Democratizing ML and bridge the digital divide

I am thinking about this topic lately. In the last few years, as I have dived deeper in ML at an academic level, building ML models, classifiers, build, train and test models, I have one question: Are all these ML tools creating a deep digital divide? Are we democratizing the use and accessibility of these models?

Only the big tech companies who have access to this ML models, data and the tools to build the models, have leap frogged the rest of the world. As we build more powerful models, this divide is going to increase exponentially.

I am in the quest to find how to bridge this divide, by making these models easy to access and for everyone to use, irrespective of what they know about the underlying tech.

My north star is a Shopify experience (where anyone can launch a business), anyone can ask simple questions on common sense problems like product launches, social media and marketing campaigns, public policy, public health policy, school outcomes etc, and based on classifiers without knowing about the mechanics of Bayes, or Random Forest or Decision tree or Logistic Regression predict the human understandable outcome.

Let's build this.