Democratizing AI is not equal to Democratizing Equity.

Welcome to data uncollected, a newsletter designed to enable nonprofits to listen, think, reflect, and talk about data we missed and are yet to collect. In this newsletter, we will talk about everything the raw data is capable of – from simple strategies of building equity into research+analytics processes to how we can make a better community through purpose-driven analysis.

*********************************

Collective Affirmation: your learning about AI should be grounded in?you, your 'why', your mission, and never in fear of loss, competition, or replacement.

The first thing I learned about designed algorithms many years ago was – if you want to operationalize the algorithm (that is, if you want it to be fully automated code), you need to have a stack of bare-minimum 4-5 must-haves (infrastructure, data, systems, etc.).

?

I have lived, worked, and seen communities where this significant need for the minimums does not hold true. I am speaking of communities where wealth inequities create a whole lot of different kinds of needs and priorities than having access to the right infrastructure.

?

So, it makes me deeply uncomfortable when AI is pushed as a "apply now or become irrelevant" kind of existential-threatening message.

?

I want to live in a world with AI where every unique voice has a space to be included and heard – regardless of the differences in our identities, socioeconomic status, or C-suite work titles.

?

I understand the potential of this technology; I truly do. But that potential is meaningless if it corners humans and their humanity, forcing us to ignore inequities and inequitable outcomes of this technology in communities around the world.

?

No, I want to remind you (and I will do this repeatedly) that your learning and adoption of AI should be grounded in you, your 'why', your mission, and never in fear of loss, competition, or replacement.

?

Learn about the joy this (AI) can bring to your humanity.

With care.

*********************************

Democratizing AI means making artificial intelligence (AI) technologies, tools, and resources accessible to a broad spectrum of people, regardless of their background, education, or economic status.

?

While there is an intention of leveling the playing field by lowering barriers to entry and enabling a more diverse group of people to develop, use, and benefit from AI, the truth is that access alone does not inherently guarantee equal and equitable opportunities for all.

In this edition with you, I want to emphasize that our work and commitment towards diversity, equity, inclusion, belonging, and justice still holds very much true. In other words, having a login credential for ChatGPT does not guarantee that you and I are fairly represented and included on the important tables.

?

Let us talk about the?why?today, and in the next edition, we will look at what the tech companies amidst us can do about it.

?

Reason 1: access should not assume ability.

Providing access to AI technologies does not automatically empower everyone to use them effectively. Individuals and communities require resources to leverage AI. Without the necessary knowledge or resources, many people might not benefit from AI, even if it is technically available to them.

?

Reason 2: socioeconomic disparities

The benefits of democratized AI are often more accessible to those already equipped with the necessary infrastructure, including hardware, reliable internet access, and other technological resources. People with existing access barriers (ability, language, nationality, ethnicity, etc.) or those living in regions with poor infrastructure might not have the same opportunities to operationalize AI, thus widening the existing socioeconomic gaps.

?

For example, creating a social post from ChatGPT is not the same as setting up a full-fledged, documented, licensed business. Additionally, as AI advances, the cost of staying current with technology can be prohibitive for many, further entrenching disparities.

?

Reason 3: biases from data representation

AI systems are only as good as the data they are trained on, and biases in this data can lead to unequal outcomes. If the process of democratizing AI does not include efforts to address and mitigate these biases, the widespread adoption of AI could reinforce existing social and economic inequalities.

?

Furthermore, without careful evaluation frameworks, the possible continued underrepresentation of marginalized groups in AI development and decision-making processes can result in technologies that do not adequately reflect or serve the needs of all communities, exacerbating disparities.

?

Reason 4: dipping or lack of commitment to DEIJ (Diversity, Equity, Inclusion, and Justice)

The market dynamics of capitalism mean that even as AI is democratized, companies and individuals with more resources are better positioned to use these technologies for profit. This can lead to a concentration of wealth and power among those already advantaged rather than a broad-based distribution of benefits.

?

Another issue is that access to ChatGPT-like tools does not mean nonprofits serving historically marginalized and vulnerable populations do not have to go through loops and hoops for essential funding. Why? Because the people looking through funding proposals are still the same – with their lack of ethnic, racial, and all forms of diversity to broaden the scope from traditional decision-making/evaluation formulas.

?

Reason 5: replacing space for real listening?

When GPTs were newly launched, someone wrote on the internet, "I can now make a GPT based on books of Black authors." The post went viral, with likes, hearts, fire, and claps, with a common theme in the comments that took this as a win for DEI.

Two issues with that:

one, AI technologies often reflect their creators' cultural and linguistic biases, predominantly from affluent, English-speaking countries. This can lead to a lack of relevant perspectives from non-English speakers or for cultures that are not well-represented in the creators industry.

?

two, replacing direct conversations, listening, and learning initiatives from the people of a certain community with a summarized para is not a win for diversity and inclusion. The intention of inclusion is centering the voice of the people from the respective communities?directly?– without paraphrasing, summarizing, or translating – but making space for real listening.

?

*********************************

To truly democratize AI in a way that offers equal and equitable opportunities for all, especially historically marginalized communities, there must be a concerted effort to center DEI principles into every aspect of AI development and deployment.

?

This means actively working to diversify tech spaces?and?non-tech spaces (tech does not operate alone), creating equitable pathways for people to gain AI literacy and skills, and ensuring inclusive decision-making processes that listen to and prioritize the needs of those often left at the outliers.

?

Democratization of AI must be about more than just access.

?

It must be about ensuring fairness, justice, and inclusivity of communities. AI has capabilities for transformative change, but unless we do it with care, transparency, and accountability, all we will end up with is perpetuating the status quo.

***?So, what do I want from you today (my readers)?

• How do you think democratizing AI can be made meaningful for all?