Artist, Curator, Retired Educator

Pausing AI. Particularly, globally pausing the race for super-intelligence, the building of data centers, and the training of these models on stolen IP, music, art, writings, and scientific research.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight
• Safe, secure and trustworthy AI

AI experts and surveys estimate the probability that AI will destroy humankind and the biosphere to be between 5% and 20%. AI has been trained on data that often favors white, first-world, and upper-class people. This cultural/social/racial bias results in unequal performance and discriminatory outcomes that permeate and harmfully affect the lives of real people, not to mention the loss of jobs and a sense of purpose. I would like to see ALL AI companies and research embed global ethical guidelines, such as UNESCO's principles, into the entire lifecycle of AI systems. Key actions include independent auditing, mandatory human rights impact assessments, human oversight to prevent discrimination, and regulatory frameworks like the EU AI Act or US Algorithmic Accountability Act.

Limiting data center growth should be used as a bottleneck for AI development. Simultaneously, these facilities pose significant environmental challenges especially in poor communities. They are making people sick. Additionally, data centers use massive amounts of energy, cause water depletion, and increase reliance on fossil fuels... passing the costs and shortages onto the human communities.

1. It appears as though local governance is distracted by the war(s) and political conflict. Non-republicans in the US, in particular, feel that they have no leverage in the current political climate. Additionally, they don't appear to sense the urgency. I personally know of situations where employees have been fired/let go only to be replaced by AI. I don't see opportunities here and I'm not sure if reskilling is realistic if AI takes on most entry-level and white-collar jobs. Without entry-level workers, you will cut off future generations of skilled workers. 2. Artists and creatives have their IP and works stolen/plagerized by AI training and products. AI reproduces copyrighted material from its training data, mimics specific artists' styles, or passes it off as original without references or disclosure. This is theft. I would like to see a slow down in AI development to put into place technology that watermarks, signatures, or cryptographically secures content to prove it was created by a human. Also, technology that asks permission and/or pays to train on creative content. 3. AI content often accelerates the spread of misinformation, commits intellectual property infringement, and perpetuates bias. AI models can also "hallucinate" confident falsehoods. The opportunities here could be: reducing hallucinations; businesses can hire developers to build systems that connect AI to authoritative, verified databases rather than letting it guess based on general training data. Also, there is a need for "Truth-Checking" Startups and AI Auditors.

Without international cooperation, humans will lose out.

The UN Global Dialogue on AI Governance (under Global Digital Compact) is a good place to start, creating a coalition of fragmented efforts and organizations. International Labor Organization (ILO), UNESCO AI Ethics, and Technical Standards Bodies must focus on actionable, inclusive, and legitimate international governance.

Artists, writers, musicians, minorities, women, and the poor.