Artificial intelligence (AI) – computer systems able to perform tasks such as visual perception, speech recognition and decision-making – is fast emerging as a useful tool in the battle against the coronavirus.
Before the virus struck, the global AI market was expected to expand at a CAGR of 46.2% from 2019 to 2025, a significant level of growth, and yet its importance in helping scientists, governments and researchers mine vast amounts of data to facilitate better decision making at an unprecedented speed, is likely to lift this figure further. Indeed, such is the potential of AI to help fight the disease that a World Health Organization report last month said AI and big data were “a key part” of China’s response to the virus.
China, where the coronavirus is thought to have originated, and where AI has grown substantially since the government’s Next Generation Artificial Intelligence Development Plan was launched in 2017, used AI technology to support measures restricting the movement of people, forecast the spread of outbreaks, make faster diagnoses and carry out scanner analyses that helped scientists better understand the disease.
Also, in just one month, AI assisted Chinese scientists to recreate the genome sequence of the virus, allowing researchers in Australia to create a lab-grown copy of the cells from an infected patient. Now an accurate genome sequence exists, it has become possible to quickly validate diagnostic tests for the virus that can be manufactured around the world and deployed to millions.
AI is also an invaluable tool in facilitating improved ease of access to scientific publications. By being able to rapidly distil key findings from a wide range of papers, researchers are able to better understand the virus and develop a more efficient vaccine as a result. For example, at the request of the White House Office of Science and Technology Policy (OSTP), and the National Library of Medicine (NLM), the National Institutes of Health last week launched the ‘Covid-19 Open Research Dataset (CORD-19)’, providing access to thousands of existing scientific publications that may hold the key to new strategies or vaccinations that could help save thousands of lives.
To bring the research to life, Microsoft applied ‘literature curation algorithms’ to enable easy search measures, whilst the Allen Institute for Artificial Intelligence (AI2) loaded the project on the AI2’s Semantic Scholar website. By applying AI to over 24,000 of research papers, allowing them to be reviewed for key terms or ideas in a fraction of a second, the scientific community is able to easily identify relevant articles that help them form answers about the nature of the virus and tactics to curb its spread. As MIT Technology Review explains; “The database not only helps consolidate existing research in one place but also makes the body of literature easier to mine for insights with natural-language processing algorithms.” The project represents the most extensive collection of scientific literature related to the ongoing pandemic and will continue to update in real time as more research is published.
Meanwhile, in a similar vein, a UK-US initiative is using AI to screen existing drugs to identify a vast collection of drug therapies that could be used to manufacture a universal vaccine. Funded by the Bill and Melinda Gates Foundation, the collection includes almost every known drug that has been extensively tested for safety, including those not yet approved for therapeutic use. The US research institute Calibr will provide a huge collection of drug molecules to the UK company Exscientia, who will use its AI-driven drug discovery platform to examine each drug’s properties. The synchrotron company Diamond Light Source will then use its facilities to examine protein structures within the identified drugs, replicating essential viral proteins for experimentation.
Whilst AI is proving instrumental in expediting these efforts, the Council of Europe warns that although AI’s contribution to the fight against the coronavirus is significant, even when a treatment or drug is identified, it is not possible to speed up clinical test phases. Therefore, as Exscientia’s chief executive Prof Andrew Hopkins explains; “The fastest this could be done is 18 to 24 months, because of the manufacturing scale-up and all the safety testing that needs to be done on a vaccine.”
Therefore, conscious of this time lag, AI is also being used to manage the stress the virus is putting on global value chains. For example, AI is being used to predict and forecast the path of the virus or its impact on populations by computer scientists from the University of Copenhagen. In the study, AI is being used to calculate which coronavirus patients need ventilators and intensive care by identifying the symptoms that seriously ill patients have in common. “We are aware of certain things that increase risk, such as age, smoking, asthma and heart problems, but there are other factors involved,” explains Espen Solem, chief physician at Bispebjerg and Frederiksberg Hospitals. “We hear about young people who end up on ventilators and older people who do well – without understanding why. So, let’s get the computer to find patterns that we aren’t able to see ourselves.” In this case, AI derived data allows doctors and hospitals to work how many patients will need a ventilator at specific times and plan the resources along their supply chain accordingly.
Away from using AI to develop a vaccine, the technology is also being used as a vital tool in keeping economies, moving – digitally restoring them – by moving and honing activities online. For example, AI is also being used to help societies adapt to the ‘new normal’ of life under lockdown. As Professor Sabine Hauert, at Bristol University, explains; “AI can be used to put people out of harm’s way, for example attention must be paid to manufacturing robots capable of cleaning hospitals, or telepresence systems for remote meetings and consultations, or simply as a way to connect with loved ones we cannot see”.
For example, China is using robots to provide faster diagnostic checks at transport hubs, and Hangzhou city ambulances are assisted by AI to speed through traffic. In addition, companies like Neolix are capitalising on human-free delivery services, with autonomous vehicles dropping off food and groceries. Indeed, many are predicting that the coronavirus may be the adoption tipping point autonomous vehicles were looking for. As Sasha Lekach for Mashable writes; “Now people see a driverless car as a helpful alternative to the crowded, exposure-risky mass transit train and bus rides we used to take. So, what used to be considered a scary, uncertain technology for many, now looks more like an effective tool to protect ourselves from a fast-spreading, infectious disease.”
In the future AI may be able to use social media data to predict human behaviour in order to alert authorities of any upcoming epidemics, allowing them more time to prepare, yet the Council of Europe warns also advises caution in how governments use AI to monitor population behaviour patterns. It suggests that technology that infringes on individual freedoms should not be trivialised on the pretext of a better protection of the population. It warns of the dangers of tracking populations activities through their devices during such an emergency, if there is no way to restore privacy once the crisis has passed. As such, it is calling for extensive conversations about how to store, protect and dispose of the data used during the coronavirus pandemic, to protect citizens privacy and engender trust in authorities.
The battle against the coronavirus is likely to be a long one, with many millions of casualties. The technology engendered by the fourth industrial revolution (4IR) could help speed up the development of our defences against the disease, by arming scientists and governments with the data, knowledge, and therefore the power to find a vaccine that keeps people safe.