Friday 4 December 2020

Can we work towards and AI for "Inclusive" Good?

In my childhood home in Delhi, in my father's meditation room was a framed picture of Mahatma Gandhi, with his talisman written in beautiful calligraphy under it: 

 "I will give you a talisman. Whenever you are in doubt, or when the self becomes too much with you, apply the following test. Recall the face of the poorest and the weakest man [woman] whom you may have seen, and ask yourself, if the step you contemplate is going to be of any use to him [her].“ - Mahatma Gandhi 

 The face that comes before my eyes whenever I am in a space of doubt is that of a poverty stricken, yet dignified, young female farmer in India. I met her only briefly as I was conducting surveys among farmers in remote rural regions of India during my Ph.D. field work She came and sat on the floor in front of me, with her hands folded – “can you help me get access to money for my food and for seeds for the next season’s crop?” I will not reveal her real name, but will refer to her as Sumi. 

  The memory of Sumi's face is enough to eliminate all doubts and bring me back at my work desk at all odd hours of the day or night. The fields of my research focus may, at first glace, appear to have changed rapidly over the last 10 years - from traditional knowledge, to patent law, to plant variety protection, to blockchain and now to ethical issues in AI and Earth Observations. But the power that pulls me and keeps me going in any and all of these fields of research is the power of that face - the power of the circumstances that the weakest and poorest segments of our society live in, endure, suffer, and some even flourish in, in their own resilient ways. 

 My research always starts with this question - how can this work positively impact the life of Sumi and millions like her? If and once I find an answer, I also find endless motivation and energy. As the world becomes increasingly aware of the need to steer research in emerging technologies in the direction of good, with due emphasis on ethics, I ask myself, can it also be steered in the direction of inclusive good? Good for all? Even those living in the remotest regions of the world? How many AI and Blockchain usecases are aimed at benefitting the poorest, including them in the mainstream, giving them a political voice, or even to putting a dignified two meals on their plates without having them resort to begging? 

  I fear the answer may be "not enough". Yet, these emerging technologies have the potential to lead the world towards more inclusive justice. Ethics and mindfulness, while having become buzz words in the AI era, need to now also become more inclusive in their content and focus. There can be no ethics in the broader sense and from a larger perspective, unless adequate attention is (also) given to the development of the poorest segments of society. With billions of dollars being given for AI research and innovation, governments and funding agencies need to ensure that a clear and sizeable portion of the research funds goes towards the inclusion of the poor - with dignity and not as charity. 

 As I surveyed scores of farmers across some of the poorest regions of India, what I noticed frequently was that no matter how poor, they all have a clear understanding of what is going on on their fields. They also have incredible insight on the impact of chemical intensive farming on their lands and on their health. But they feel helpless in the face of the sheer pressure to adopt and maintain "high yeilds" in a chemical intensive and chemical driven environment. Caught between debt traps and increasingly unproductive soil conditions, they are driven away from their farms, towards a life of labour either on another (rich) farm, or to the cities and slum life. 

  It is said that charity too can be classified, ranging from lowest to highest classifications, based on the type and duration of impact. While giving money or food in charity is good, it is classified in a lower category, as both food and money will last only for a very short duration - perhaps only till the next meal (depending on what the money is used for). But giving knowledge is the highest charity. If education can be imparted to these farmers on how to revive their fields, stop being reliant on expensive chemical inputs and become entrepreneurs and innovators on their own fields, saving and sharing their own farm saved seeds, and sharing their know-how, for a small price, with researchers or other farmers who are in similar farming conditions, our poor farmer/laborer community can be converted to a proud and productive force of innovators. 

  I have spoken and written about this earlier in this post, and subsequently, Natalie, our post-doctoral researcher for a UK funded project also published a paper on this idea. But what needs considerable deliberation and hard work is HOW. In our position paper to the government of India (see here), we have recommended a three pronged approach to accomplishing this goal. These prongs are: 

(i) reviving traditional ecological knowledge (TEK) systems; 

(ii) educating farmers and rural agricultural extension officers on means of useing these TEK systems on farm to sustainably increase profits and productivity, and become providers of know-how and materials rather than mere takers, and 

(iii) develop and implement an AI and blockchain or distributed ledger backed solution to ensure equitable, transparent and traceable transfer of know-how and materials from farmers (grassroots) to diverse stakeholders. 

These prongs, if implemented, would help diversity directions of value and knowledge flows and give back dignity to small and marginal farmers engaged in sustainable agriculture using indigeneous seeds and local biodiversity. Earth Observation and Remote Sensing data can further support payments for sustainable agriculture that revives and improves agrobiodiversity in situ - this incentiving such activities that bring both environmental and economic gains to small farmers. 

 Regulations in Europe are changing to facilitate sales of so called "heterogenous materials", i.e. seeds that are not uniform, but genetically diverse and variable. Such seeds have been found to be more resilient to climate change as well as to biotic and abiotic stresses. Through EO, AI and Blockchain, one can track the progression and climatic impact of these farming systems on one hand, and incentivize optimal and increasing adoption of sustainable farming methods by more and more farmers (especially small and marginal farmers) - for their own beenfit and for the benefit of the environment and biodiversity. Policy and regulatory changes the world over should finance, support and promote the creation, improvement, dissemination and adoption of such technologies. 

Only then can we say that we are truly using state of the art technologies, not just for "good", but for inclusive good.