Verdant’s founder Brian Dolan recently keynoted DataCon LA where he spoke about data strategies for humanitarian applications of AI. Brian knows a few things about the topic since Verdant itself is focused on solving the humanitarian issues of today by addressing the UN’s Sustainable Development goals. This involves working on technology solutions including Health-tech, Ag-tech, and a marketplace for addiction recovery services.
The key to building a company that is addressing the UN’s SDG’s is to think top-down and identify the big problems we want to solve.
Brian Dolan
For too long, we’ve been building commercial AI products that people can’t stand., Chatbots are one example that comes to mind. But the technology’s true potential is also being explored by academics and hospitals in areas such as early cancer screenings and clinical trial matching.
Many researchers and data scientists are in fact like Brian, dreamers who dream of using AI to build a better world. It starts with addressing the hard problems. The environment is one such challenge, and many issues we face, such as climate change, rising seas, and degradation of soil are addressable through technology.
There are key differences, however between Commercial AI and Humanitarian AI.
- In commercial enterprises, we find focused stakeholders in data-rich environments with people who are empowered decision-makers, with the power to move the needle on their projects.
- Humanitarian AI, on the other hand, may have hundreds of disparate sources who are not aligned in their goals and who deal with issues from multiple angles and without a key decision-maker or policy by committee. In these environments, we are usually dealing with summary statistics vs detailed internal data.
It makes the challenges around humanitarian AI that much more difficult to overcome.
Brian’s aha moment in Humanitarian AI came when he approached it as a human effort. Contrary to the belief that you must focus on the solution, Brian found two key approaches that work best for humanitarian AI efforts:
- Ignore the problem. Instead, model the stakeholders and the ecosystem first. Examine how they can affect each other, determine what information goes back and forth, and how the stakeholders come to a decision, and select amongst the options.
- Identify the datahub. The datahub is the central point of communication between the stakeholders, the key place where data is exchanged, which can be a person or a place.
Verdant now uses this approach to evaluate humanitarian issues in search of solutions in a variety of fields.
Humanitarian AI for Agriculture
In the Ag-tech space, we are building a marketplace for biomass. Every year farmers throw away thousands of tons of agricultural waste, which could be turned into electricity and other products. It’s estimated that this wasted biomass represents $25 billion that could be going back into farmers’ pockets. While a “Craigslist for biomass” has been attempted, it is too simple of a solution that ignores the analysis that needs to happen.
Biomass is a complicated substance with complex thermochemical properties that need to be communicated to the labs that serve as a go-between in exchanges. In looking at the ecosystem, we found that the lab is the key place where data is exchanged. In building products to address the biomass supply chain, we have focused on labs as the central point of focus and developed our UX to be tailored to their needs.
Humanitarian AI for Recovery Centers
Drug addiction impacts 21 million Americans today and is the 9th leading cause of death. This is not for lack of available recovery options, as there are plenty of beds and treatment centers to treat these patients. The problem is matching patients to the right treatment facility, which is time-consuming. Most addicts in crisis can’t afford to wait. We identified 4 key stakeholders in the process:
- Addicts
- Insurance companies
- Patient advocates
- Recovery facilities
It’s the advocate who sits with the patient at the time they are trying to select the right treatment service. This person must navigate care options as well as understand the complicated psychology of an addict, decode insurance reimbursement options, comb through specific recovery center offers, and more. The patient advocate is a key data piece, or “datahub”. The technology we are building in partnership with our startup studio, True Match Recovery Index will empower advocates to make decisions faster and create liquidity in the market.
The key is to think holistically, which is where modern AI started, with Norbert Wiener who developed analytical processes and invented methods that focused on systems thinking such as stochastic control and game theory.
Here at Verdant, we believe that starting at the top down and looking at the system as a whole is how AI for Humanitarian causes can scale and make a positive Earth impact.
To watch Brian’s full keynote at Datacon LA, click here!
