Alivever.com - Projects

Project 1: AI Data Services
2024/Jan/1
This project aims to design and provide a set of universal data services and retrieval service interfaces, allowing any cloud-based AI training frameworks and AI programs to conveniently invoke these data services. This significantly simplifies the complex data preparation involved in AI training. AI researchers will no longer need to manually search for and download datasets, manually process data, and finally input the data into AI in command-line form. They simply need to directly utilize Alivever's data services and retrieval services to immediately gain powerful data capabilities. This effort involves aggregating, organizing, categorizing, deduplicating, denoising, authenticating, anonymizing, and indexing globally renowned public datasets. Subsequently, a universal data structure and cloud service interface will be designed. Ultimately, this will provide services for cloud-based AI systems in the form of iterators, search engines, and recommendation engines. This service will help AI companies significantly reduce their R&D and operational costs. Typically, AI companies require extensive downloads of public datasets and complex data processing before they can commence AI training. This consumes a significant portion of AGI companies' budgets and time, sometimes up to 30%. However, the "alivever" data service project can directly help AGI companies reduce the costs of collecting and organizing datasets to almost zero. Additionally, combining data search engine services with LLM (Large Language Models) helps reduce the volume of large pre-training models. This not only significantly reduces LLM's daily operational dependence on GPUs but also further lowers the operational costs of AGI companies.

Project 2: My Story
2024/Jan/1
In this world, the vast majority of recorded texts encompass human civilization (such as history, religion, philosophy), human experiences (such as natural and social sciences), and prominent figures (such as kings and celebrities). However, there is very little content that documents the complete lives of each ordinary person. Yet, the true human civilization is actually composed of every individual. History and experiences are merely outcomes, not reasons. Although we already have social media platforms like Facebook and Instagram, their content remains highly fragmented, sometimes distorted merely for flaunting purposes. In reality, in Europe, many ordinary people have already written biographies of themselves and their families, or are preparing to do so. In Asia and other ancient civilizations, people record the deeds of their ancestors through oral traditions. However, with global aging and escalating conflicts, many individuals have no descendants, and even numerous large noble families have lost their heirs. As time passes, the memories, experiences, and feelings of more and more individuals will gradually disappear. This is an outcome we are unwilling to see and a problem we aim to address through this project. We aspire to solve this challenge through this project. We will attempt to find a method to objectively and comprehensively document the lives of every individual from a user-first perspective, preferably at a cost close to zero. We hope to assist every citizen in every country in recording their lives so that they and their descendants can revisit these memories and experiences at any time. We will actively assist any ordinary citizen in organizing their personal biographies. Biographies can take the form of text, historical photos, videos, and on-site interviews, among others. All personal biographies will strictly adhere to personal privacy and data security regulations. By default, personal biographies can only be accessed by their owners and their heirs. The content of a biography can only be publicly released with explicit written consent from its owner. Naturally, we will also actively collect all publicly published biographies and incorporate them into our database. We will comply with copyright laws and will not directly provide the text of biographies. Instead, we will offer services externally through an "AI data interface" in a format understandable by computers (non-human language).

Project 3: My Shadow
Not started yet
This project will, based on the "My Story" project, utilize AI to train the content of biographies and ultimately generate an AI virtual persona. All private biographies will strictly adhere to personal privacy and ethical standards in technology. 1. AI virtual personas will only be created with the consent and payment of the biography's owner. 2. AI virtual personas will only use data from the respective owner's biography. It is prohibited to use multiple different biographies' information to train a single AI. 3. AI virtual personas are strictly prohibited from accessing the internet and can only be accessed by the owner of the AI virtual persona and their successors. 4. AI virtual personas will only exist in software form and are prohibited from being integrated into any form of hardware robot body.