The author suggests strategies for improving AI development, including creating a multiparty stakeholder group to protect different cultures and minor voices, and combining different AI-adjacent technologies to balance their strengths and weaknesses. The article also suggests that the AI industry may need to employ millions of new humans as analysts, subject-matter experts, and developers to help scale and regulate the technology.
Key takeaways:
- The development and evolution of AI technology is compared to raising a child, with the need for instilling values and understanding, but the current approach lacks a human touch and can lead to biases and inaccuracies.
- Context is crucial in AI programming, and cultural, regional, and individual differences can easily be overlooked, leading to skewed data and potential misuse.
- There are concerns about the concentration of power in a few tech companies that control AI and cloud technologies, and the need for a multiparty stakeholder group to ensure the protection of different cultures and minor voices.
- AI development will likely require the involvement of millions of new humans as analysts, subject-matter experts, and developers to ensure the technology's accuracy, fairness, and regulatory compliance.