◼ TECHNOLOGY Bloomberg Businessweek March 1, 2021
17But that spring, week by week, more people
started to pay attention to Gu’s work. Prominent
biologist Carl Bergstrom tweeted about Gu’s model,
and not long after, the U.S. Centers for Disease
Control and Prevention included Gu’s numbers on
its website. As the pandemic progressed, Gu, a
Chinese immigrant who grew up in Illinois and
California, found himself in regular meetings with
the CDC and teams of professional modelers and
epidemiologists. Traffic to Gu’s website exploded,
with millions of people checking in daily to try to
find out what was happening in their states.
As more forecasting models began to appear
through 2020, Nicholas Reich, an associate profes-
sor of biostatistics at the University of Massachusetts
at Amherst, collected 50 or so and measured their
accuracy at his Covid-19 Forecast Hub. “Youyang’s
model was consistently among the top,” Reich says.
In November, Gu decided to wind down his
death forecast operation. Reich had been blending
the various forecasts and found that the most accu-
rate predictions came from this “ensemble model,”
or combined data.
“Youyang stepped back with a remarkable sense
of humility,” Reich says. “He saw the other mod-
els were doing well and his work here was done.”
A month before stopping the project, Gu had pre-
dicted that the U.S. would reach 231,000 deaths on
Nov. 1. The actual number was 230,995.
The IHME’s Murray has his own take on Gu’s exit.
He says Gu’s model wouldn’t have picked up on the
seasonal nature of the virus and would have missed
the winter surge in cases and deaths. Machine learn-
ing algorithms are based off the past and cannot
account for changes such as virus variants and how
well vaccines may or may not work against them,
according to Murray. For its part, the IHME called
the early peak of the virus, then erred by predicting
a steep decline in deaths until it adjusted its model.
“Since then, we are the only group that has gotten
it right consistently,” Murray says.
Reich’s assessment of the IHME’s figures is more
tempered: After the early misses, “recently it has
been a reasonable model,” he says. “I would not say
it is one of the best, but it is reasonable.”
Gu was unmoved by Murray’s comments about
his forecasts. “I’m very appreciative of Dr. Chris
Murray and his team for the work they did,” he
says. “Without them, I would not be in the posi-
tion I am today.”
To the extent that we can learn from this data
story, Reich says the CDC and others should be
quicker to combine models and distribute blended
data. “We have to have people ready, instead of
going around knocking on people’s doors,” he says.
After taking a bit of a break, Gu, now 27 and
living in a New York apartment, did get back into the
modeling game. This time, he’s tracking how many
people in the U.S. have been infected by Covid- 19,
how quickly vaccines are being rolled out, and
when, if ever, the country might reach herd immu-
nity. His forecasts suggest that about 61% of the pop-
ulation should have some form of immunity—either
from the vaccine or past infection—by June.
Before the pandemic, Gu hoped to start a new
venture, possibly in sports analytics. Now he’sconsidering sticking to public health. He wants to
find a job where he can have a large impact while
avoiding politics, bias, and the baggage that some-
times comes with large institutions. “There are a lot
of shortcomings in the field that could be improved
by people with my background,” he says. “But I still
don’t know quite how I would fit in.” �Ashlee Vance▲ GuTHE BOTTOM LINE This spring, when health institutions were
unable to predict the path of the virus, a young data scientist built a
better model as a hobby—highlighting flaws in the system.“Youyang was
the one person
actually
looking at the
data and doing
it properly”