AI and the Fight Against COVID-19
As artificial intelligence (AI) and machine learning (ML) become more prominent in today’s world, they must adjust to every situation, whether it be minor and light or major and crushing. More specifically (on the latter side), AI must – and has already begun to – adjust to worldwide crises and pandemics, such as COVID-19. As the search for a cure continues, several companies have found ways to help the general public as well as patients, health care workers, and researchers during this unprecedented time:
Automatic support and informative databases: In the past, chatbots were better known for being automated customer service representatives or supports. They would offer assistance in real time by automatically addressing client questions and concerns. Swami Sivasubramanian, VP of Amazon Machine Learning, says that the website robots of Clevy.io, a French artificial intelligence startup, introduces the same concept with one main exception: it is programmed to answer questions about COVID-19. As it collects millions of messages each day, the system is able to address the concerns of the public without exhausting federal resources, such as government health institutions. Similarly, AI has also been used to collect and store information on COVID-19 in a cloud-type environment, allowing researchers to submit their findings and compare notes. Brian Uzzi, a consultant on AI, uses the Allen Institute of AI as a key example. The institute created a digital library of reliable sources, all related to COVID-19. The database, called the COVID-19 Open Research Dataset, or CORD-19, contains hundreds of thousands of scholarly articles on the virus that can be compared and contrasted as we search for more answers. It is possible that massive cloud libraries like these could help researchers finally discover a cure.
Monitoring crops: When COVID first hit, the world was thrown into a panic. Stores and supermarkets were practically ransacked until the shelves were bare. To this day, it is still difficult to get our hands on specific items (and not just toilet paper and hand sanitizer). According to Sivasubramanian, Mantle Labs has created, and is currently utilizing, a crop monitoring system that keeps track of agricultural growth and development through satellite images. By taking snapshots of crops, the system can determine whether or not a plant is healthy and ready for harvest. And, if the plant is afflicted in any way, the system can alert the farmers so they can try to save it. This system also helps to determine how much food will be produced in a standard harvest. Because of this, farmers and retailers can better prepare to provide for consumers.
Tracing the virus and its spikes: AI and ML keep track of data clusters across the globe in order to best predict when and where the next pandemic will occur and which populations are most at risk. Some AIs are so advanced as to identify new diseases and viruses as they emerge. According to Sivasubramanian, BlueDot was one of the first AIs to identify COVID’s arrival in Wuhan, China. This AI not only tracks data, but is able to discern and interpret at least 65 languages in order to scan the worldwide news for reports of COVID. Meanwhile, other AIs are used for a more direct approach. According to the staff of AI Trends, AIs (more specifically, ClosedLoop.ai) that observe and predict COVID spikes in specific areas use this information to have their employees call these residents to discuss handwashing, social distancing, and the procurement of essential items. The only downside to this method is its inefficiency in underdeveloped nations. Some countries and nation states do not have the resources to maintain proper hygiene and may even struggle to distribute essential items. While there is room for improvement in the method, its premise teamed with the appropriate technology has already proven to save a plethora of lives. Meanwhile, AI technology is also capable of tracing potential virus transmission from person to person. According to Uzzi, AI apps like DP-3T (aka SwissCovid) store the information of social contacts and, evidently, the whereabouts of users’ phones so that if one person tests positive for COVID, others who recently came in contact with them will be alerted. There is controversy with this method however; while the information collected is anonymous, it is still personal information collected via Bluetooth, which raises privacy concerns among users.
To provide the best possible healthcare: If a potential victim of COVID-19 needs to be tested or treated by a healthcare professional, the patient would trust that their caretaker would be in the proper position to treat them. The wellness of the caretaker would depend on whether or not they exhibit symptoms, have had a full night’s sleep, and show signs of increased stress levels. According to AI Trends, AI technology like the type used in the Warrior Watch study monitors these variables. The Warrior Watch study itself involved the use of Apple Watch to keep an eye on the health and fitness of select healthcare workers. And while the Apple Watch did not necessarily scan or monitor the transfer of bacteria within a patient, it did monitor heart rates, which are directly linked to respiration. Because COVID-19 is a respiratory virus, irregular heart beats could stand as an indicator of a healthcare worker being sick or considered too “at risk” to tend to patients. This, in turn, could ensure safe and effective treatment to those infected.
COVID-19 has left millions of people feeling powerless, because this is an unfamiliar virus we do not yet know how to control. But thanks to AI and machine learning, we can control how we react to the pandemic. We can learn about the virus and its effects through chatbots and large databases. We can plan for enough resources to get us through the crisis. We can trace the virus to know where it is going to hit and when (thanks to AI, we know that it will likely recur in the Fall). And moreover, we can trust that our healthcare workers will be in the right condition to treat and administer to us. According to Oren Etzioni and Nicole Decario at Wired.com, at this moment, AI systems such as DeepMind’s AlphaFold are set to study and predict the shapes of the virus’s protein structures and amino acids. This means that once the organization has an idea of what proteins are involved, they can then start developing an immunization that fits with the protein builds. AI has more than just the potential to give us the knowledge we need to find a cure that will prevent another disaster like this from happening in the future - it has the high probability and likelihood to do so. And with all of these resources available to the public and scientific communities, this pandemic will be stopped.
