My ideal AI+blockchain project (2)

Source: Daoshuo Blockchain

There is an article in "Golden Finance""A Quick Look at 11 Mainstream AI Projects" lists and briefly introduces 11 currently popular crypto ecosystems. AI related projects.

I probably read the official introductions of these projects. I have selected some of the projects that are introduced in more detail and share with you my views on these projects.

- Grass: https://app.getgrass.io/

AI needs to be used A large amount of data and resources are collected. The wider the scope of collecting these data and resources, the more beneficial it is to improve the performance of AI.

So the focus of this project is to make use of the network resources (Internet connections) held by netizens around the world. These netizens can rent their idle Internet resources to AI, so that AI can collect global network resources and data from different regions and from different angles.

In this process, since AI will use the personal Internet connection of netizens, it will inevitably involve the private information of netizens stored on the client. Therefore, the project places special emphasis on protecting personal privacy.

This project immediately made me think of projects such as decentralized storage.

Projects such as decentralized storage also focused on "using idle resources in individuals' hands" at the beginning ----- storage resources, but later developed to basically Users who participate in mining on the Internet rarely use their "idle" resources to mine. The vast majority of them use "newly purchased" resources to participate in mining.

Since then, I have doubts about any project that claims to "use idle resources" to perform XX operations: Will it develop into using proprietary resources? The device participates in XX operations.

In this project, will users avoid using the device they are currently using to protect their privacy stored locally, and ultimately purchase a new device? What about participating in mining?

Currently, this project is actively encouraging users to participate in mining. Judging from the feedback on the Internet, it seems that most users use the same equipment they are currently using. Everyone cares more about the amount of mining revenue rather than whether this project will infringe on their privacy.

But whether there will be changes in the future, we have to continue to observe. If it becomes similar to decentralized storage, and everyone uses new equipment to participate in mining and starts to rent a private network or computer room with particularly good network conditions, then not only will it deviate from the original intention of the project team in terms of "utilizing idle resources" , and will also weaken the effect of decentralization to a certain extent.

In addition, the main emphasis in the project introduction is to use resources and collect data in a decentralized manner, but after these resources and data are used, How will the results be processed?

The project party does not seem to have talked about it.

Judging from the existing technical means, using the traditional centralized method to operate is technically the most mature and convenient. However, there are currently no successful and large-scale technical solutions for decentralized and efficient data processing.

If the operation is still carried out in the traditional way, is there any possibility that the user's network data will be abused? After all, the bad records of Internet giants make me have to be wary of this.

Overall, in my opinion, this project belongs to the general direction of using blockchain technology to partially solve the problem of computing power or data collection in AI. I am absolutely sure and agree.

However, the description of many issues in the information currently released by the project team is still somewhat vague, so there are still some issues worth noting.

Reference link:

"Quick Overview of 11 Mainstream Encrypted AI Projects":

https://www.jinse.cn/blockchain/3677551.html