April 30, 2020

Enlisting AI in the push to combat Covid-19

by James in Article

New initiatives to use artificial intelligence to battle outbreak and help mitigate future pandemics

Thomas Siebel had big ambitions and a well-planned rollout this spring for the C3.ai Digital Transformation Institute, a research consortium headed by artificial intelligence software company C3.ai in partnership with Microsoft, research universities and businesses across a range of key industries.

With a mission “to attract the world’s leading scientists to join in a coordinated and innovative effort to advance the digital transformation of business, government, and society,” the initiative, known as C3.ai DTI, was set to launch its first call for proposals when the coronavirus pandemic hit.

“The agenda was digital transformation writ large — think smart cities, think logistics, think digital oilfields, think precision medicine,” says Siebel, founder and chief executive officer of C3.ai.

The DTI had raised $400 million “in cash and in kind” and was gearing up to distribute awards up to $500,000 “and access to a lot of computing power,” along with the C3.ai platform, with the stipulation that published research would go into the public domain.

“We were getting ready to announce it when the plague hit,” Siebel says from his home in Northern California.

“I always thought the first call for proposals would be for data anonymization,” he says, describing a tool to allow researchers access to troves of data such as reservoir information or health care statistics while protecting the anonymity of the source.

Instead, DTI’s first round of grants will go to researchers with ideas about applying artificial intelligence techniques to help fight the spread of Covid-19 and mitigate future pandemics.

Siebel, who founded and led Silicon Valley giant Siebel Systems before starting C3.ai, has identified “ripe avenues for impactful breakthroughs from AI” that include genome-specific Covid-19 medical protocols, biomedical methods for drug design and repurposing, modeling and prediction of Covid-19 propagation, and the efficacy of interventions.

He hopes the results of the initiative will have broader implications for public health and the containment of infectious diseases, including the development of more accurate models for infection and mortality rates.

More than $5 million in DTI awards ranging from $100,000 to $500,000 are to be announced in coming weeks.

Siebel says oil and gas industry partners have been enlisted in advancing the research, first with Covid-19 and then the broader mission, but as of press time a formal announcement had not been approved.

C3.ai has also launched a related initiative, the C3.ai Covid-19 Data Lake, which uses the company’s software platform to aggregate large sets of data into what Siebel describes as “a unified, federated image” that researchers can use to analyse data from many different sources – at no cost.

Unifying the data sets helps researchers generate insights faster without having to engage in time-consuming “data wrangling” and integration efforts to make the information useful.

“This is really quite sophisticated, and it will be available to the world at no cost to do research on Covid-19,” he says of the new service.

“We’re not making any money here. All we’re doing is to try to have an impact on a very important topic.”

He adds: “I think the probability that we don’t have some positive impact on this dialogue is zero.”

The C3.ai Covid-19 Data Lake launched 22 April and a second data drop on 15 May is expected to increase the size of the lake by a factor of three, Siebel says.

“We expect this data set to expand at a very rapid rate.”

He would like to sign up more partners from the oil and gas industry “to provide money, guidance, data scientists and computing capacity” — and data sets that could possibly be mined for information that correlates with the spread of coronavirus.

Read the full article here.