Artificial intelligence apps like ChatGPT and DALL-E, which can generate remarkably coherent text and graphics in response to short prompts, started taking the world by storm late last year. Known as generative AI, these apps open up new business opportunities and ethical questions related to property rights, privacy, misinformation, and more. Flying under the radar is a growing group of social entrepreneurs who are leveraging the new technology to address pressing social issues, putting AI ethics at the center. Among them are Bangalore-based social entrepreneurs Sachin Malhan and Supriya Sankaran who co-founded Agami in 2018. Ashoka’s Hanae Baruchel spoke to Sachin to gain insight into the role generative AI could play in democratizing access to justice in India and beyond.
baruchel: There’s so much buzz around generative AI right now that it’s hard not to be skeptical about some of its uses. Why are you so interested in its potential in the context of access to justice in India?
Malhan: More than 1.4 billion people live in India and only about 10 percent of the population has access to justice because it is far too expensive for the average person. AI has the potential to absolutely lower transaction costs and level the playing field by helping people understand things like what their rights are; what to look for if and when they need a lawyer; or what legal questions to ask. AI can also help lawyers and individuals determine whether a title deed meets the standard. It can reduce investigation time and help unclog court files. If we can reduce some of those costs to near zero, it could lead to a massive explosion of access to justice in countries where the system is massively underfunded, whether in Southeast Asia or Africa.
But for that, we need public-facing innovators to build the middle tier of AI for Justice, and then a bunch of entrepreneurs to build solutions that serve people from all walks of life. Most people in our space will create AI to help large corporations navigate litigation, handle documents, and generally serve the high-paying class. No doubt we’re about to see an incredible wave of innovation, but will it be affordable? Will it be aimed at public purposes?
Hanae Baruchel: What has this rapid evolution in generative AI meant for organizations like yours?
Sachin Malhan: For our own work developing an ecosystem of AI for Justice solutions in India, it’s potentially revolutionary. We used to spend hundreds of hours teaching the computer to recognize and structure different types of data. For example with one of our OpenNyAI apps –in Hindi, ‘nyay’ means justice– we wanted the computer to recognize what a court decision looks like and highlight key facts to create summaries of judgments. This meant that we had to annotate 700 to 750 court records ourselves before it could begin to understand the patterns. This is long, painstaking and expensive work. With the refinement of GPT, LaMDA and other major language models, you could now dump 500,000 reviews or even a million at once and it would do the annotation practically on its own, “unattended”.
baruchel: You’ve already started integrating generative AI into your work. Can you give an example?
Malhan: Yes. We’re in the middle of a little pilot called Jugalbandi, where we train ChatGPT to answer any question related to government rights in India, such as eligibility for affordable housing scheme. We input the government scheme information – the clauses, the eligibility criteria, etc. – to ensure accuracy and explainability, and ChatGPT adds an interactive layer to it.
baruchel: You mean I could go to your app and say, “I am in Bombay. Can you help me?”
Malhan: Exactly, and the system would reply, “What kind of support are you looking for? Is housing interesting?” And you could say, “Oh, yeah, housing would be great.” It will start asking things like “How old are you? Do you have an existing home? Do you have family members?” It will communicate with you at your own level of conversational comfort.
The key here is that it works even if you are semi-literate or illiterate in your own local language, because we integrate Bhashini ULCA, an open-source data project that enables speech recognition and translation from a dozen Indian languages to another. So I could ask ChatGPT a question in Hindi or Bengali and it would answer me both by text and voice message in my own language. For the first time ever, someone in a remote village in India can ask questions and get immediate answers about the government entitlements they might be eligible for. This is a potential game changer, as most research shows that access to essential services fails at the last mile because people don’t know what’s available to them or how to use existing systems.
baruchel: How do you consider the risks of applying AI in such high-stakes situations? When you talk about government rights and social services, we are really talking about the most vulnerable segments of society.
Malhan: Things are moving so fast right now that this is a real and legitimate concern. Most people don’t even take the time to think about fair use or privacy. That’s why it’s been so important to us to build this middle layer of AI applications as a collaborative, open source effort. Somebody’s going to build these tools whether we do or not but if we manage to build it as part of a community effort with a really diverse group of people who are impact oriented and can offer perspectives on the things we should be paying attention to will be much better equipped to mitigate unintended consequences.
baruchel: What is missing for more people to build out technology in this way?
Malhan: We need to create the spaces where entrepreneurs, innovators and academics interested in building better AI and better AI applications can think together about the hard questions. In India, we are working with a wide range of technologists, grassroots organizations and lawyers to solve problems as they arise and design this middle layer of AI for justice in a way that works for everyone. We need to build a global Justice AI entrepreneurial ecosystem to develop the parameters for conversational AI privacy rules, conversational AI bias, and more. Things move so fast that we don’t even have time to anticipate the problems. That’s why when Sam Altman, CEO of OpenAI, was asked, “What do you think we’re not talking about?” he surprised a lot of people when he said, “Universal Basic Income.”
Follow them for more information about Agami’s work Twitter.