BBC Science Focus - 03.2020

FE ATURE INTERVIEW

DRCAMILLA PANG

Camilla completed her

degree in biochemistry at the

University of Bristol, and

went on to gain a PhD in

bioinformatics from

University College London.

Her current work uses

disease and cancer data,

along with machine learning

methods, to find patterns

that can be used in

healthcare and lead to the

development of therapies.

Camilla is an active

volunteering researcher for

cancer research groups at the

Francis Crick Institute in

London, and also works on

her own projects within

psychology, science

communication and the arts.

Camilla was diagnosed with
autism spectrum disorder
when she was eight years
old. She struggled to
understand the world
around her, and once asked
her mother if there was ‘an
instruction manual for
humans’. Her first book,
Explaining Humans,is a guide to navigating
life, love and relationships using the lessons
she’s learned in her scientific career so far,
drawing on examples from how the different
proteins in the human body can reflect the
different roles in a social group, to the way how
light refracts through a prism helping her to
break down fear into something manageable.

IN WHAT WAY DO YOU STRUGGLE TO
COMMUNICATE?
Well, firstly open-ended questions aren’t my
forte, I think that’s quite important to know.
This whole book highlights why open-ended
questions are hard; I’m on the autistic
spectrum and I also have ADHD, but that
doesn’t define me. I don’t fall victim to my
neurodiversities. They empower me and I
wouldn’t have been able to do my PhD or write
my book,Explaining Humans, without them.

WHAT IS THE BIGGEST THING THAT HAS
SCIENCE TAUGHT YOU ABOUT HUMANS?
Science is still teaching me. It’s not a concrete
thing and this is one of the reasons why I’m in
love with it. It’s something that is ever-
evolving, and so are people.
It’s very fluid, that’s one of the lessons it has
taught me. I like to be on the cutting edge of
science, and I think that inadvertently makes
me more fluid as a person, and less rigid.

WHERE DOES SCIENCE INTERSECT WITH HOW
WE MAKE DECISIONS?
Take machine learning [using patterns in data
to generate algorithms that can predict future
outcomes]; it’s inherently based on
psychology. That’s what we’re trying to do, to
mimic the human brain and then scale it up so
we can gather some kind of insight and
intuition.
Machine learning uses what we know to
anticipate what we should do in unknown

circumstances, aka the future.

They aim to replicate the

convoluted logic and decision-

making processes that happen at

smaller scales, but ever so

accurately, in the human brain.

But currently, as much as AI

can iterate data until the cows

come home, humans are much

more superior, albeit slower, at decision making.

Our decisions are a tad more involved and

layered than computers. Neurotypically, we

don’t refresh our biases every time from random.

Rather we gather them through experiences,

errors and successes, which fine-tune our

intuition to infer outcomes.

HOW DOES YOUR KNOWLEDGE OF MACHINE

LEARNING HELP YOU MAKE DECISIONS?

Imagine I am a committed Uber Eats user and

have a strong sense of routine (which isn’t too

far off). The app in question may predict a

suggestion and cater – quite literally – to my

needs, the evening before I’ve even opened the

app. The same logic could be paralleled in the

attempt to anticipate the whims and woes of

close friends or partners, which is more of an

experienced, convoluted and recursive process.

Or in decisions that concern many people.

Things like; do I, or do I not react to when

someone is really annoying me? Or, what kind

of thing do I say when someone needs a hug?

These decisions are nuanced and require some

level of known experience, some learning from

mistakes and a degree of self-reflection.

This imitates a more snazzy version of AI:

deep learning. Deep learning is an adaptive,

reinforced training, an expertise acquired

purely through its freedom and ability to learn

from its mistakes. Although, deep learning

operates in an accelerated and low pressured

environment, where housework and the fear of

human judgment doesn’t get in the way.

SO, IS THERE REALLY ANY SCIENCE BEHIND MY

FAILED RELATIONSHIPS?

I don’t like the word ‘failed’. I think that’s a very

binary way of thinking. I mention this in

chapter one ofExplaining Humans, it’s box

thinking, and you’re either one or the other. You

fail or you succeed. But evolution is a bit more

flexible than that, so a failure in one context 2

“To be neurodiverse and

show it, takes a lot of guts.

You have to know how to

train it, to use it, and not

be hindered by being an

odd shape”