Eton is trampling on
the freedom to think
Jenni Russell
Page 34
one-dimensional and they let
Hassabis pretty much go where his
intuition takes him. Now it’s proteins,
tomorrow it’s quantum chemistry.
“The leaves are the applications,” he
told me using a tree analogy, “but the
roots are the big scientific questions.”
That’s what motivates him, “the
fundamental building blocks, the
Nobel prize-level problems”. His life’s
work has evolved into harnessing the
incredible computational power of
artificial intelligence to solve those
problems and to solve them
relatively soon. It is the ultimate
blue-sky science. You couldtheoretically do the same at our
leading universities if you gave them
the same resources. Rather than
spending £375 million on a failing
satellite company, as the government
did last month in a doomed quest for
a post-Brexit British global
positioning system, ministers might
consider spending it on the next
Demis Hassabis.
Science needs no borders, just a
sense of the possible. Speaking to
Hassabis yesterday, I found his sense
of excitement infectious. I just wish I
was 20 again. But then, at the end of
our conversation, he added:
“Sometimes I think my brain is going
to explode!” Let’s hope not, because
that would be one hell of a big mess.Science is fast-forwarding into the future
A year of rapid research advances has raised hopes for plastic-eating enzymes and cures for the diseases of old age
LEE JAE-WON/ALAMY
new-generation drugs — that the
right incentives need to be created to
develop solutions.
This extraordinary, accelerated
year in science has given us some
models to think about. For all the
“Britain got the vaccine first”
boastfulness yesterday, the thing that
has mattered least has been what
country or which group “won” the
race. Let Vladimir Putin claim the
laurels if that’s what he needs to
hear. For us the priority is getting the
job done.
The same lesson applies to
Deepmind. Hassabis’s company is
owned by the American tech giant
Google, at which point some of you
will boo and mention tax payments
and YouTube videos of terrorist
atrocities. But it’s not so simple. Like
the rest of us, Google people are notDemis Hassabis is given the freedom
to go where his intuition takes himlatter years of our lives into healthy
and pain-free ones.
A little farther down the line we
should be able more quickly to
create new proteins altogether,
although I’m worried that Prince
Charles, that long-time worrier
about genetically modified crops,
might not approve. Demis Hassabis,
the Londoner who is chief executive
of Deepmind (and a genuine hero for
our times) envisages creating
enzymes that could, for example,
devour pollutants such as the plastics
we dump in our oceans.
Yesterday, António Guterres,
secretary-general of the United
Nations, made a speech about
climate change that would be better
described as “desperate” than
“impassioned”. Looking ahead to the
international climate conference to
be held in Glasgow next November,
Guterres said that the planet is
broken, facing an extinction crisis
and the destruction of biodiversity
by pollution.
The good news is that with new
climate promises from the incoming
Biden administration and from
China, we could be heading for a
2.1C degree temperature increase by
the end of the century, instead of the
feared 3C degrees. The bad news is
that many scientists regard even this
lower figure as disastrous.
Nothing can detract from the need
to reduce emissions by pre-emptive
action. But here, too, the advances of
recent times may be able to help us.
For example, Hassabis thinks that his
company’s advances in AI can help
to make buildings much more energy
efficient.
This is not just about what is done
but how it’s done. We know from the
slowly unfolding disaster of
antimicrobial resistance — in which
our antibiotics lose their
effectiveness and aren’t replaced byA
Martian eavesdropping on
the national conversation
this week might well have
concluded that our idea of
a scientific puzzle is how
many Scotch eggs it takes to
constitute a substantial meal. One or
two, with or without a pickle?
The Martian would be wrong. I’ve
seen it argued by lockdown critics
that historians will look back on this
period as having been dominated by
a hysteria akin to that of the
17th-century witch trials. It’s
remarkable how such imagined
future scholars tend to agree with
whoever is imagining them.
My own guess is that this period
will instead be remembered for the
global acceleration in scientific
knowledge and capacity that we
have glimpsed in recent weeks.
Because something amazing is
happening.
By now we all know about the
vaccine. Or rather, the first of several
vaccines. Eleven months after the
first Covid-19 case was documented,
ten months after a Chinese
researcher sequenced the virus’s
genome and shared it with the
scientific community, we will start
immunising people against it.
The speed at which this has
happened is staggering.
But this week saw a development
that could lead to even greater
advances. The British-based artificial
intelligence (AI) company,
Deepmind, had managed to predict
the shape of proteins. Work that
used to take researchers years can
now be accomplished in days.
Proteins, the building blocks of life,
adopt complex three-dimensional
shapes which determine what they
do and how. Accurately predicting
these shapes gives us the ability to
use proteins for our benefit or to stop
them being used against us.
No wonder joyous scientists have
been throwing their lab coats up in
the air. I was struck by the comment
from Professor Dame Janet
Thornton, of the European
Molecular Biology Laboratory, who
had not expected such an advance in
her lifetime. She said: “This is not the
end of anything; it is the beginning
of many things”.
Like what? Deepmind’s
breakthrough is an enabler not a
creator. But it lays the ground for an
exponential increase in discoveries
about human life. The shape of
proteins will help diagnose what is
happening when someone developsAlzheimer’s, how cancers differ, what
viruses are doing and eventually how
to make drugs to combat a host of
diseases and enhance human health.
In the past few days I’ve been
reading about the science of ageing,
and how protein malformation and
malfunction may account for a lot of
the most unpleasant conditions
associated with it. I don’t think it’s
far-fetched to imagine that in the
years following this week’s
announcement we could see new
drugs that could eventually conquer
diseases like malaria and sleeping
sickness, and others that turn theHassabis thinks AI can
make buildings much
more energy efficient
What motivates him
are ‘the Nobel
prize-level problems’
Comment
@daaronovitch
David
Aaronovitchred box
For the best analysis
and commentary on
the political landscape
thetimes.co.ukthe times | Thursday December 3 2020 1GM 33