SCIENCE sciencemag.org 22 MAY 2020 • VOL 368 ISSUE 6493 811IMAGE: HEALTHMAPHealthMap uses artificial intelligence and data mining to spot disease outbreaks (colored dots) and issue alerts.
United Kingdom. “There’s such a wealth of
data, we will need some sort of tool ... to me
that tool is machine learning.”
Well before COVID-19 hit, CDC began
an annual competition to most accurately
predict the severity and spread of influ-
enza across the United States. The compe-
tition receives dozens of entries each year;
Biggerstaff says roughly half involve ma-
chine learning algorithms, which learn to
spot correlations as they are “trained” on
vast data sets. Roni Rosenfeld, a computer
scientist at Carnegie Mellon University, and
colleagues have won the contest five times
with algorithms that mine data including
Google searches, Twitter posts, Wikipedia
page views, and visits to the CDC website.
Teams involved in the flu challenge have
now pivoted to COVID-19. They are apply-
ing AI in two ways. One aims to spot the first
signs of a new disease or outbreak, just as
HealthMap did. That requires the algorithms
to look for ill-defined signals in a sea of noise,
a challenge on which a well-trained human
may still hold the upper hand, Pollack says.
AI could also assess the current state of
an epidemic—so-called now-casting. The
Carnegie Mellon team aims to now-cast
COVID-19 across the United States, using
data collected through pop-up symptom
surveys by Google and Facebook, Google
search data, and other sources in order to
predict local demand for intensive care
beds and ventilators 4 weeks out, Rosenfeld
says. “We’re trying to develop a tool for
policymakers so that they can fine-tune
their social distancing restrictions to not
overwhelm their hospital resources.”
Although automated, AI systems are still
labor intensive, notes Rozita Dara, a com-
puter scientist at the University of Guelph
who has tracked avian influenza and is
turning to COVID-19. “By the time you get
to AI, it’s the easy part,” she says. To train
a program to scan Twitter, for example, re-searchers must first feed it examples of rele-
vant tweets, painstakingly selected by hand,
Dara says. AI may also struggle in a rapidly
evolving pandemic, where correlations be-
tween online behavior and illness can shift.
AI has misfired before. From 2009 to
2015, Google ran an effort called Google
Flu Trends that mined search query data
to track the U.S. prevalence of flu. At first
the system did well, correctly predicting
CDC tallies roughly 2 weeks ahead of time.
But from 2011 to 2013, it overestimated flu
prevalence, largely because researchers
didn’t retrain the system as people’s search
behavior evolved, Yom-Tov says. Searches
for news reports about the flu were mis-
interpreted as signs of infection.
“I don’t think it’s an inherent problem,”
Yom-Tov adds. “It’s something that we’ve
learned from.” He and colleagues from Uni-
versity College London recently posted a pa-
per to the arXiv preprint server showing they
could correct for that media-related bias.
Nations struggling to adequately test
for the new coronavirus might be tempted
to use automated surveillance instead.
Biggerstaff says that would be a mistake.
When the flu re-emerges this fall, he says,
only testing will be able to distinguish out-
breaks of it and COVID-19. But AI might help
policymakers direct more testing to hot spots.
“The hope is that you would actually have the
two working together,” says John Brownstein,
an epidemiologist at Boston Children’s who
co-founded HealthMap in 2006.
Some researchers question whether
AI systems will be ready in time to help
with COVID-19. “AI will not be as useful
for COVID as it is for the next pandemic,”
Dara says. Still, machine learning in
epidemiology seems here to stay. Pollack,
who sounded the alarm about COVID-19 the
old-fashioned way, says she, too, is working
on an AI program to help scan Twitter for
mentions of the disease. jO
n 27 May, NASA will launch people
into space from U.S. soil for the first
time since 2011, when the space shut-
tle Atlantis roared aloft on its final
voyage. This time, astronauts will be
riding to the International Space Sta-
tion (ISS) not on a NASA rocket, but aboard
vehicles bought from the private space com-
pany SpaceX: the Dragon 2 capsule atop a
Falcon 9 rocket.
The occasion marks yet another mile-
stone for the private California company,
which over the past decade has gone from
underdog to dominator. SpaceX now handles
about two-thirds of NASA’s launches, includ-
ing many research payloads, with flights as
cheap as $62 million, roughly half the price
of a rocket from United Launch Alliance, a
competitor. SpaceX’s goals are not limited to
low-Earth orbit: Last month it won a contract
to build a Moon lander, and it is steadily test-
ing a huge heavy-lift rocket, called Starship,
that could carry people to Mars.
Researchers see both benefits and risks in
the company’s increasing power. It has low-
ered the cost of spaceflight through innova-
tions such as reusable stages and fairings,
saving NASA money. With its outsize capacity,
Starship could cheaply put large telescopes
in orbit and heavy science experiments on
moons and planets. Yet SpaceX, with a fast-
and-loose Silicon Valley mindset, has over-
looked the potential for its technologies to
contaminate night skies and pristine planets.
Some worry the company, led by brazen bil-
lionaire Elon Musk, could jeopardize NASA’s
long-standing culture of safety. “NASA tries
to model everything to the nth degree,” says
David Todd, an analyst at Seradata, which
tracks launches and satellites. “SpaceX works
on the basis of ‘test it until it breaks.’”
Between 2006 and 2008, the first three
flights of its Falcon 1 rocket ended in failure.Crewed launch
deepens ties
between NASA
and SpaceX
Bold company’s rise
brings benefits and risks
for space science
SPACE SCIENCEBy Adam MannNEWS | IN DEPTHPublished by AAAS