Problem Statement

As AI technology is rapidly becoming an essential part of our lives, it is increasingly being centralized in the hands of a few large corporations. The current centralized nature of AI models introduces a "rich gets richer" cycle in which only companies with access to large, labeled datasets can benefit from AI. Even though AI companies like OpenAI are announcing for future monetization plans, it is yet to be a relief. The monetization plans seem to be operated in an opaque manner, eventually leading users to abide by the firms’ decisions.

Even more, there is no idea whether users’ personal and precious data sets will be kept safe. Truth to be told, we all may have probably once noticed news articles of tech giant firms misusing users’ data.

Besides the huge boom of artificial intelligence, sometimes, we also discover out-of-date or incorrect responses. As a result, ironically, we are currently facing chronic problems in the AI industry including privacy invasion, inaccuracy, and absence of data sovereignty.

1. Content Creators' Intellectual Property Infringement

Centralized AI platforms often rely on massive amounts of data, including content created by individuals (artists, writers, musicians, etc.) without their explicit consent. These include:

• Loss of property ownership: Scraping content from public and private domains to train models can result in creators losing control over how their work is used.

• Lack of transparency: how creators' intellectual property is integrated or replicated, leading to situations where AI-generated content mimics or directly uses parts of original works without proper attribution.

• Ethical issues: As creators might not receive recognition, compensation, or the ability to restrict usage of their creations, undermining their ownership.

2. Opaque Profit Sharing

Centralized AI platforms generate significant revenue from the data and content they process, but often fail to fairly compensate those who contribute:

• Monetization without compensation: The creators of the data used for training are rarely compensated, even though their contributions are critical to the platform’s success.

• Platform-centric profit model: AI companies usually take the majority of profits, using the data from users, while offering minimal (if any) royalties or profit-sharing with the actual content creators.

3. Inaccurate Responses

Centralized AI models can provide inaccurate or misleading information due to:

• Training data limitations: AI systems depend heavily on the data they're trained on, which may be outdated, biased, or incomplete, leading to inaccurate results.

• Context misunderstanding: These platforms often fail to grasp the full context of user queries, resulting in surface-level or incorrect responses that don’t address the user’s specific needs.

• Lack of real-time verification: AIs cannot always verify information at the moment, meaning it might provide outdated or erroneous facts that can mislead users in decision-making.

4. Unreliable Responses

Responses from centralized AI platforms may be unreliable for several reasons:

• Bias and misinformation: These platforms can reflect the biases present in their training data, leading to unreliable or skewed responses that perpetuate misinformation.

• Opaque algorithms: The models operate as “black boxes,” meaning users often have no insight into how or why the AI arrived at a particular response, making it hard to trust the results.

• Generalization issues: Since centralized AI models are designed to serve a broad audience, they may oversimplify or generalize complex topics, leading to shallow and unreliable advice, particularly in niche or specialized areas.