By Steve Kilpatrick
Co-Founder & Director
Artificial Intelligence & Machine Learning
The field of artificial intelligence is growing rapidly. New research makes AI more effective every day. Already, machine learning algorithms can process millions of data points. Computer vision applications can identify objects from images in real time, and natural language processing has advanced to the point where you can talk to Alexa in your home.
In the wake of these new developments, many technologists are turning their eyes toward the future. As AI becomes more sophisticated, it will increasingly permeate every industry and aspect of daily life. Some have a negative view of AI’s future implications. They believe AI will cause massive unemployment, lead to greater wealth disparity, and possibly see the beginning of militarised machine intelligence.
AI certainly is potent, and we shouldn’t downplay the ethical concerns associated with AI. Still, the same power that makes AI so intimidating can also be applied for good. AI could help us solve some of the world’s toughest problems. It can provide us with new approaches and ways to understand and model the world around us. The end result could range from curing diseases to ending poverty. In this article, we’ll explore AI’s potential for good and what initiatives are pushing AI in a benevolent direction.
Applications Across Industries
Artificial intelligence is promising anywhere we have lots of data, complex relationships between data, or real-world systems that are difficult to model and predict. Nearly every major industry and issue in our modern society has potential use cases for AI. Urban planners need to model traffic. Doctors want faster, more comprehensive test results, and emergency responders need ways to prioritise rescues and coordinate supplies in the event of a disaster. AI can help with all these scenarios and more.
According to research from Accenture, the AI healthcare market is expected to reach $6.6 billion by 2021. Robot-assisted surgery, virtual nursing, and preliminary diagnosis are all areas where AI could make healthcare more effective and cheaper at the same time.
Some exciting AI healthcare applications are already upon us. Automated image detection and diagnosis are a key area where AI is showing major promise. A recent study found that AI was better than trained dermatologists at diagnosing skin cancer based on images.
AI could also help people with disabilities participate more fully in society. Microsoft’s Seeing AI app is already available for free and helping people with visual impairments to read product packaging, count currency, recognise friends, and even describe colours and scenes. Similar products like TapTapSee allow users to take photos and the AI will recognise and describe aloud what’s in the photo.
AI is also powering a revolution in drug discovery. What was previously a time consuming and expensive process of trial and error can now be modelled using algorithms in many cases. Pfizer, for example, is using IBM’s Watson to do just that in its search for cancer drugs.
Another area where AI shows great potential is crime prevention and response. Police departments and law enforcement agencies around the world are experimenting with new AI applications to make our neighbourhoods safer. Technologies like ShotSpotter use AI combined with smart city listening devices to triangulate the location of gunshots and direct police to the scene. Other companies, like Chinese company Hikvision, are working on ways to make cameras that can scan license plates, recognise faces, and detect anomalies like suspicious packages.
Intel is also working on using AI to help recover missing children. Working with the United States’ National Center for Missing and Exploited Children, Intel has developed a machine learning algorithm to organise, investigate, and prioritise tips for referral up to the FBI and other law enforcement agencies.
Of course, there are drawbacks to using AI in law enforcement. Machine learning algorithms can show racial and other biases when they’re trained on biased data. Any crime prevention scheme should involve humans supervising and reviewing the information AI provides.
Humanitarian aid is increasingly a data-driven endeavour. It makes sense that artificial intelligence should be able to help us answer difficult questions that arise when disaster strikes. Intelligent algorithms could help relief workers and first responders prioritise who to save and plan the most efficient way to get food and supplies to those in need.
On a global scale, AI in humanitarian aid is also promising to model and explore the most effective responses to global issues of poverty, famine, and displacement. Thoughtful use of AI is important here, too, though. Like crime prevention, machine learning algorithms’ biases could play an unwanted role.
Artificial intelligence can help relieve pressure on the education system, both in developed and developing countries. For starters, AI tutors and customised curricula could make learning more personalised to each child. The other side of the coin is AI algorithms that can automatically generate feedback for teachers on pedagogy and training. In addition, AI can automate many of the administrative tasks involved in being a teacher, freeing educators up to spend more time interacting with students.
Around the world, such systems would allow children in developing countries to access the same quality of education as major industrialised nations. Online learning will become increasingly prevalent. AI could help children learn at their own pace and play to their natural strengths as learners.
AI algorithms are also helping us model and improve our environment. Farming has benefited from AI systems that can detect anomalies and tell farmers where to apply more irrigation or soil amendments, leading to higher yields. Modern energy grids will use AI to manage supply and demand, allowing for a faster movement to renewable sources.
Around the world, cities are installing sensors and other devices that allow for systematised management of city services. Water systems could use AI to keep drinkable water flowing for residents and manage waste. Traffic lights could be optimised down to the second to get the most efficient flow of traffic. AI could also find and then coordinate traffic signals along the quickest route for emergency vehicles.
AI for Good Initiatives
Many governments, organisations, and companies are working together to make this benevolent AI future a reality. The Artificial Intelligence for Good Foundation is applying AI research toward the UN’s global sustainable development goals. The UN itself is organising global AI summits around the world to address global problems.
In addition, IBM and XPRIZE are offering a $5 million reward to the winner of its AI award. The competition is purposefully open-ended, giving teams the opportunity to develop a novel solution for any grand challenge facing the world.
Not to be outdone by IBM’s sponsorship of the XPRIZE, Microsoft has dedicated entire teams of resources to advancing AI for good. The Seeing AI app for the blind, above, is just one example of a Microsoft AI for Good initiative. They’re working with organisations around the world to build solutions to disaster recovery, education, refugee crises, and human rights.
Sharing the Benefits of AI
The benefits of AI can and should be available to society as a whole. The latest advances in the AI for good movement have proven that AI’s extraordinary potential can solve big, difficult problems for humanity. It’s exciting to see everything AI has accomplished so far and imagine what might be possible in the near future.