Designing AI for Everyone: The Promise of Socially Validated Systems

Adrian Blackwood
3 min readApr 18, 2023

I’ve Spoken at three events in the last week about building products with Ai, and no one is talking (yet) about social validation, and how we can achieve it.

Today, I want to emphasize the criticality of socially validated AI. As AI technologies are increasingly integrated into our lives, it’s essential that they reflect our societal values and norms, as well as respect ethical and legal frameworks. Socially validated AI refers to systems that have been designed, trained, and evaluated with the involvement and feedback of diverse stakeholders, including the affected communities, domain experts, policymakers, and civil society organizations.

Who cares? We all will, in the near future.

As Ai grows we need a shared way to agree on Ai default norms and Ai behaviors.

The reasons why socially validated AI is so important are manifold. Firstly, it ensures that AI systems are more accurate, reliable, and trustworthy, as they are tested against a broader range of scenarios, perspectives, and biases. Socially validated AI also helps to reduce the risk of unintended consequences, such as discrimination, bias, or harm to individuals or society. By involving a diverse range of stakeholders, socially validated AI can surface potential risks and trade-offs that might not be apparent to the developers or users of the system.

Moreover, socially validated AI promotes transparency and accountability, which are crucial for building public trust and confidence in AI. By making the design and decision-making processes more visible and inclusive, socially validated AI enables users and stakeholders to understand and challenge the choices and assumptions behind the AI system. This can help to prevent the “black box” effect, where AI systems are opaque and unaccountable, and can lead to unfair or undesirable outcomes.

Activism isn’t enough, we need a business rationale to drive social validation.

OpenAi’s future architecture features prominent placeholders for Social Input, but lacks the systems, process and oversight to maintain, enforce and promote such a monumental shift in collective societal alignment.

First Area: inputs.

Sell more products with a “socially validated product’ label.

Product builders have a crucial role to play in ensuring that AI products and applications are socially validated. They must recognize their responsibility to design and code products that are inclusive, fair, and ethical. This requires a proactive effort to involve diverse stakeholders in the development process, from the ideation stage to the testing and deployment stages. Builders must also be aware of the potential biases and limitations of the data and algorithms used to train AI systems, and take steps to mitigate these risks. Additionally, builders should prioritize transparency and accountability in their products, ensuring that users can understand and control how AI systems make decisions. Ultimately, it’s up to product builders to create socially validated AI that is aligned with our collective values and aspirations, and that contributes to a more equitable and just society.

Second Area: Feedback.

The place for activism.

“AI for good” social organizations also have a significant impact in promoting socially validated AI and shaping public sentiment towards AI. These organizations are dedicated to leveraging AI for positive social impact, such as promoting sustainability, healthcare, education, and human rights. By demonstrating the potential of AI to solve complex social problems, these organizations can help to counterbalance the negative narratives and fears that often surround AI. They can also play a critical role in raising awareness of the importance of social validation and advocating for ethical and responsible AI practices. Through their advocacy and innovation, “AI for good” social organizations can help to build a more positive and inclusive vision of AI, which can encourage broader adoption of socially validated AI in all sectors of society.

No Conclusion, just a direction -> forward.

In summary, socially validated AI is essential for creating AI systems that are aligned with our values, goals, and needs as a society. It’s a collaborative and iterative process that involves ongoing engagement and feedback from a diverse range of stakeholders. By prioritizing social validation, we can build more trustworthy, transparent, and inclusive AI systems that can benefit everyone.

Adrian explores the future of reality through four exciting areas: Applied Ai, Spatial Reality, health-tech, and Ambient Intelligence. find out more at: adriankyleblackwood.com

--

--

Adrian Blackwood

Adrian explores the future of reality through four exciting areas: Applied Ai, Spatial Reality, health-tech, and Ambient Intelligence.