30 January 2021 | New Scientist | 15Technology Quantum computing
Matthew Sparkes Leah Crane
AN ARTIFICIAL intelligence that
can grade the skill of a pianist
with near-human accuracy could
be used in online music tutoring.
Brendan Morris at the
University of Nevada, Las Vegas,
and his colleagues selected almost
1000 short video clips of people
playing piano from YouTube and
got an expert pianist to manually
grade each on a 10-point scale.
The researchers used half of
these videos and their associated
grades to train a neural network,
a form of AI, creating a model that
can assess piano playing in unseen
videos. They then used the other
half of the clips to test the model.
The team ran the test three times,
first giving the AI access to just
audio, then just video and finally
to both together. The audio-only
version was 65 per cent accurate
when compared with the grades
assigned by an expert, increasing to
74 per cent for video-only and 75
per cent for the combined version
(arxiv.org/abs/2101.04884).
Morris says that although the
software works well, it is unclear
how the AI chooses its grades. This
lack of transparency is a common
problem with neural networks.
“We don’t know specifically
what it’s doing. As with a lot of
AI, we can’t say exactly what’s
happening,” says Morris.
The software could be doing
anything from identifying a pianist’s
ability to play two widely separated
notes with one hand to their ability
to quickly play notes at large
intervals, he says. The AI has
probably identified hundreds
of small clues like these and
taken them all into account
in its assessment, he says.
Even without fully understanding
how it works, Morris hopes the
technique will eventually prove
useful in tutoring, potentially
reducing costs and lowering
barriers to learning piano. ❚
Being graded by
an AI could improve
your piano playing
QUANTUM computing can be
chaotic, but key properties of
that chaos may actually help
us develop useful devices. That
is the finding of a study of the
behaviour of quantum bits, or
qubits, which has shown that
their chaotic nature may be
easier to predict than thought.
Quantum computers use
qubits as the basic unit of
memory, the same way regular
computers use bits. However,
while a bit can only be in one
of two states – a 1 or a 0 – a qubit
can be in a combination of the
two, giving it an advantage over
its classical counterpart. Whenqubits are grouped together,
such as on a quantum computer
chip, they can display chaotic
behaviour, each oscillating
between different values. That
can make changes in their final
states hard to predict.
Alexandre Zagoskin at
Loughborough University in the
UK and his colleagues simulated
this behaviour for certain types
of quantum computing systems.
They found that systems of five
or more qubits display not just
chaos, but hyperchaos, which
makes their behaviour even
more unpredictable.
“Chaos appears when a small
difference in initial conditions
causes a very fast-growing
difference in the trajectory of
the system’s behaviour,” says
Zagoskin. “In hyperchaos, the
trajectories run away from one
another in many directions.”
However, the chaotic
dynamics of the system
can be reined in by changing
the properties of the energyentering the system. That could
make hyperchaos useful as a
random number generator,
one of the potential early uses
of quantum computers.
More importantly, though,
the researchers found that
adding more qubits didn’t make
the hyperchaos in the system
grow exponentially as they
had suspected it would.
Instead, it grew linearly –
each additional qubit added
one more layer of chaos. This
means the system is easier to
describe mathematically – and
hence simulate – than if adding
qubits made chaotic behaviour
shoot up (npj Quantum
Information, doi.org/frvr).
“This system shows very
non-trivial and new quantum
phenomena which we have
never seen before,” says Shiro
Kawabata at Japan’s National
Institute of Advanced Industrial
Science and Technology. “In
this sense, this system can
be regarded as a new type of
quantum simulator, a ‘chaotic
quantum simulator’.”
Simulating quantum
behaviour with classical
computers is a challenge. If itwere possible to simulate lots of
qubits precisely with a classical
computer, there would be no
need for quantum computers.
“We can calculate how a
small group of qubits will
behave, but we cannot extract
from this information about
how a realistic, large group
of qubits will behave,” says
Zagoskin. “This is a bottleneck
in quantum computing.”
This work won’t help us
simulate the specifics of a large
group of qubits, Zagoskin says,
but it may help in figuring out
their general behaviours, such
as how to control a system to
minimise chaos.
He likens it to building a
model aeroplane in the process
of engineering a real one: it
won’t behave exactly the same
as if it were life-sized, but it can
nevertheless help guide the
final design.
Martin Weides at the
University of Glasgow, UK,
says that understanding how
and when hyperchaos arises
“will be extremely valuable for
the design of future large-scale
quantum simulators and
computers”. The researchers
have already started the next
step in this – to test the
theoretical work in actual
quantum computers. ❚Hyperchaos could help us build
better quantum computers
VC
HAL/GET
TY
IMAG
ESA conceptual
image of a quantum
computer chip“Systems containing
five or more qubits
display not just chaos,
but hyperchaos”