This article is derived from the theme sharing and Q&A of Xiaoyu Smart Media on August 22
Sharing guests:
Ming Guo, the founder of Soteria, a Silicon Valley technical expert, worked as a software engineer in Silicon Valley for 20 years, and was a rocket expert. Founded Make Mountain Accelerator, using Silicon Valley-Shenzhen resources to help smart hardware startups around the world.
Ming Guo, a Silicon Valley blockchain technology expert, deeply analyzed the three major reasons why the blockchain is "difficult to implement" from the technical and social aspects in the last theme sharing, and explained that "big data" and "privacy computing" will bring Come to the blue ocean application:
"Big data and privacy computing are the blue ocean applications of the next generation blockchain"
https://www.odaily.com/post/preview/5147441
In the interactive questioning session after the theme sharing, many friends in the community raised a lot of in-depth questions about the shared content. Due to time constraints, most of the questions could not be answered immediately. and answer. Soteria will continue to explain the details of various aspects to you from different aspects in the articles and topic sharing that will be published immediately, so please pay attention. The following are selected Q&A sessions:
Q1: What are the current practical applications of this theory (referring to privacy computing)?
A: This theory can show its talents in the field of smart home at present. For example, "cross-domain" "privacy computing" can be realized, especially related to health data. For example, through "privacy computing", your ordering app can pass "privacy computing" on your health data while protecting privacy. Provide more targeted services.
Q2: If giant data is not compatible with the current Internet, what kind of new pattern will be available for human beings in the future?
A: The "new pattern" in the future is a new type of economy built on a generalized decentralized network -- SSDE -- an endogenous decentralized economy. Her technical foundation includes "big data" and "privacy computing". ".
Q3: How do you protect personal privacy? We use WeChat and Alipay's products, and the privacy is all in the hands of WeChat and Alipay. Unless users don't use their products, how can there be any privacy? The key is that we have to use WeChat and Alipay. You see that you are also using WeChat for operations. Do you think you have privacy on WeChat? Who said that personal data has no value on the centralized Internet? Doesn’t Taobao Jingdong rely on personal data to make user portrait recommendations? Is there no value generated in the process of recommendation? Now that you know that the current economic model and economic foundation cannot be used, how can you change the economic model and economic foundation to make your blockchain work?
A: In the digital world, your "privacy right" is "ownership", so the "private" data that has been handed over is no longer yours. In the current centralized Internet architecture, there is no right to personal privacy. Therefore, it is wrong to say that "privacy is in certain Internet applications". Data that is no longer in your hands (in the data centers of Internet application giants) is not your data, nor is it your "privacy" — Because of the data you handed over, your "privacy" and "ownership" are lost, you no longer own those data, and you no longer have "privacy".
"Personal data has no value in the centralized Internet" means that the value is captured by the Internet giants and has no value for you (users), because users "contribute" value to the Internet giants for free and "voluntarily", and the value does not belong to user. By value we mean "economic value", that is, do the Internet giants who use your data pay you "economic incentives"? No! So we say that there is neither privacy protection nor economic incentives for the use of "privacy" in the current centralized Internet architecture, so privacy is worthless.
In this highly centralized and monopolized Internet, the blockchain cannot provide real "privacy protection", because there is no living space for business models that can make "privacy" valuable. Then, to achieve true "privacy protection", it is necessary to find an economic model/economy that gives value to "privacy" (that is, to "privacy owners"—users, not to "privacy data grabbers") entities, and the application scenarios associated with this new economic model, these application scenarios should be temporarily unable to intervene by the current monopoly Internet giants (due to technical and economic reasons), we believe that this new economic model is SSDE - Endogenous Decentralization In the globalized economy, this new application scenario is "big data + privacy computing".
This new decentralized economy and the current centralized economy will coexist for a while, but in the long run the traditional Internet business model will die. This process is like the beginning of the industrial revolution. The new industrialized economic model coexisted with the traditional small-scale peasant economy for a period of time. The industrial revolution will carry out many economic activities that do not exist in the small-scale peasant economy. In the end, the small-scale peasant economy will disappear and the application of the small-scale peasant economy Scenario—Traditional agriculture is replaced by modern industrial agriculture.
Q4: The keynote speaker said it very well. In order for data to be traded, the right must be confirmed first, and the transaction can only be done under the premise of solving privacy, but I don’t understand the difference between private computing and encrypted computing?
A: The difference between "private computing" and "encrypted computing": the main thing is that "private computing" is "calculated" by the data owner and the demander together. In the process of "calculation", the "demander" It does not have access to the data of the "owner", but only to the "calculation results". However, there is only one party in "encrypted computing", that is, the database owner. They have already obtained the original data, so in fact, "encrypted computing" does not have "privacy protection".
Q5: What is the difference between you and platON?
A: This question is about the specific technical path, and I can continue to discuss it in the follow-up sharing. To put it simply, my understanding is that Dr. Xiao and the others are in a very cutting-edge technical direction, that is, "generally secure multi-party computing", and the engineering challenges are relatively large. We are currently using synthetic solutions such as "privacy protection transactions" and "light chain smart contracts" that are closely integrated with the blockchain, especially solutions that combine new economic incentive paradigms to solve the problem of privacy protection.
Q6: What are the development stages of the SSDE-endogenous decentralized economy, that is, what will be the specific implementation process?
A: SSDE — the problem of endogenous decentralized economies, we will discuss and expand in depth in the next live sharing.
Q7: I don’t know why I suddenly remembered the App application measures related to privacy protection when I watched the iOS13 release video. I feel that what you are talking about is the technology widely used in iOS13 that Apple will release this fall. In the future, when Apple's App needs to use your private information to help you make decisions, your information will remain on your mobile phone, and Apple will only provide the App and the algorithm. Your personal information is not circulated on the Internet (encrypted), and Apple does not have your private information. In this way, there will be no privacy issues, and what Google has done for many years is that any information it can get about you is stored in its system (encrypted). He can use these data to provide better services, but the worry is Who will ensure that your privacy is not used for other purposes. If you think about it this way, you can understand what a decentralized talk is. iOS13 and iPadOS will be released in September, and the app for iPadOS is also the same concept. May I ask if my understanding is correct?
A: Apple's new generation of iOS devices (such as the iPhone) does have a hardware implementation of the "Data Enclave" that protects privacy "locally"; the private data protected in the "Data Enclave", such as fingerprints, It will not be uploaded to Apple's data center; the demand for fingerprints is realized through "local" "verification", which is indeed related to the "privacy computing" based on decentralized "data sovereignty" we advocate Where it resonates. However, the "private data" protected in Apple's "data chamber" is mainly static "identity data", such as identity information and fingerprints, etc. It cannot protect other "private data", especially "behavioral privacy data".
Q8: Everything can be calculated, and calculation can be private. What about the data security of smart hardware?
A: My understanding is that the main problem of current privacy computing is not only a technical problem, but a problem of economic model and economic foundation. In our current centralized Internet and resource-bound credit economy, privacy computing Technology cannot be used. The data security of smart hardware is actually the issue of "data sovereignty" I mentioned above, that is, the data security of smart hardware is that these data are never shared with the outside world and are only used for "private computing".
Q9: I don't quite understand, like bank data will never be out of the database? So how to build a value evaluation model? After all, not everyone's data is valuable to all demanders. The robo-advisors that have been called for many years in China have all been reduced to gray marketing tools.
A: In fact, we need a new type of economy. For example, blockchain can help us realize "decentralized" "finance". Then there is no so-called "bank data". Privacy computing" can be realized.
Q10: A person's data sovereignty is a prerequisite for data monetization, but having data sovereignty requires the ability to control their own data, but many services require us to open personal data to use services. How can this be broken? Even if the new generation of human beings can enjoy data sovereignty, the data that has been made public in our generation can’t make decisions for themselves, right? The data security of smart home will not be particularly high, breaking through a router may break all the data...
A: "You need to open personal data to use services" - this is a typical business model of the current monopoly and centralized Internet. This business model itself is contrary to and incompatible with "data sovereignty", so it can be "realized" The "data sovereignty" of "data sovereignty" cannot adopt the current centralized Internet business model, but needs to introduce a new type of endogenous decentralized economy—SSDE, in which "privacy computing" in the "big data" application scenario is currently available The earliest SSDE business model.
After the endogenous decentralized economy develops, the current highly monopolized dopamine drug-centered Internet business model that relies on privacy violations to capture value will collapse and die. In this process, the next generation of more stringent privacy protection GDPR may provide assistance, but in the end, users still need to vote with their feet, provided that users need to be provided with new service models and scenarios that can protect "data sovereignty", such as "big data". "Privacy computing" in application scenarios.
"The data security of smart homes will not be particularly high, and breaking a router may break all the data"-Under the current technical architecture of the centralized Internet, this problem is unsolvable. However, in the decentralized network in the SSDE/"data sovereignty" era, the solution is to "drain the fire from the bottom of the pot" and completely remove the "data". In the network, "data" is a black hole. In this kind of network, you can't "see" any "data" at all, you can only see "computation" (privacy computing), usually various smart contracts that "package" "privacy computing". These smart contracts realize the "data sovereignty" you have.
Q11: Data is indeed not used for trading, which is acceptable. However, without data as a basis, artificial intelligence will lose its foundation. How to consider this issue?
A: Indeed, after the realization of "data sovereignty", the current centralized Internet giants will not have enough data to train their "artificial intelligence". But this does not affect the development of artificial intelligence at all. In the era of "data sovereignty", the "computation" for training "artificial intelligence" will be done locally, not in the data centers of Internet giants. This is also advocated by the so-called "edge computing" scenario story, because the amount of data that can be used for artificial intelligence training far exceeds that of the traditional "big data" centralized technology architecture. Processing power, this is what we call "big data" - and this "big data" comes from sensor data that is closer to people. Users have a stronger demand for privacy protection for this type of data, and they may not mind the data collected by the traditional Internet. Privacy "big data" (such as data on users' news browsing or online shopping habits), but they will be very concerned about the privacy protection of "big data", such as health data and home life privacy data.
In fact, the current "big data" collected by the centralized Internet is a kind of "aggregated data". Through "machine learning" on this type of data, the "dopamine drug" business model of Internet giants can be optimized and perfected, but the "artificial intelligence" There is no help for the advancement of this kind of "big data" and the "behavioral" data of "smart monomers" are too far apart. What can really promote the development of "artificial intelligence" is actually the "training" calculation performed on "huge data" close to the "continuous behavior" of "intelligent monomer", so the real artificial intelligence training can only be performed locally This is also an application scenario of "big data" in the era of "data sovereignty" - "realization" by training "artificial intelligence" "privacy computing" smart contracts.
Q12: The teacher said about the new economy, I want to know how the old and new economies can coexist?
A: (answered earlier) This new decentralized economy and the current centralized economy will coexist for a period of time, but in the long run, the traditional Internet business model will die. This process is like the beginning of the industrial revolution. The new industrialized economic model coexisted with the traditional small-scale peasant economy for a period of time. The industrial revolution will carry out many economic activities that do not exist in the small-scale peasant economy. In the end, the small-scale peasant economy will disappear and the application of the small-scale peasant economy Scenario—Traditional agriculture is replaced by modern industrial agriculture.
Q13: I would like to ask you, personal data, privacy protection, privacy computing...these are all inseparable from personal identity authentication, and the current various centralized identity authentication cannot be opened up globally. My understanding is that the establishment of a decentralized economic system and For development, decentralized identity authentication must be solved. This is a basic requirement, and all kinds of decentralized applications need it. How did you solve it? Or which one has a feasible solution to this problem?
Qunyou Claire: The identity of the decentralized system should be your private key, right? It should be different from the centralized system.
Qunyou Jialin: The identity information is multi-dimensional, and the identity ID should be unique in a system, and the private key alone is not enough. Then everyone has a lot of private keys, how to confirm the right? Identity, confirmation of rights, contribution, income...
Group Friends, Great Country, Great Era: This kind of thinking is still stuck in the present.
Qunyou Jialin: Please explain in detail how decentralized identities are verified and operated in a decentralized economic system? Or what about your artificial decentralized identity? Or it is not needed.
Group friends, big country, big era: In the future, the world will be dualistic, and ID cards will not be used in decentralized places. Machine learning, it is useless for you to hold your ID card, and it can only be used in traditional centralized scenarios. Resolving this contradiction is a false proposition, people will not take you to play at all.
Qunyou Jialin: I am not talking about a centralized (identity authentication) system... According to what you said, there is no need for privacy protection in a decentralized system? What about personal data? Doesn't zero-knowledge prove that this kind of technology solves the privacy and confirmation of personal data? Personal information, literary and artistic creation, copyright... How do these work in the decentralized system?
Group friends, big country, big era: First of all, most decentralized systems are open source, and the data can be read by anyone. There is no privacy at all. Just like Bitcoin, every transaction is public.
Second, the concept of privacy is established for a specific talent, you have your data, I have my data, and my data does not want you to see the concept of privacy.
Third, the digital world only believes in mathematics, and there is no premise that it must be bound to the physical world. We do not need to bind any information in reality to control a data account, which also isolates everyone's identity information.
Fourth, the premise of commercial activities is based on trust. In many cases, you only need to know that you can get the money without knowing who I am. But the current situation is that you not only know who I am, but also know what I like to eat and drink. Well, who did you have a date with yesterday? This is obviously too much.
Fifth, zero-knowledge proof is in the field of cryptography, not unique to the blockchain. It can isolate a person's original data from calculations, so that your privacy can be protected to the greatest extent.
Sixth, the blockchain can only solve the trust problem in the digital world, but cannot solve the trust problem in the physical world. Problems in the physical world can use the blockchain as an auxiliary solution. In the real world, you should rely on the law to solve it. Find the police to find the police to solve, the blockchain is not a panacea.
Ming Guo (speaker): @Jialin The current popular concepts have many misunderstandings about the relationship between the concepts of "identity", "trust", "right confirmation" and "decentralization". What I'm going to say below may be more thrilling/shocking, and I will have a special article to detail it in the future. Simply put, I think that the "decentralized" world does not need "identity", "trust", and "right confirmation". Therefore, "decentralized identity", "decentralized trust", and "decentralized right confirmation" are all false propositions -- oxymoron.
In the decentralized world, there is no "identity" corresponding to "person" in our centralized world; every "transaction" and "interaction" you have in every "decentralized" network can be There is a different, separate ID/address and corresponding "private key" -- in a decentralized world you can have an infinite number of clones. In fact, "identity" in our centralized world is a means/weapon/device of "resource binding", a kind of oppression/coercion, and a kind of original sin. A decentralized world does not need a centralized "identity".
trust"trust"One of the characteristics of Bitcoin is trustless trust -- "trust" without trust. The latter "trust" is actually just rhetoric, and the essence is "trustless". Some people say that you believe in "mathematics" and "hash" -- this is not true, because whether you "trust" mathematics and hashes or not, you cannot fight against mining machines, so it is "no trust" -- really go The centralized network must be "trustless", and "no trust" also leads to the need for "identity", so you can join and withdraw from the decentralized network at any time, and the decentralized network does not require identity verification, nor does it, nor does it Rely on any "login restrictions"/access control.
In the same way, the decentralized world has no way to "confirm" the "identity" in the centralized world. You can use cryptography or blockchain to "witness" a "discussion", but you cannot "identify" the identity in the real world. The "identity" projected to the "discussion" in the decentralized world is bound, which means that the "truth" in the decentralized virtual space cannot be "owned" (that is, "identity" and "discussion" cannot be bound) , the binding of "identity" and "discourse" can only be achieved by introducing a centralized coercive system/coercive mechanism.
Group friend Jialin: I thought about this before, but it was a bit vague at the time, but it is much clearer now, thanks for the explanation. At the same time, I also understand + agree with this concept.
Ming Guo: @Jialin These questions are very important, because the "decentralized" world is something we were not familiar with before. I have a series dedicated to the philosophical issues of these "brain-burning" "decentralized worlds". Speaking of "identity", you asked a question here first, and I will give a simple reply): Related links
Qunyou Jialin: At present, there are many blockchain identity authentication projects, and there are also many other projects that hope to connect or connect the physical world and the digital world. After all, the physical world has a long history of development, and the digital world is still too young. Let's see the development. I also hope that the decentralized digital world will develop better and bring more freedom and rights to individuals.
Ming Guo: @Jialin still needs a lot of work to connect the physical world and the digital world, such as artificial intelligence...
(full text)
(full text)
