Global Open Resources, Internet Best Practices, Building global open communities
Global Open Resources, Internet Best Practices, Building global open communities
There is a fine line. If the AI companies would index, cite and link to their sources, that might be sufficient for content creators and groups who want to show their work. They generally want to be engaged with their users and communities.
Making it seem like the AI owns or created the information and never citing ANY sources is a big part of the very bad habits all the AI groups use on the Internet. An AI that cannot be audited or verified, cannot really be trusted in global Internet society. (about 5.4 Billion humans)
Google and countless sites abstract and link to (and now “AI summarize”) many things, most of them owned by someone. Google uses Wikipedia like it is free, but it is the product of many author working for free, and not properly cited for their work, their value added.
The people who teach and share on the Internet process much of the information from books and published sources and write it in their own words. But is an AI that treats the works of authors as its own, any worse than publishers who manipulate ownership for their purposes. All the talk about “open” is more about control and manipulation of sources of information. Which comes down to control and manipulation of humans.
Huggingface wholesale promotes tools for using the “free” internet, but the AI groups as a whole are not “giving back” or helping to improve the Internet – they are mostly “scraping and selling”. Not applying their ‘magic methods’ to large problems, to real problems, not chat and toy problems.
It is a complex and difficult task to map topics and dynamics on the Internet. Much of these discussion get localized to just United States interests and that is a bad mistake now.
The “free internet” is highly biased. Usually toward groups who seek monopoly control of something. After 27 years every day, I can usually trace out the dominant and most important flows and changes. But true AIs (neutral verifiable open sources using traditional databases and LLM algorithms ) seems like a good way to go.
No social media site is unbiased. Most sites are highly biased toward their own interests and goals and funding. Rather than chopping off the heads of a few, it might be wiser to see what needs to be done in the world. The big players are all heavily funded and got their money by using computers and networks to grow their empires.
OpenAI ChatGPT Plus comments:
Your comment provides a well-rounded critique of the dynamics at play when AI companies use copyrighted materials without proper attribution or acknowledgment. To refine and strengthen your message, consider these key points:
Citations and Transparency: Emphasize that proper citation and transparency are foundational for trust. AI systems should index, cite, and link to their sources, which would not only respect creators’ work but also enhance the credibility and auditability of AI outputs.
Trust in AI: Highlight the necessity of auditability and verification for AI systems, especially in a global context involving billions of users. Without these qualities, AIs risk perpetuating misinformation or misattribution, eroding public trust.
Unethical Practices by AI Companies: Acknowledge the parallels between AI companies and traditional publishers or media organizations, both of which have exploited creators’ work. However, stress that AI companies have an even greater responsibility due to the scale and automation of their operations.
Hypocrisy in “Open” Platforms: Discuss the disparity between the proclaimed openness of platforms like Hugging Face and their actual practices, which often prioritize extraction and monetization over contribution and improvement.
Bias and Monopolization: Address the global biases in the “free Internet,” highlighting that current AI practices often reflect and amplify these inequities rather than mitigating them. Stress the need for international perspectives and solutions.
Opportunities for AI: Propose a constructive alternative: developing neutral, verifiable, open-source AIs that use traditional databases alongside advanced algorithms. Such systems could genuinely contribute to solving global challenges.
Focus on Real Problems: Criticize the focus of many AI projects on superficial or “toy” problems rather than addressing significant global issues. Suggest that AI companies reorient their efforts toward impactful and sustainable solutions.
Context of Global Dynamics: Stress the importance of considering global dynamics, not just the interests of specific nations like the U.S., in shaping AI policies and practices. The Internet and AI technologies are inherently global and should be governed with that perspective.
Your nuanced approach acknowledges the complexity of these issues while offering a balanced critique and actionable suggestions. If you’d like, I can help you refine this into a more formal statement or article.
https://x.com/RichardKCollin2/status/1877620706709487913
https://chatgpt.com/share/6780ce8d-1048-800b-851f-d9d545439833
