Difference between revisions of "A Global Network of Personal Knowledge Container"

From PKC
Jump to navigation Jump to search
(Created page with "=Context= Given the imminent waves of products and services based on the late-breaking Artificial Intelligence Generated Content(AIGC) technologies, the goal of protecting personal and organizational data sovereignty is becoming ever more challenging. To address this issue, I propose the following approaches: =Goal= Provide functionally equivalent data processing, machine learning, and social/smart contract verification and execution capabilities to all members in socie...")
 
Line 23: Line 23:
#Create a self-governing community of [[PKC]] operators, that shares and exchanges data and knowledge assets.
#Create a self-governing community of [[PKC]] operators, that shares and exchanges data and knowledge assets.
#Operate regular events to promote best-practices, and late-breaking approaches to better protect data sovereignty while uphold social justice.
#Operate regular events to promote best-practices, and late-breaking approaches to better protect data sovereignty while uphold social justice.
#
#Initially, raw data collection, data cleasing, and [[LLM]] creation and certification using [[PKC]] will be the main services and products.
Because LLM is the "engine" for Generative Pre-trained Transformers (GPTs), which are evolving into a pervasive human-machine interface to a wide range of business and human activities, having both decentralized and highly aggregated computing resources to share a common LLM creation process is critical. While the pooling of computer resources assists in increasing efficiency and broadening the capabilities and performance of trained LLMs, the decentralized component aids in limiting the bias of majority judgments. From a social and economic standpoint, LLM is becoming into a new kind of asset that requires regular updating with a tight data processing technique, including data entry and automated data accountability policy compliance verification. Using locally adapted LLMs will invariably promote local information density, safeguard localized economic and political interests, and, most importantly, protect information/knowledge sovereignty. This is critical in terms of dispersed social fairness. (No one should let a centralized authority tell them what and how to think.)
=Boundary Conditions=
 
#The proposal assumes that the existing blockchain and data center infrastructures can handle the demands of large-scale machine learning computation and data processing required for the proposed service, which may not be the case.
The data collecting and filtering controlling models should be implemented as "Smart Contracts," which are available in high-level source code for both human and machine inspection. Large Language Models and raw data assets will be managed as independent objects to be published, acquired, or sold to parties of interest in line with "Smart Contract-Bound" principles. Existing Blockchain infrastructures, like as Bitcoin, Ethereum, Solana, and Polygon, can manage payment and awards for data and knowledge contributions (by writing code or evaluating "smart contracts"), as well as payment implementation.
#The proposal does not address potential security and privacy concerns around the use of blockchain and Smart Contracts in the management of sensitive data assets.
 
#The proposal also assumes that there is a demand for custom-made or localized Large Language Models, and that organizations and individuals are willing to pay for them, which may not be the case.
A public event in June 2023 might expose to the general public that blockchain mining can now be switched to continually execute Machine Learning Computation on the material given by anybody with a Distributed Identity (DiD) account ready to accept the duties stated by Smart-Contracts. As a result, a workflow and a global marketplace for material exchange will be developed, and computer resources will be directed toward the productive development of new and evolving knowledge. It will also be advantageous to use the large-scale public event to teach the audience about a brand-new industry, allowing for further examination of the combined opportunities given by Web3 and AIGC.
 
- edited by Grammarly and Quillbot
- edited by Grammarly and Quillbot


Revision as of 23:52, 16 February 2023

Context

Given the imminent waves of products and services based on the late-breaking Artificial Intelligence Generated Content(AIGC) technologies, the goal of protecting personal and organizational data sovereignty is becoming ever more challenging. To address this issue, I propose the following approaches:

Goal

Provide functionally equivalent data processing, machine learning, and social/smart contract verification and execution capabilities to all members in society.

Success Criteria

Allowing individuals and organizations of all levels of privileges to share a common data exchange protocol that enables fair and freewill participation.

Inputs

  1. Provide a personalized cloud solution, namely Personal Knowledge Container(PKC), which enables individuals and organizations to scale their data solutions from personal computing devices to high-end data centers.
  2. It is necessary to enable many instances of independently administrated, and geographically distributed PKCs, so that data and localized knowledge can be organized and curated based on individual and distributed wills. In other words, large-scale citizen participation across diverse locations and cultures is highly desired.
  3. Use of publicly known cryptographic data protection protocols and identity-protecting Smart Contract execution and Blockchain-based transaction services could help set up the initial platform. See Zenroom and The EU Interfacer Project.
  4. Provide public computing resources to filter and train publicly submitted data content to create diverse and highly functional Large Language Models(LLMs).
  5. Create a knowledge asset brokerage institution focused on creating, filtering, sharing, and buying/selling LLMs to businesses and individuals who require locally tailored or customized LLMs. Allowing small firms and individuals to participate in the design and composition of LLMs fosters system variety and a broader user base, which encourages freedom of expression.

Activities

  1. Develop and collect existing technical solutions to build the PKC and data protection infrastructures.
  2. Organize educational and promotional events to make these solutions and potential opportunities/threats known to a large number of people
  3. Encourage exchanges and interconnections between different PKCs, forming a larger computing network that becomes robust in a decentralized manner.
  4. Identify productive and rewarding data resources, algorithms, scientific concepts, and governance models to promote the sustainability of this global network.

Outputs

  1. Create a self-governing community of PKC operators, that shares and exchanges data and knowledge assets.
  2. Operate regular events to promote best-practices, and late-breaking approaches to better protect data sovereignty while uphold social justice.
  3. Initially, raw data collection, data cleasing, and LLM creation and certification using PKC will be the main services and products.

Boundary Conditions

  1. The proposal assumes that the existing blockchain and data center infrastructures can handle the demands of large-scale machine learning computation and data processing required for the proposed service, which may not be the case.
  2. The proposal does not address potential security and privacy concerns around the use of blockchain and Smart Contracts in the management of sensitive data assets.
  3. The proposal also assumes that there is a demand for custom-made or localized Large Language Models, and that organizations and individuals are willing to pay for them, which may not be the case.

- edited by Grammarly and Quillbot

Related thoughts: A Global Network of Personal Knowledge Container

References


Related Pages


Retrieved from "https://pkc.pub/index.php?title=A_Global_Network_of_Personal_Knowledge_Container&oldid=38322"