Utilitarian AI: Four Reasons To Be Hopeful
Around the world, there is a wealth of amazing initiatives using AI to solve our most pressing challenges. Reasons to be hopeful in an era of war, climate disaster, and AI anxiety.
While there — rightfully — is a lot of attention going towards what could go wrong with AI, it’s time we also have a broad, inclusive discussion about what we want to achieve with AI advancements.
At a talk for start-up founders in June 2023, Sam Altman, CEO of OpenAI, was quoted saying:
“What if we can have things that we can’t imagine today? It’s not even like we solve climate change and every disease is cured. That’s all great. But that’s boring. And that’s obvious and predictable, it is certainly going to happen. What’s the really great stuff?”
Against that, one could argue that curing all diseases and solving climate change is not boring at all — in fact, that’s “the really great stuff.”
The quote is indicative of the state of Big Tech: eccentric billionaires compete to be the first to offer commercial space tourism, others spend millions and millions of dollars prepping their private underground bunkers.
In other words, the people currently dominating tech and AI developments tend to see things through a sci-fi lens, somewhat detached from common people’s very real problems.
As a result, food distribution, rising inequality, climate change, and other pressing global issues don’t receive the same attention from the tech elite and entrepreneurs as space travel, private transportation, and weird AI gadgets do.
With AI, a technology still in its formative years, we have a unique chance to get things right.
Around the globe, there are a wealth of amazing initiatives using AI to solve pressing challenges. Actual reasons to be hopeful in an era of war, climate disaster, and AI anxiety. Let’s have a look at four of them.
South Africa: AI fighting spatial apartheid
In South Africa, the country with the highest levels of inequality globally, satellite images and AI are used to fight spatial apartheid.
When the three founders of the nonprofit Distributed AI Research Institute noticed that significantly less state funding was going to areas predominantly inhabited by Black people, they decided to take action. Since then they’ve been working with a purpose “to put data using AI into the hands of marginalized groups.”
The trio has done so by gathering millions of satellite images as well as geospatial data from all of South Africa. They have then used the images and data to train machine-learning models and build an AI capable of labeling areas as wealthy, non-wealthy, non-residential, or vacant.
Their AI is intended to aid researchers and public service institutions in identifying underused land that could be repurposed for public housing and services, thereby addressing the pressing housing crisis and improving the quality of life for residents in underprivileged townships.
They also work in advocacy and policy-making by providing accurate data to organizations and government agencies, hoping their data can help in the reallocation of resources and in planning infrastructure to meet the needs of growing township populations.
“We want the work to push the government to start labeling these townships so that we can begin to tackle real issues of resource allocation,” Raesetje Sefala, one of the initiators, says.
Worldwide: Eco-acoustics and AI protect forests
Rainforest Connection is an NGO operating across six continents using AI and eco-acoustics to protect forests.
The NGO has built a ‘Guardian Platform’, a solar-powered acoustics monitoring system placed in tree canopies. The physical platform captures environmental sounds that are streamed in real-time to the cloud, where advanced artificial intelligence and machine learning algorithms dissect the audio.
When the AI detects irregularities, such as chainsaw noises or sounds from illegal loggers, it triggers an alert system that helps local rangers and conservationists take action.
Through ecoaccoustics and classifications via AI, the platform can also help monitor biodiversity and endangered species.
The Rainforest Connection is operating in 37 countries, monitoring 736,200 hectares.
India: An AI-powered chatbot breaks barriers
In rural India, villagers turn to the chatbot Jugalbandi, which breaks language barriers and democratizes access to government information on opportunities and regulations. The language barrier exists as a lot of government information is written in English, while India has 22 official languages. According to the initiators, a “major barrier to thriving in Indian society is effective literacy, awareness of rights and benefits, and access to technology.”
Initially introduced in the village of Biwan outside Delhi, Jugalbandi has expanded its services to cover ten of India’s 22 official languages. Through WhatsApp, users can send text or audio messages in their local language, which are then processed and responded to by the chatbot. This process involves speech recognition, language translation, and text-to-speech technologies.
This is one of the early success stories of generative AI: the chatbot is powered by GPT-4. Microsoft, investor in OpenAI, writes that the chatbot already has aided “a farmer who needed help applying for pensions for his aged parents. Another wanted to know why his government assistance payments mysteriously stopped — and how to restart them. A university student needed a scholarship to fund her studies.”
And on to the last case: Using AI to enhance resource allocation.
Worldwide: Better allocations of funds in humanitarian catastrophes
In the last few years, a new field has emerged: Humanitarian predictive analysis. Humanitarian predictive analytics involves using large amounts of data to help machine learning and statistical models predict the details of humanitarian crises, like pandemics, famines, natural disasters, and movements of refugees. By predicting these events accurately ahead of time, organizations can better prepare by arranging emergency funding, supplies, and staff in advance.
Some of the biggest humanitarian organizations have started utilizing AI in their aid efforts. The Red Cross’s Forecast-based Financing program and UNHCR’s Project Jetson are prominent examples. And UN’s World Food Programme uses AI to forecast food insecurity, especially in conflict zones.
This might sound dystopian to some, but “forecasting and early warning systems have always been a component of humanitarian action”, as K4D writes. And results show that predictions with AI can significantly improve timely resource allocation and decision-making during crises.
Still, challenges around data quality and algorithmic bias highlight the enormous responsibility that anyone using new technologies around vulnerable groups has.
Final thoughts
While there — rightfully — is a lot of attention going towards what could go wrong with AI, I believe that it’s time we have a broad, inclusive discussion about what we want to achieve with AI advancements. The brilliant use cases potentially outnumber the dystopian ones, but mostly so, if people and interests outside Big Tech (and Big Money) join the conversation and imagination. With AI still being a technology in its formative years, now is a good time for that.
What other cool solutions should I know about? I would love to hear about AI initiatives that make you hopeful.
