171. What is Biocomputing? Interview with Ewelina Kurtys

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
BETH BARANY (00:00):
Hi everyone.
Welcome to, or welcome back toHow to Write the Future Podcast.
I'm your host, Beth Barany.
I am an award-winning sciencefiction and fantasy writer who
is super excited to write inboth genres and both of my
series feature strong womenheroines going out into the
world, being bold, beingadventurous, and I started this

(00:22):
podcast because I really.
Want to focus on the fact thatwith story we can reimagine what
we want as humanity.
And I vote for positiveoptimistic futures'cause what we
envision we can help make it so.
And I believe writers have astrong gift as well as

(00:44):
responsibility to help make thathappen.
Occasionally I interview subjectmatter experts and I have one of
them with us.
Today who I will bring in just amoment.
So I just wanna say Welcome,welcome Evelina.
So happy to have you here withus.
That's so great.
We're gonna talk aboutBiocomputing in a moment, is
that right?

EWELINA KURTYS (01:04):
Yes, correct.

Yes, it's such an interesting topic.
I've been hearing about it forabout the last nine years since
I've been writing my sciencefiction mysteries, my Janey
McCallister mystery series, andreally fascinated by this topic.
So if you could take a momentand introduce yourself, and then
we will dive into the questions.

Hello.
my name is Ewelina Kurtys.
I'm a strategic advisor at FinalSpark, one of the three startups
in the world, which try to buildcomputers from living

neurons.
That's amazing.
That is amazing.
So why don't we just start offwith: What is bio computing?

So Biocomputing is actually a new field.
So the terminology is not yetfully established.
I prefer to use Biocomputing,but you can also use other terms
like wet ware computing ororganoid intelligence.
So this is a new field, in whichwe are trying to use iving
neurons to process informationand we want to use them as

(02:05):
processors, to processinformation as today we do with
computers.

So why is this technology important?
And maybe also if you could saya little bit about why you were
attracted to this field.

So at first I will say why it's important
because we can see today thatartificial intelligence can have
a scalability problem, becauseAI is using increasing in amount
of energy.
Actually, the amount of energywhich is used is increasing
exponentially.
So today is still manageable,but we can expect that in the

(02:42):
future it can become a problem.
So we try to solve this problemby using living neurons for
computations because they are 1million times more energy
efficient than digitalcomputers.
So that's, that's the reason whyit's important and why we
believe that it's a future forAI.
And why I work on this, thereason is because I'm

(03:03):
neuroscientist.
I have done research in brainimaging.
I was always fascinated withbrain.
And after I left academia, Istarted work in industry.
I started to discover new areasand I discovered artificial
intelligence.
I realized how important it is,how many crazy things you can
do.
I become fascinated.

(03:24):
I trained myself in the fieldand I was working a lot in the
commercial applications ofartificial intelligence and now
I work on step farther, youcould say, on the future of
artificial intelligence, whichis biocomputing.
So this is how we see future ofAI in around 10 years when we
can do, we can run artificialand intelligence algorithm on

(03:47):
the living neurons.

That just blows me away.
I remember hearing years agoabout using crystals or quartz
or diamonds and, and organicmaterial for computing.
So is that just my sciencefiction brain putting things
together that maybe haven't beenput together or is that also
like a part of, of biocomputing?

Well, it's not part of Biocomputing, but it's
part of bigger field, which iscalled"unconventional
computing".
And there are actually manyideas.
Also, quantum computing is oneof the type.
So unconventional computing iseverything different than
digital, what we have today.
So there are actually many ideasand people try to make
computation with fungi, withbacteria, with DNA.

(04:34):
So there are many, many ideas onhow to do computations in
different way, and we belong tothat field.

So the, the broader field of, what did you call it?

unconventional computing.

Unconventional computing.
Okay.

And then there are many, many different fields
inside, including biocomputingon living neurons.

Okay.
And so can you paint a picturefor me?
Are you there working with petridishes or like what, at what
scale are the, is the organicmatter?
And maybe today, what is it thatyou can have it do?
I have heard about, fungus, and,and what it can do.

(05:14):
Um, but I'm wondering, yeah.
What does that look like in, in,in your field specifically or
what you're working onspecifically?

So, at the moment it's everything is very
small.
We work on the 3D structures ofliving neurons, which we call
neurospheres.
So they're such a round block ofcells, 10,000 cells each.
We put them on the electrodes.
So at the moment is very,everything is small.
It's a small scale, and we tryto discover the basic algorithm

(05:44):
on how to program neurons.

Okay.
And, and when you say small, canyou give me a comparison?
Is it like a small, as a coin,smaller than that?
Do you have to use a microscope?

Smaller.
Definitely, so it's actuallylike half millimeter diameter is
one neurosphere, which we put onthe electrodes.
And actually, if you want to seethis visually, you can go on our
website, finalspark.com.
We have section live.
And there is a camera view fromour laboratory on how it looks.

And so looking into the future, you mentioned 10
years from now AI will be run onbio computers.
is that hard, like a hardprediction?
Or is that a wish?
Is it somewhere in between?

No, it's our estimation.
We have been made some specificplans for our research.
We are talking currently toinvestors.
We are seeking 50 million Swissfrancs of investment.
And we think that with thisinvestment we can accelerate our
research.
With investor, we plan to solvethe problem of learning in

(06:43):
vitro.
So how to teach neurons, somebasic algorithm in the next two,
three years.
After around three years foradvanced algorithm to match the
performance of digitalcomputers, and then around three
years for scaling.
So we want to make hugestructures, even a hundred
meters long of neurons, whichwill be so-called bio server.

(07:06):
It will be remotely availablecomputational power the same way
as today cloud computing.
So we assume it will take us 10years to arrive to this.

And you said that one of the advantages over the
energy usages of the computersrunning AI systems is some kind
of energy efficiency withBiocomputing.
How can that be the case?
Is it because everything is somuch faster or it uses less
energy to power the same kind ofcomputations?
And again, is this more of aguess or is this like based on

(07:38):
actual experiments that you'redoing in the lab?

Well, we cannot really measure this, this yet,
but there are some publicationswhich compare human brain to
digital computer.
And they estimate that humanbrain is around million time
times more energy efficient.
And we know also that tostimulate human brain, we would
need small nuclear plant.
And we can run on one banana forall day.

(08:02):
From this are these calculationsabout energy efficiency and of
course, in our bio computing atthis stage we can store one bit
of information.
So this is a little bit early tomeasure real efficiency of
Biocomputer.

Can you explain to me again what you were just
saying about a nuclear powerplant?
Are you saying it would take onenuclear power plant to make our
brains go?
Tell me.

No, no, no.
To simulate, to simulate whathappens in our brain with
digital technologies.

Ah, okay.
Because the human brain is somuch more energy efficient than
any computer.

Yes.
And actually there are a lot ofconsiderations about this.
I also wrote recently a blogarticle about this, which is on
my LinkedIn, comparing differentaspects.
So we know that brain isactually processing information,
encoding information, totallydifferent way comparing to
digital.
So brain is encoding informationin time and space, so it matters

(09:00):
when exactly and where in yourbrain neurons are active.
So this is totally differenttype of encoding than zero- ones
in computers.
Also, in brain you have a lot ofrecurrent connections, which
are, a little bit more difficultin digital, not so common.
And also, you have a lot offiltering information.
So we can treat in hierarchicalway information.

(09:25):
So not everything has the sameimportance like in the computer,
although there are somesolutions for that.
But, more or less in thecomputer, every information is
the same importance.
We can filter, we can focus onmost important things and let's
say ignore less important ones.
Also, we can, for example, whenwe analyze images with our

(09:46):
brain, we can only for example,we can detect.
So we can spend energy only whenwe detect the differences in our
field of view, while computerwould analyze all the pixels,
for example.
So these are just, there aremany, many examples on how brain
is more efficient also because,memory and computation happens

(10:07):
in the same, place in.
in the brain.
So that also increases theefficiency because you don't
have to spend energy on, on thechanging the location of the
information, like in, usualdigital computers.
So there are many ways, how youcan explain this energy
efficiency of the brain.

That's fascinating.
I'm very fascinated by the ideathat the brain is actually
making choices, about ourreality before we're consciously
aware.
That pattern recognition issolely based on our previous
experience.
And if we want to have newexperiences and new
understandings, we actually haveto input new things that we have

(10:46):
never seen before.
So our brain can make newpatterns.
So that essentially our brain isa predicting machine and all
coming down to being superenergy efficient.
One brain scientist talks aboutit being: our brain, it wants to
keep us alive.
Therefore, it wants to save asmuch energy as possible.
Therefore, it is making constantpredictions about every single

(11:07):
word I'm gonna say next.
Everything I might evenunderstand in terms of coding
emotion, what it means whensomeone makes a certain facial
gesture-- all of it.
It's all based on thispredictive model happening way
below our conscious awareness.
So is Biocomputing somehowlooking at that is, oh, how do

(11:27):
we take advantage of that?
I know it is early days yet, butthat's something that I'm
fascinated about.
So I was wondering how does thatapply, if at all, to the field
of Biocomputing?

Well, I think for now it's a little bit
difficult to compare because weare at the stage of one bit of
information.
So let's see what the futurewill bring and then we will see
with experiments.

Absolutely.
So, I know this is a briefinterview today and, and it is
such a vast topic.
So where can people find outmore about this work, of
biocomputing?

Our website is a good resource.
We have also some articles.
We have a lot of information,final spark.com.
We also have Discord account.
We build a community there.
There are a lot of technicaldiscussions also.
So I recommend this.
Also on our website, you canfind the link to Discord.
Also we have LinkedIn page and,X page.

(12:22):
So we try to communicate ourwork widely.
And we also talk a lot topodcasters and journalists
because we would like that thisnew field of biocomputing is
more recognized in the society.

That's really wonderful.
And as I, I told to you before Ieven started recording, I have a
nephew who wants to go into thefield of biotech and for all the
young people who may belistening to this, what
recommendations do you haveabout where to put their
attention in terms of, of theireducation?
I know my, my nephew is doinghis biology and chemistry as

(12:58):
well as computer sciences.
So yeah.
any recommendations for theyounger generation who are
curious about thisunconventional computing and
specifically biocomputing?

So I always recommend people to follow what
they're interested in becausethe future is unpredictable and
we cannot really know what willbe the jobs and what will be the
market in 10 years or 20.
So I think the surest bet is tofollow what you're interested
in.
And then you can always changeand you can learn something new,
some new skills if there will benecessary.

(13:31):
But if you do what you like,it's much easier also to learn
new stuff and to keep going.
It's much more pleasant.
So I really recommend to followthe interest.
And, if someone is specificallyinterested in biocomputing, they
are recommended to check ourpaper in Frontiers.
And it's a blend of biology andengineering.

(13:52):
So there is also such a trend,maybe in education.
And also you can see this inresearch and in private
companies there are more andmore projects which combine
biology and engineering, and Ithink it's quite hot now.
However, still I reallyrecommend to do what someone is
interested and enjoy because thetrends can always change.

(14:13):
So we don't know, but there is alot, currently a lot of
interesting projects oncombining biology engineering,
so people who understand biologybut can also code, code and
maybe build some hardware.
So that's interesting mix.

I really love that advice.
I wish someone had handed memore engineering skillsets when
I was younger.
I had the mindset of a builder,but I didn't have the skillset.
And I mean now I am a builderof, of media and podcasts and
film and books and courses, etcetera.
So, I want to say, let me putmyself here in the spotlight.

(14:48):
Ewelina Kurtys, thank you somuch for being a guest on How To
Write The Future.
I think your insights and inputwill be so fascinating to both
writers and non-writers andanyone who cares about the
future.
So thank you.
Thank you so much for being withus today.

Thank you too.

Everyone.
Write long and prosper.

All Episodes

Episode Transcript

Popular Podcasts

Dateline NBC

Stuff You Should Know

My Favorite Murder with Karen Kilgariff and Georgia Hardstark

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}171. What is Biocomputing? Interview with Ewelina Kurtys

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Dateline NBC

Stuff You Should Know

My Favorite Murder with Karen Kilgariff and Georgia Hardstark

All Episodes

171. What is Biocomputing? Interview with Ewelina Kurtys

Dateline NBC