Artificial intelligence, the next and most amazing frontier of technological revolution, is basically an idea that computers can perform functions typically associated with the human mind. The term “artificial intelligence” was coined by John McCarthy in the research proposal for a 1956 workshop at Dartmouth. McCarthy shared how a computer can improve itself (that is, learn or evolve) and think creatively
Machine learning (ML), a subset of AI, focuses on learning from data, identifying patterns and making decisions with limited human intervention. ML systems can make better decisions than humans because they can take many more factors into account and analyze them in milliseconds. The backbone of AI and ML are well-constructed algorithms.
Generally, an algorithm takes some input and uses mathematics and logic to produce the output. An algorithm is like following a recipe: prepare the ingredients, heat the oven to 250c, and cook a homemade pizza (output) in 15 minutes In stark contrast, an ML Algorithm takes a combination of both inputs and outputs simultaneously in order to “learn” the data and produce outputs when given new inputs. Let’s imagine oven is too hot. Through ML, the system learns from the past that the oven gets too hot and so turns it down. Thus, in ML, an algorithm is a set of rules given to the computer program to help it learn on its own.
There are three main types of machine learning algorithms: (a) Classification Algorithms involves dividing the dependent variable into classes and then predicting a class for a given input (e.g. classify emails as spam or non-spam). (b) Regression Algorithms are used for predicting continuous variables (e.g. stock prices). (c) Clustering Algorithms involve assigning the input into two or more clusters based on feature similarity. We can, for example, find all transactions which are fraudulent in nature.
AI is a rapidly growing field that’s already all around us — from Google Maps to Amazon Alexa to Snapchat filters. Humanoid robots and self-driving cars are already reality in some countries. New smartphones are using computer vision technology to recognise your face with high enough accuracy to unlock the phone. AI is employed in the financial world in fraud detection. Chatbots like Siri are able to understand our words and even converse! Netflix is able to use massive amounts of data to suggest a movie you might like.
On societal front, AI’s powerful capabilities could be harnessed and added to the mix of approaches to address some of the biggest challenges of in the world of medicine. AI-based applications could improve health outcomes and the quality of life for millions of people in the coming years. AI is being applied to genomic data to help doctors understand and predict how diseases spreads, meaning more effective treatments can be developed. Other prime applications include clinical decision support, patient monitoring and coaching and automated devices to assist in surgery or patient care.
Researchers at the MIT Media Lab have applied reinforcement learning, a capability in which systems essentially learn by trial and error, in clinical trials with patients diagnosed with glioblastoma (the most aggressive form of brain cancer) to successfully reduce toxic chemotherapy and radiotherapy dosing. This example is particularly exciting as it shows capabilities still in development being applied to social good use cases; reducing chemotherapy doses helps improve quality of life of cancer patients and reduce the cost of their treatment.
Artificial intelligence could also be game changer in detecting and managing Alzheimer. A World Health Organisation report says that they hope AI can reverse “two decades of failed experimental therapies for Alzheimer’s disease.” One of the difficulties with Alzheimer is that by the time all the clinical symptoms manifest and we can make a definitive diagnosis, too many neurons have died, making it essentially irreversible. Currently, diagnosis requires extensive evaluation, physical examination, a battery of tests, and neuroimaging by MRI, CT, or PET scan. This can take months, and even then, it’s impossible to deliver a definitive diagnosis. We can use the data information from a large data set from various diagnosis procedures to create an AI system, which could help with the diagnosis of a of Alzheimer’s disease using ML approaches. A machine-learning algorithm could hopefully diagnose early-stage Alzheimer’s disease, helping to get the patient the treatments they need.
Though clinical applications have been slow to move from the computer science lab to the real-world, there are hopeful signs that the pace of innovation will improve. Advances in healthcare can be promoted via the development of incentives and mechanisms for sharing data and for removing overbearing policy, regulatory, and commercial obstacles.
Artificial Intelligence continues to bring incremental benefits to human life. While we develop principles, methodology, and frameworks to ensure that AI is not misused, the artificial intelligence systems have amazing potential to build a better health and yes, a beautiful world.
Zenc, Ozgur “Notes on Artificial Intelligence, Machine Learning and Deep Learning for curious people” Towards Data Science, 2019.
Berruti, Federico, et al. “An executive primer on artificial general intelligence” McKinsey Quarterly, 2020.
Chui, Michael, et al. “Applying AI for social good” McKinsey Global Institute, 2018.
Marr, Bernard “What Is Machine Learning - A Complete Beginner's Guide” Forbes, 2017.
Stone, Peter, et al. “One Hundred Year Study on Artificial Intelligence: Artificial Intelligence and Life in 2030” Stanford University, 2016.
“Artificial Intelligence Algorithms: All you need to know” edureka, 2018.
Smith, Dana “Artificial Intelligence Can Detect Alzheimer’s Disease In Brain Scans Six Years Before A Diagnosis” University of California San Francisco Research, 2019
AI4ALL website ai-4-all.org