Original article was published by Praveen Mishra on Artificial Intelligence on Medium
The Story Of Newly-Found Friendship: AI, ML, and The Pandemic
This article talks about how this pandemic has given the AI-ML duo a much-needed push and how this newly-found friendship is working out for good!
One thing that has been a constant motivation behind the evolution of technology is — we, as humans, like to stay in control. And I believe this is the most common reason why we are obsessed with AI & ML in general.
- “I am planning to start a new business. Which industry will become the most profitable in the next 5–10 years?”
- “I am a student. For my specialization, I would like to choose a subject that stays relevant in the future.”
- “Which stock will give me maximum returns in the next 15 years?”
The COVID-19 outbreak caught everyone off-guard and no one really knew how to stay in control of this new situation. Businesses suffered, if not shut down, people died, healthcare facilities were over-utilized and much more!
Artificial Intelligence and Machine Learning came to the rescue. They have been around for long enough but the pandemic made them more significant than ever.
In this article, I will talk about how this pandemic has given the AI-ML duo a much-needed push and how this newly-found friendship is working out for good!
Predicting The Future: Humans vs AI (& ML)?
The thing about letting humans alone do future predictions that they are less likely to be apt and if they turn out to be true, the person behind it is considered to be “extraordinary”. And, in most cases, they are!
Even though we have witnessed pandemics in the past, healthcare facilities & Governments started wondering how did they miss anticipating the COVID-19 outbreak. The question stays — will we be able to predict the next virus outbreak like Coronavirus using AI & ML?
AI, coupled with ML, is more reliable for future predictions as they work with data and never on biasness or any other anomaly. But, time and again, we have seen that human predictions have failed and so have AI predictions.
The Irving Fisher Prediction
Irving Fisher was a noted economist of the 20th century. He was called “the greatest economist the United States ever produced” by Milton Friedman — another American economist. Given his stature, one would assume that his predictions could never go wrong but that was not the case.
Just days before the Wall Street Crash of 1929, Irving Fisher said that — “Stocks have reached what looks like a permanently high plateau” which obviously wasn’t correct!
AI Prediction Gone Wrong
Back in 2018, a woman was killed by a self-driving Uber in Arizona. There are multiple scenarios like this in which, even after multiple tests, AI (& ML)-powered solutions have failed.
So, the question now is AI (& ML) predictions vs human predictions the right battle when the combined solution is more efficient?
Artificial Intelligence, Machine Learning & Human Intelligence — The Better Combination For The Pandemic
One of the perfect examples that can be cited as a perfect example, in this case, is — Synergistic Anticipation of Geopolitical Events (SAGE). Led by Aram Galstyan at Information Sciences Institute (ISI), SAGE is all about using AI & non-experts to predict geopolitical events.
Experts have predicted that in the next 10–20 years AI & ML will help humans to do their jobs more efficiently. And, it is safe to say that one of these jobs would be predicting the future.
The good news is that even though AI-and ML-powered tools couldn’t predict the Coronavirus outbreak, they can definitely play a major role in curbing its spread.
How Pandemic Has Changed Our Inclination Towards AI & ML?
Keeping aside the technicalities for a second, there are many things that businesses and healthcare facilities have realized during this pandemic. Here are a few of them.
Remote Work Is The New Normal
Before the pandemic, remote work and working from home were considered a luxury. Now, remote work has become the new normal.
Companies in the IT sector and other sectors are finding ways to define processes to make remote work as beneficial & productive as working from the office. AI & ML are helping remote teams to function efficiently during the pandemic.
Microsoft even used an AI engine on their latest launch — Surface Pro X to make video chats “less weird”.
- AI-Powered Ticketing Systems are helping remote teams to effectively handle in-house concerns and customer complaints during the pandemic.
- The increase in remote conversations has been streamlined by AI. LivePerson reported that its Conversational AI-powered chatbots & messaging tools are witnessing a “significant” increase in volume due to the pandemic.
- Machine Learning is facilitating robust access to data and the remote computational needs of companies during this pandemic.
Some Businesses Can Survive in Lieu of Severe Budget Cuts
Several businesses have suffered huge losses due to this pandemic leading to severe budget cuts.
According to a study by the National Bureau of Economic Research, more businesses closed in 3 months of this pandemic (February to April) as compared to the entire Great Recession!
As such, companies are now aggressively learning to adopt AI — and ML-powered automation tools and stay functional in light of severe budget cuts.
In a survey by Algorithmia, 43% of IT leader respondents agreed that AI & ML matters much more than they initially thought because of the pandemic.
Accurate Predictive Analytics in Healthcare is More Important Than We Anticipated
The COVID-19 outbreak, in multiple ways, has been a “stress test” for healthcare facilities all across the globe.
In response to the COVID-19 outbreak, researchers from a diversity of fields have risen up to the occasion and notably, among them are ML researchers.
In the past two decades, both AI and ML have shown promising results in the medical field. Even to a point that they have been applied to address some of the top societal challenges — one of which is predicting the spread of disease.
Bringing AI & ML capabilities to fight against COVID-19 will definitely give medical facilities an upper hand as compared to the usage of traditional analytics and current clinical decision-making techniques.
Machine Learning can be used to effectively identify data-patterns from complex datasets to provide unprecedented insights regarding disease spread, patient outcomes, drug discovery, and more!
AI, ML, and The Pandemic — For Businesses
Businesses are doubling their AI/ML spent in the light of the ongoing pandemic and for all the right reasons!
AI & ML have been around for long but their potential was not fully realized by companies. The pandemic has brought the much-needed urgency in the implementation of these technologies.
Machine Learning technologies empower modern-day computers to mimic human intelligence on voluminous datasets, identify patterns, and predict the spread of this pandemic.
Artificial Intelligence, on the other hand, is learning to carry out tasks that were earlier executed using traditional methods, more efficiently and as per their urgency.
Be it a small business or large scale enterprise, businesses are finding new ways to operate efficiently while meeting the norms of social distancing. The areas where these technologies can help your operations, of course, will depend on your business.
Here are a few instances of how AI & ML are helping businesses.
- Using Machine Learning capabilities (of AWS) to predict daily orders, Domino’s Pizza delivers in record time.
- iFood, a Latin American food delivery company, was able to decrease net delivery distance covered by 12%, reduce downtime of personnel by 50%, and increase SLAs from 80% to 95% using Artificial Intelligence.
- Certipass, an accredited body that offers certification of digital skills, is using Amazon Rekognition for automated identity verification of their candidates during tests in the ongoing pandemic.
Other cases of AI and ML implementation in businesses during the pandemic include customer support chatbots, predicting cash flow, document/identity verification, and much more.
AI, ML, and The Pandemic — For Healthcare Organizations
Pandemics are a real test of Artificial Intelligence & Machine Learning. Right after 2000, Healthcare facilities across the globe have been trying to use these technologies to predict disease spread, anticipate virus outbreak, prevent the spread of false information, drug discovery, self-analysis, and such others.
- AI & ML can keep people informed about the official rules and communication. For instance, Clevy.io, a French start-up, has launched a chatbot for people to easily access official government communications about the COVID-19 outbreak.
It is powered by information from the French government and the World Health Organization in real-time and has sent over 3 million messages to date! Orléans, Strasbourg, Nanterre, and other French cities are using Clevy.io to decentralize the distribution of verified information directly to their residents.
- Machine Learning is also being used to analyze large amounts of data about the virus spread and anticipate early warning signs. The partnership of Chan Zuckerberg Biohub and AWS Diagnostic Development Initiative (DDI) is a classic example of this.
Using Machine Learning, the Biohub has developed a system to detect and quantify the undetected spread of COVID-19.
- The early access of information can truly help healthcare organizations and businesses to prepare themselves for a pandemic as powerful as the ongoing COVID-19.
BlueDot — an AI-driven algorithm, combs through the entire Internet, new reports in 65 different algorithms, plant & animal disease networks, and official proclamations to send warnings way in advance.
Dr. Kamran Khan (Founder and CEO, BlueDot) says — “What we have done is use natural language processing and machine learning to train this engine to recognize whether this is an outbreak of anthrax in Mongolia versus a reunion of the heavy metal band Anthrax.”
BlueDot even successfully predicted the location of the Zika virus outbreak in South Florida. And with the release of advancements in NLP & ML such as Google’s BigBird, tools like BlueDot will become more & more powerful.
CDC, WHO, and similar organizations are receiving an enormous amount of data about the spread of the COVID-19 virus and related events. Notably, what many other researchers also pointed out, collaboration & cooperation between healthcare organizations, at the global level, are crucial to realize the true potential of Machine Learning and Artificial Intelligence!
Besides collecting and analyzing huge datasets, ML and AI can also help healthcare facilities in multiple other ways during this pandemic. For instance, ML techniques can considerably alleviate the task of medical imaging, like chest CT scans, combined with expert input to detect the COVID-19 virus.
AI & ML — The Duo is Here to Stay
The pandemic has sealed the deal for AI & ML and we all have high hopes for this newly founded friendship.
However, one of the biggest problems that these technologies face in unprecedented situations like the COVID-19 outbreak is the lack of reliable datasets (from the past). For now, organizations are feeding as relevant datasets as possible at the beginning of an outbreak. Hopefully, researchers will find a more definitive solution to this as well.
What are your thoughts on this? Let me know in the comments!
 Swami Sivasubramanian (VP of Amazon Machine Learning, AWS), How AI and machine learning are helping to fight COVID-19, World Economic Forum
 Sathian Dananjayan and Gerard Marshall Raj, Artificial Intelligence during a pandemic: The COVID‐19 example, Wiley Online Library
Nathan Peiffer-Smadja and team, Machine Learning for COVID-19 needs global collaboration and data-sharing, Nature Machine Intelligence
Finding a role for AI in the pandemic, Nature Machine Intelligence