Gilles Sabrié | The New York Times
Renowned AI labs such as DeepMind, OpenAI, Facebook AI Research and Microsoft have remained relatively silent as the coronavirus has spread around the world.
“It is fascinating to see how quiet he is,” said Neil Lawrence, the former director of machine learning at Amazon Cambridge.
“This (pandemic) shows what bulls are – most AIs. It’s great and it will come in handy someday, but it’s no surprise that in a pandemic we fall back on proven techniques. “
These techniques include good old-fashioned statistical techniques and mathematical models. The latter is used to create epidemiological models, which predict how a disease will spread across a population. Right now, they’re much more useful than areas of AI like reinforcement learning and natural language processing.
Role of AI
Sure, a few useful AI projects happen here and there.
In March, DeepMind announced that it had used a machine learning technique called “free modeling” to detail the structures of six proteins associated with SARS-CoV-2, the coronavirus that causes Covid-19 disease. Elsewhere, Israeli start-up Aidoc uses AI imagery to report abnormalities in the lungs, and a British start-up founded by Viagra co-inventor David Brown uses AI to search for drug treatments Covid-19 .
Verena Rieser, professor of computer science at Heriot-Watt University, pointed out that autonomous robots can be used to help disinfect hospitals and that AI tutors can help parents bear the burden of home education . She also said that “AI companions” can help with self-isolation, especially for the elderly.
“On the outskirts, you can imagine doing things with video surveillance,” said Lawrence, adding that cameras could be used to collect data on the percentage of people wearing masks.
In addition, a facial recognition system built by the British company SCC has also been adapted to identify victims of coronavirus instead of terrorists. In Oxford, England, Exscientia is examining more than 15,000 drugs to see their effectiveness as treatments for coronaviruses. The work is done in partnership with Diamond Light Source, the UK’s national “synchotron”.
But the role of AI in this pandemic is probably more nuanced than some had anticipated. AI is not about to get us out of the forest anytime soon.
“It sort of indicates how excited AI was,” said Lawrence, who is now a professor of machine learning at the University of Cambridge. “The maturity of the techniques is equivalent to the internet noughties. “
Why hasn’t AI had more impact?
AI researchers rely on large amounts of well-labeled data to train their algorithms, but at the moment there is not enough reliable data on coronaviruses to do so.
“AI learns from large amounts of data that have been manually tagged – a time-consuming and expensive task,” said Catherine Breslin, machine learning consultant who worked on Amazon Alexa.
“It also takes a long time to create, test and deploy AI in the real world. When the world changes, as it has done, the challenges of AI will be to collect enough data to learn and to be able to build and deploy technology quickly enough to make an impact. “
Breslin agrees that AI technology has a role to play. “However, they will not be a quick fix,” she said, adding that even if they did not end the virus directly, they could make people’s lives easier and more fun when they were in detention.
The AI community is thinking long and hard about how it can make itself more useful.
Last week, Facebook AI announced a number of partnerships with academics across the United States.
Meanwhile, DeepMind’s polymath leader Demis Hassabis is helping the Royal Society, the world’s oldest independent scientific academy, on a new multidisciplinary project called DELVE (Data Evaluation and Learning for Viral Epidemics). Lawrence also contributes.