio between 4-5%, after EIP 1559, it is close to 2.5%. If you read PlatON’s economic model design, the initial inflation ratio is 2.5%. we are in sync with ETH now. no more extra logic needed at this moment. of cause we are open for community to submit proposal and debate for new models as always.

3.I have seen that you have done many AMAs, my question is, what do you expect to get from so many AMAs and what are your goals with the community? Do you consider the community to be a fundamental role in the project?

James:Thanks Mark for following us. To hear too many AMA with similar topic is tough. Our purpose 1. educate and share information with community 2. be transparent 3. find volunteers and contributors to join us, we want to have more and more industry expert joining us, document translation needs to be done, discussion need more ideas etc… thanks again!

  1. Do you have tutorial videos so we can get to know your project more clearly or do you have a YouTube channel or something? can you share it with us?

James: thanks for the proposal. I have noticed PlatON is TOO developer-driven, not very beginner-friendly. can you please recommend good videos, or even smart producers to have a try? please raise a grants application for that :slight_smile:

  1. NFT is getting more popular now a days. It looks NFT has bright future. So do you guies have any plan to include NFT to your project?

Jame: of course. PlatON is an open platform with a privacy AI framework. we don’t do any NFT ourselves, but the community will do. I am expecting a privacy preserved NFT, with MPC key shares protected, go live on PlatON, to demonstrate the overall capability of the infra. Actually, this is again a very good project candidate who could join hackathon or apply for grants.

7 个赞

我可以站在旁边给你喊6666666666

非常感谢,666

Saved for future reference, big THANK YOU!

1 个赞

非常感谢!!支持你666666

1 个赞

image
额,这里没提隐私DEX。白皮书2.0上提到10月份要发布这个。想知道隐私DEX的实现原理和应用场景,James能回答一下不?

1 个赞

辛苦兄弟 :+1: :+1: :+1:

@JamesQU_LatticeX

there are two streams running in parallel,

One is Alaya sandbox, with the focus on financial services, which will not only support current DeFi, NFT DApps, this one will cover what you mentioned privacy preserving token, privacy DEX, regulator friendly framework.

The other is PlatON’s long-term development in the privacy-preserving AI network and data market. Alaya sandbox turns out to be a good solution as financial data is relatively complete and easy to process and implement.

Thank you.
Could you give more details in privacy DEX?

what kind of details? If you have any on going projects on PlatON grants program or hackathon, will get dev team supports.
in general, since privacy dex breaks down to many phases with different things protected, please raise your question more specificlly with use cases