Using Artificial Intelligence to fight the impact by COVID-19
HomeBlogUsing Artificial Intelligence to fight the impact by COVID-19
Although the hype around Artificial Intelligence may sometimes be misleading, it is fair to say that AI is capable of providing solutions to problems that otherwise we would be struggling to solve. This is true for applications such as natural language processing (NLP), computer vision, image recognition, cancer risk assessment, personalized recommendations among many others.
Due to the severity of the current public health crisis, there are ongoing efforts to use AI to help cope with COVID-19. Combined with epidemiological knowledge, AI is helping to track the spread of COVID-19. On the clinical side, there are new methods that use AI to support triage from lung X-rays images, such as the CAD4COVID method from Delft Imaging. Other methods are being used to repurpose known drugs to help treat patients.
Besides trying to directly tackle the disease, some companies have created AI tools that can be used to ensure social distancing such as the one from Landing.ai. Companies such as Datakalab are creating methods to detect if persons are wearing facemasks. Venture studio Cough for the Cure is building an algorithm to diagnose if a patient has COVID-19 based on the sound of her cough.
Other studies use AI and NLP for monitoring mental health using information posted on social media. These type of studies provide information that can lead to early warning of depression and anxiety and or substance abuse.
The applications to deal with COVID-19 are great examples of the possibilities to use AI for social good. However, the damages caused by the pandemic are also economic. AI can also help to fight the economic fallout of the pandemic. COVID-19 is affecting the most basic processes of our economy, as well as touching nearly every industry. AI can be a key tool for industry and policy makers by providing insights into how the new situation is adversely affecting processes and automatically searching for the optimal ways to restructure them. Companies with some level of AI adoption will probably have better chances for smoother and faster transitions. What are potential uses of AI as businesses try to respond to the “new normal”?.
Supply chain restructuring. Supply chains continue to be disrupted due to the social distancing guidelines. To continue serving their purpose, many business supply chains need to be restructured for logistics network optimization and management of sales and demand. AI can help with this task by analyzing the influence of the different factors in the supply chain networks, which will be translated to optimized workflows.
AI-driven robots. In a world where social distancing is the way out of the pandemic, robots can fulfill different functions helping to minimize human interaction. Smart robots are guided by computer vision and artificial intelligence algorithms. These robots can help from loading/unloading boxes to quality checking, autonomous driving among many others.
AI models for targeted marketing. Purchasing patterns have been altered by the lockdown measures. Marketing for potential customers targeting will have to change to reflect the new reality. How and what we buy is rapidly changing. AI models will be useful to reflect these changes to relearn the potential clients to target for marketing campaigns in order to maximize reach.
New demands will introduce new use cases for AI. This can already be seen in new attempts from companies to strengthen high demand products with AI. An example is the live caption in the (recently made free for everyone) Google Meet. This video conferencing platform incorporates benefits such as AI-powered live transcriptions from video, which is possible thanks to Google’s speech recognition technologies. These types of AI technologies can be repurposed without too much work in a way that allows them to create value for users in old or new products.
However, there are still critical concerns when adopting AI for organizations. We have to be aware of fairness, transparency and other ethical concerns around AI systems. It is known that many forms of bias may be reproduced and reinforced by AI, that could be given by factors such as imbalanced data collection, sample bias (using a person’s race or gender for the models), measurement bias such as when different data distributions occur for training and test data or even algorithmic bias. These issues should be studied and tackled before the AI deployment; otherwise the AI can harm more than benefit humans. Automation of logistics and distribution tasks may solve problems created by the need for social distancing but may also have an undesirable impact on the labour market.
The intrinsic abilities of AI to adapt to changing patterns and to find interesting insights from data become very useful when coping with the disruption and uncertainty caused by COVID-19. Companies using AI-guided working models will find it easier to adapt to the new conditions. There are also opportunities to introduce AI on your business by directly using consultancy services, by training your employees to be able to use cloud-based AI services or by hiring a team of AI specialists. It is also likely that the emphasis on the type of AI projects in demand will shift direction, which represents an opportunity for innovation.