March 10, 2025 • 39 mins
Artificial Intelligence is on the rise – even this description could be written by AI! But what does AI think about Jews? In this episode, Tami and Dash talk to Afro-Brazilian, Aussie-Jewish writer and academic Guido Melo about the risks of AI stereotyping and prejudice.
Articles related to this episode:
https://thejewishindependent.com.au/what-generative-ai-thinks-about-jews
https://thejewishindependent.com.au/what-travel-taught-me-about-my-identity-as-a-black-jew-ish-man
Email your feedback and voice memos here: ashamed@thejewishindependent.com.au
Subscribe to The Jewish Independent's bi-weekly newsletter: jewishindependent.com.au
Tami and Dash on Instagram: tami_sussman_writer_celebrant and dashiel_and_pascoe
X: TJI_au
YouTube: thejewishindependentAU
Facebook: TheJewishIndependentAU
Instagram: thejewishindependent
LinkedIn: the-jewish-independent
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:07):
Are you interested in issues affecting Jews in real
life and online?
But a little bit ashamed thatyou're barely keeping up to date
.
Speaker 2 (00:15):
Well, you've come to the right place.
I'm Dash Lawrence and in thispodcast series, your spicy third
cousin, Tammy Sussman, and Icall on experts and each other
to address all the ignorantquestions that you might be too
ashamed to ask.
Speaker 1 (00:31):
Join us as we have a go at cutting through some
seriously chewy and dewy topics.
Speaker 2 (00:37):
Welcome to the Jewish Independent Podcast.
Ashamed to Admit, a shame toadmit.
Hello everyone, I'm Dash.
Speaker 1 (00:54):
Lawrence, executive Director here at the Jewish
Independent, and I'm Tammychronically ill, but make it
cute, sussman.
I've spoken about my arthritison this show before, but I
haven't spoken about theimmunosuppression, have I?
It's a little bit boring.
Speaker 2 (01:09):
No, I'm interested to learn more.
Speaker 1 (01:12):
Are you actually?
Speaker 2 (01:14):
Does that sound disingenuous?
No, definitely.
I want to hear what's happeningfor you this week.
Speaker 1 (01:20):
Members of the Jewish community have autoimmune
issues, and mine is autoimmunearthritis, inflammation, pain,
fatigue, and so I takemedication for that.
And the great thing about themedication is it helps my pain
and my fatigue.
And the bad thing about themedication is that it suppresses
(01:44):
my immune system.
So I catch everything Every dayis do I want to be in pain and
exhausted or do I want to have apersistent cough?
This week I've chosenpersistent cough.
Luckily for you, Dash, it meansthat I'm not really able to
talk much today.
(02:05):
I've prepared something for youto read.
I wanted it to be a surprise.
Speaker 2 (02:12):
Okay.
Speaker 1 (02:13):
Okay, so I'm sending it to you now.
I'd like you to read the bodyof the email in its entirety.
Speaker 2 (02:22):
So email's arrived in , tammy.
Thank you for sending it.
So email reads hello Dash, Iwanted to honour your Irish
ancestry and the Jewish familyyou've created with your Jewish
partner, so I asked AI to writea limerick about you.
Can you please read it now?
Thanks so much, tammy.
Speaker 1 (02:44):
Okay, before you read the limerick, I did put partner
in capital letters, because youalways correct me when I say
wife, you say partner.
Speaker 2 (02:55):
It's correct, we're not married.
Speaker 1 (02:56):
Because you're super cool, you live in North
Melbourne.
People don't get married,they're just partners.
Speaker 2 (03:04):
So the limerick goes.
There was once a fellow namedDash whose love for the Jews was
no clash.
He'd donate his cash and joineach festive bash and stand with
them, bold in a flash.
Tammy, that is terrible.
I mean who, what, where, why,why?
Speaker 1 (03:27):
Why did I get ChatGPT to write a limerick about you?
Speaker 2 (03:37):
Yes, why do you think Today's?
Speaker 1 (03:38):
episode is all about AI Mm-hmm, Specifically how AI
imagines Jews.
Speaker 2 (03:45):
There are many great advantages to AI, I think both
of us also see some of itslimitations and some of its
problems.
In fact, I'm now recalling that, by way of example, in season
two, when we had David Baddielon the program, you shared with
(04:08):
David and the listeners and ifyou haven't heard that episode,
you must do One of our greatepisodes to date but at the end
of that conversation you sharedwith David an image that you
asked ChachiBT to create of himgolf and bringing in Jewish
elements.
And what did AI spit out?
Speaker 1 (04:30):
A few golf balls and some cats.
And to represent his Jewishness, there was no menorah, there
were no Shabbat candles, therewas no Sidor or Talit prayer
shawls, it was just moneycandles.
Speaker 2 (04:45):
There was no Sidor or Talit prayer shawls.
It was just money, just money,just money.
Astounding.
So clearly there are some realproblems with AI.
It's the way that it sourcesits information and then the
types of stereotypes that itperpetuates.
This is a problem because inyears to come, we're going to
see an increasing number of AIimages used across our news
(05:11):
media, across the internet, Imean it will be ubiquitous.
Speaker 1 (05:14):
That's right.
In fact, after I published thatreel where I exposed the
problematic image of DavidBaddiel that AI had created, one
of our listeners, Guido Mello,commented and said that this is
what his research is zoominginto.
Speaker 2 (05:34):
Guido has recently published an article on the
Jewish Independent this is howAI Imagines Jews, and in that
article Guido shares hisresearch not only on the way
that AI generates stereotypical,anti-semitic images and ideas
of Jewish people, but also theway that people respond to those
(06:00):
images and the ways that theimagery reinforces some of the
biases and the perceptions thatpeople have.
So really encourage ourlisteners to take a look at
Guido's article, but for thetime being, have a listen to our
(06:20):
conversation with Guido Mello,who is a master's researcher
from Victoria University here inMelbourne.
He's a published author and amultilingual columnist in
Portuguese, spanish and English.
A Jewish community membercompleting his conversion, guido
focuses on AI erasure andbiases in artificial
(06:41):
intelligence.
His writings can be found acrossmultiple countries, including
Australia, the Americas andAfrica and, of course, in our
humble little publication, theJewish Independent.
Enjoy this fascinatingconversation about the scary,
dark but ultimately hopefulworld of AI.
(07:03):
Guido, we're going to talkabout a recent article of yours
(07:29):
in the Jewish Independent, whereyou shared with us some really
fascinating research looking atstereotypes and prejudice and
anti-Semitism in AI, a newemerging field.
But before we get there, tellTammy and I and our listeners
about your unconventional pathto life in the Australian Jewish
community.
Speaker 3 (07:48):
I was born in Brazil and I was born in a city called
Salvador.
Not like the country, there'sno L, so it's just Salvador, and
it's in the northeast of Braziland it's the most African city
outside of Africa.
You know most African cityoutside of Africa.
You know like 75% of thepopulation is African
(08:08):
descendants.
And for people who doesn't know, all Brazilians, africans or
black people really came becauseof the transatlantic slavery
trade.
And another thing that peopledon't know is that most of the
humans that came from Africaenslaved to the Americas, came
(08:30):
to Brazil.
The population of the UnitedStates of Africans is about 50
million.
In Brazil it's about 120million.
Speaker 1 (08:38):
Wow, I didn't know that.
Speaker 3 (08:39):
Yeah, so if Brazil was a country in Africa, it
would be the second largest,just behind Nigeria.
Speaker 2 (08:46):
Do you know much about the story of your
ancestors and when and how theyfound their way to Brazil?
Speaker 3 (08:52):
No, absolutely not.
So what happened is there wasan event which is kind of a
history erasure, which issimilar to what I studied in my
research.
History erasure which issimilar to what I studied in my
research.
This Brazilian presidentdecided to eliminate all of the
traces of slavery to show thatit was important for community
(09:15):
to move on.
So he burned all the documents.
Then we can all start like akind of an year zero, like Pol
Pot tried to do in Cambodia.
And he did that.
So it burned out the documents.
On my father's side, which isentirely West Africa, on my
mother's side, funnily enough,my grandfather, who is a white
(09:36):
Portuguese guy, apparently hadblue eyes.
So my mom says, and his name isOliveira, 50% of the Oliveiras
are Jewish converts who, afterthe Inquisition, were forced
into conversion to Catholicism.
So you know, maybe I'll startthere.
Speaker 2 (09:56):
This is going to become particularly relevant
when you tell us about how it isthat you came into the Jewish
community and into Judaism,because you weren't born Jewish,
as we'll soon discuss.
Tell us about your upbringing.
Speaker 3 (10:09):
This is a shame to admit moment.
Recently, the Brazilian filmindustry won an Oscar for a
movie called I Still here, orAinda Estou Aqui in Portuguese,
and the movie depicts adictatorship in Brazil that
began in 1964 and lasted until1985.
(10:29):
So I lived, actually livedthrough a dictatorship, you know
, a military dictatorship, andmy shape to the mid-moment is,
paradoxically, my father, who isa very left-wing Marxist guy
who taught me like when I was 13, he gave me Karl Marx to read
because he's like oh, you needto read the manifesto, the
(10:51):
communist manifesto, and I did.
He was from the Air Force, myfather was from the Air Force,
so the military dictatorshipthat oppressed Brazil and
oppressed people like me wasalso what defended me or
protected me because, as I said,my father was in the Air Force
(11:14):
during the dictatorship of themilitary.
It's important to note for mepersonally and for the record
because this is a record thatmight last forever that my
father often told me how he didnot participate in the
dictatorship and he refused toengage and he was very left-wing
, and I think that's importantfor me, you know, for his record
(11:35):
.
But we were part of thedictatorship, which is so
connected to what Israel istoday.
For me, because things are socomplicated and people want an
easy answer, it's like oh,you're part of the dictatorship,
you're the bad guy, and you'renot part of the dictatorship,
you're the good guy.
And for me, that's why maybefor me it's so easy to
(11:57):
comprehend Even with someonethat didn't grow up with, you
know Israel in the forefrontthat yes, there's things that we
can criticize, but within thatsystem that clog, there's humans
, and all these humans havethese particular stories.
And, yeah, I don't see thingsas bad and as good, and I often
(12:21):
tell my kids there's no good andthere's no bad.
You have to remember that, thatno one is really good.
And I often tell my kidsthere's no good and there's no
bad.
You have to remember that, thatno one is really good, no one
is bad.
People live circumstantiallyand they will do bad things.
Speaker 1 (12:34):
Guido, that's very wise words, but we've skipped
over a little bit.
We still don't know how youcame to Australia and found
yourself in Melbourne's Jewishcommunity.
Speaker 3 (12:47):
Talk us through that so when I lived in Rio, I used
to work in a cyber cafe fixingcomputer printers and plugging
people to the internet, and Imet this young, well, ashkenazi
girl.
At the time we were both veryyoung and, yeah, we got together
(13:11):
.
Speaker 1 (13:12):
So she was in Brazil.
Speaker 3 (13:13):
She was visiting Brazil.
She was, you know, doing awalking holiday from London,
from England.
She came to Rio, met me.
I convinced her to stay.
She lived there for almost ayear, over nine months, and then
we decided to get engaged, youknow, and we chose, you know
like, we could live here, wecould go back to Australia.
(13:35):
She said I have to finish myuniversity.
We came back to Australia andfrom that moment on I was a
member of a Jewish family.
You know, and I don't know,like, maybe a few months later I
was going to a bar mitzvah withone of the cousins and I went
to that.
There's a beautiful synagoguein Melbourne, Turak, I think
(13:56):
it's called the MelbourneSynagogue.
It's like a very big dome.
Speaker 2 (13:59):
This is Turak Shul.
I think it's the second oldestsynagogue in Melbourne.
Speaker 3 (14:04):
yeah, yeah, on Turak Shul, I think it's the second
oldest synagogue in Melbourne.
Yeah, on Turak Road, and I wentto a Bahamian for that and I
was like, oh, I like this and Ireally you know, I can't explain
it.
Well, I think the simplest wayis that my father, he was really
an admirer of Jewish thinking.
We read Kafka, or you know Isaid Marx, and he always talked
(14:27):
about Jewish composers and wewatched many Holocaust films and
he was very interested.
There was very a big connectionbetween.
He could see parallels betweenthe African, black suffering and
the Jewish suffering, so he wasvery interested.
Speaker 2 (14:45):
You're in Melbourne.
This is in the 1990s, I imagine.
Speaker 3 (14:48):
No, no, no, early 2000s, so 2003.
Speaker 2 (14:52):
Okay, the early 2000s , but certainly I don't imagine
there were a lot of newlyarrived Afro-Brazilians in the
Jewish community at that time.
Speaker 3 (15:02):
No, there wasn't and there isn't today.
Speaker 2 (15:04):
Yeah.
Speaker 3 (15:05):
I'm like the single African guy in my congregation,
which is fine.
Speaker 2 (15:11):
Sometimes the community is accused by itself
of being exclusionary, beingdiscriminatory to some members
of the diverse Jewish communitythat we have in Melbourne and in
Sydney.
Was that your experience in theearly 2000s?
Speaker 3 (15:26):
Well, the way I see it is, racism always exists in
all aspects of my life, but Idon't have any case where I was
singled out in my congregationor even every congregation I've
been because I was black.
But I know, you know I can'tcontrol individuals.
You know, like I remember therewas this guy that was looking
(15:47):
at me once at shoe Not the shoeI currently go under, one of the
shoes.
I've been to many and he waslooking at me.
I think I said Shabbat Shalomor whatever, and he's like
didn't reply and I was like,okay, you know, whatever, you
know like, you know like it's,but I it's, it's just one person
.
But yeah, so I I reallyconnected to my wife's ex-wife
(16:08):
she's at next now, but we'resomewhat friendly because we
shared three kids, but yeah, Ilove the Shabbos.
I was like everyone was yellingat each other.
She was yelling at her mom andshe was like, wow, this is, I
wanted that.
I like that.
I tell people I joined Judaismfor the challah and for the
latkes and then I got October7th, you know, like as a side.
Speaker 1 (16:34):
As a side dish.
Speaker 3 (16:35):
And I always felt connected.
I realized later on I had a lotof Jewish friends.
I basically was Jewish already,without you know, having done
the conversion, and I found thisyou know, great rabbi, and
that's sort of that's thejourney I'm living now, but it's
like I really felt that wasalways part of my life.
(16:56):
And my rabbi says you know.
Rabbi Allison says you know,you're just returning.
You're just returning and Ifeel like that you know.
Speaker 1 (17:07):
So we've heard a little bit about your context
and your relationship to Judaismand Jewishness, which is
important for our listeners toknow, especially when it comes
to the work that you're doingnow, and that is in academia.
So what originally led you toacademia?
Speaker 3 (17:31):
Well, I was always a reader.
I loved nonfiction.
I'm really interested in theways the word works and the
system works and power anddynamics and gender and race and
history.
And when I moved to Australia,I only had year 12.
So I completed year 12.
(17:53):
And my ex-wife and I, we openeda business and I started to make
money out of this businessselling women's clothes and
women's jewelry from SouthAmerica, mostly from Brazil, but
also from Argentina, peru,chile and I remember talking to
my father.
He passed away in 2017, butthat's like you know, 2005 or
2006.
(18:13):
And I said to him you know, Imoved to Australia.
Now I could study, but thewhole point of study for me was
to make money.
And I said to him I'm alreadymaking money, but, yeah, so in
2019, as my relationship reallywent south, like I realized we
were divorcing, I thought youknow what I always wanted to do?
(18:37):
This?
I'm not really into fashionanymore.
I want to do academia, but Istill was thinking I should do
one degree, one for me and onefor the business.
So I did a diploma ofprofessional writing and editing
, which I could use forcopywriting for my website at
the time, and then I did thebachelor's of digital media and
(18:57):
writing, and digital media iswhere I started to use, you know
, the tools that I could use formy business and I realized
really digital media was veryEurocentric.
Like I remember, I went to aclass and this woman was
teaching us about film and shesaid the first films were made
(19:19):
da, da, da, da da.
And she was telling somethingabout Europe and she never
mentioned Bollywood.
And it's like really, you gotpaid to teach film and you're
not even going to mentionBollywood, not even going to
mention polio, not mention, noteven sure, not teach, but not
mention.
So this is eurocentric.
And I was typing my tiktok andI saw the, the filter, and the
(19:40):
filter showed me my features butlike it sort of thinned my nose
and made my eyes green and mademy head straight and and I was
like here is the most beautifulversion of you.
And I was like no, that's notbeautiful, you know for me, and
that's what led me to theresearch.
(20:02):
I was like what else AI istelling people?
That's the most beautiful?
It's right or wrong.
Maybe even two years ago Ityped Jewish people.
Two years ago it was likepeople in grey with sad faces.
That was Jewish people.
Speaker 1 (20:19):
So you're saying, if you typed into something like
chat GPT, show me an image of aJewish person, it would feed
back.
Something to you that lookedlike it would feed back
something to you that lookedlike Two years ago.
Speaker 3 (20:33):
It looked like Holocaust figures and sad people
, people wearing grey and skinny, and it was really horrible.
Now, two years later, that wasin the beginning.
Now, two years later, itimproved, but then it became
(20:56):
ultra-stereotype into, like youknow, bearded men.
So basically all Jews areAshkenazis.
Speaker 1 (21:00):
According to AI, according to AI.
Speaker 3 (21:03):
Most Jews are men, like basically they don't have
women.
And it's funny because if youtype family, they would put two
guys.
They don't have women.
And it's funny because if youtype family, they would put two
guys.
They would put a guy andanother guy, and a girl and a
boy or a guy and a guy and a boy.
But they wouldn't put a woman,like they couldn't.
The AI cannot see a Jewishwoman unless if you say Jewish
(21:25):
woman, if you say Jewish family,it thinks it's all men.
And if you put Zionists, youhave this army looking, people
looking very bad at the camera.
Speaker 2 (21:37):
They look aggressive.
We encourage the listener totake a look at some of the
photos that Guido unearthedthrough his research, and we'll
put the link to that in the shownotes.
But, guido, can you just help?
Tammy and I understand themechanism that sits underneath
AI, so if you were to look underthe hood, what exactly is
running it that generates theimages that we see?
(22:01):
I use AI frequently, but I'veactually got really very little
understanding and knowledge ofwhat it draws on.
Speaker 3 (22:10):
Yeah, it's really not transparent.
Exactly what goes inside ofthat jush?
It's basically a big cooking ofinformation that is collected,
scavenged, really, over theinternet, but what we know based
on whistleblower here orcomments there or kind of paper
(22:32):
here and there is that a lot ofthe data that was trained by
ChatGPT or Midjourney, which isthe visual software that I use
in my research, is informationaround the internet.
What happens is, because it'sanglophone, so it uses mostly
English-speaking information andit's mostly focused on Europe
(22:57):
and the North sort of globalNorth, which I don't like that
term, but you know by likepeople that are European and
European descendants.
Most of the information comesfrom those places.
It also uses, like even thingslike Reddit and obscure blogs.
So what happens is you'regetting all these male-centric,
(23:18):
chauvinistic, at timesEuropeanized views of the world.
Again, it's very important forme personally not to say AI is
bad or even that Europeans arebad or whites are bad.
That's not what I'm saying.
I'm saying that the AI toolsthey utilize, they center those
(23:43):
people despite the existence of,you know, despite Europeans and
really white people being aminority in numbers in the
global, even though thesecountries are the leading
countries in the world in manyaspects, in the industrial
aspects, every search orsomewhere saying most of our
(24:15):
data is Eurocentric, it wastrained by males and those
buyers will appear.
If you want to get somethingdifferent, you have to be more
specific.
And just a disclaimer so peopleunderstand.
Speaker 1 (24:27):
In your article you wrote that this idea of a
homogenized Jewish people isdangerous and, in your view,
anti-Semitic.
Can you speak more to that andthe negative repercussions of
presenting Jews in that way?
Speaker 3 (24:47):
Yeah, so the reason this idea is anti-Semitic is
because I can't prove it, butit's very likely that, with the
information we have, that the,the diaspora, are the main
drivers of the narrative aboutJewish people.
Again, there's not a problemwith that per se.
(25:27):
I'm not combating that.
That could be a differentconversation.
I'm more saying because that'sthe information out there,
that's what these tools areutilizing to create the big Jews
, and also because, historically, jewish men have had more
opportunities to have theirvoice heard in many aspects.
(25:48):
And I think the dangers in thatis that it shows us as Jewish
people, as one specific type,with one specific gender, with
one specific idea.
And if there's one thing that Iknow about being connected to
the Jewish community for overtwo decades is that we are so
different, we look so different,you know, not just physically.
(26:09):
You know I met Black Jews, Imet Indian Jews, I met Asian
Jews.
You know I met Asian Jews, Imet disabled Jews, I met queer
Jews and tall short, all sortsof people.
And if that's the image that'sbeing shared to the world,
especially with Gen Zs, ifthere's a study that says 46% of
Gen Zs rely on AI to tell themwhat to do or what to say.
(26:33):
And if that's the image thatthey get, homogenization and
stereotype can what to say?
And if that's the image thatthey get, homogenization and
stereotype can lead to violenceand violence can lead to harm to
us People if they see us asthese people that are Zionists,
men, beaded, whatever it's, justlike it kind of would lose our
(26:56):
own self-identity, our idea ofsecrecy, and it becomes easier
to harm us, to attack us.
Speaker 1 (27:01):
I think and we've seen that happening recently,
when there's been a lot ofdenial about Jewish people being
Indigenous to the land ofIsrael and that's because many
people just see Jews asAshkenazi inverted commas white
and they don't actually see thefull picture.
Speaker 2 (27:25):
Guido, we should mention that the research you've
done has been conducted throughshowing these images to
research participants.
Showing these images toresearch participants.
So tell us about how theparticipants responded to the
(27:46):
images of these stereotypicalAshkenazi bearded Jewish men.
What was the response?
Because this is the reallyintriguing part of your research
.
Speaker 3 (27:56):
What my research found is, as the bias embedded
on AI tools, such as generatedimage systems, increases, so
does the bias on the usersincrease.
And I think this is a bigdanger because if people believe
(28:18):
what they've seen, becausethere's an underperformization
of AI tools, because they speakto me I have an AI, I talk to
Chachi Pichu often, you know,had a proper conversation with
it and unless, if you are reallyswitched on and analytically
(28:38):
thinking critical, you willbelieve it's a human that's
telling you information orthere's an anthropomorphization
of it and people believe inthose informations.
And for me, what my researchshows showing images to people
is that they became more racistand more stereotypical and
accepting.
(28:59):
I don't see even a mitigationof it because the companies have
no intention of retrainingtheir materials and when they
try to do mitigation, they mightremove entire words that need
to exist for historical context.
For example, the New York Times,this African-American artist.
(29:19):
She was doing a work withenslaved ships and she couldn't
Google the word enslaved becausethe word enslaved got disabled.
So if you disable the wordHolocaust, for example, because
you wanted to protect Jews,let's say you want to do, for
example, because you wanted toprotect Jews, let's say you want
to do a good thing, you want toprotect Jews against people
(29:41):
utilizing that, but then if youdisable that word, you erase the
entire history, and that'swhere AI erasure comes from.
The mitigation won't solveanything.
Disclaimer I think ourgovernment should take AI.
As you know, we build roads.
We should be building AIregulated by government and with
(30:06):
support of institutions,universities, like we should
have AI.
That's an Australian AI,because it's like it could have,
for example, people saying AIwould say Australia began in
1788.
And it's like sure, but youshould add European Australia
started in 1788.
Speaker 1 (30:24):
I'm glad you touched on that, because I wanted to
talk about something that youmentioned in another interview
you did on I think it was theABC when you mentioned the
Alicia Keys Superbow Bowlincident.
Speaker 3 (30:36):
Yes.
Speaker 1 (30:37):
So just to enlighten listeners who aren't aware, it
was last year's Super Bowl.
Speaker 3 (30:42):
Last year's Super Bowl 2024.
Speaker 1 (30:45):
2024 Super Bowl.
Alicia Keys was the halftimeperformer.
Those who were listening livewould have heard that there was
this lovely human moment whereone of her notes was slightly
off.
It's fine.
Everyone moved on.
Then, after the event, thehalftime performance was
replayed.
(31:05):
It was perfect.
So whoever reconstructs thathad used AI to fix her off note
and then you spoke about that inthe interview as some kind of
AI erasure.
Speaker 3 (31:21):
Yes.
So AI erasure is a conceptthat's tangential for my
research that's going to come ina paper that I'm writing
currently and I've been writingarticles in different places
which is the idea that, becauseAI is so pervasive, the
individuals, institutions,governments can use AI tools and
(31:42):
generative images generated bywant to erase their
contributions to the RomanEmpire.
Stalin removed adversaries fromphotographs because he wants to
(32:04):
erase those contributions tothe adversaries, so that always
existed.
So AI erasure what happenedwith Elissa Keyes is that she
missed a note and then used AItools to recreate her voice, and
what happens in thesesituations is that the public
can't go to history.
It's a lie, it never happened,and what would that be for the
(32:26):
future?
Speaker 1 (32:27):
But you mentioned that AI should be regulated by
governments.
My Jewish fear is that what ifwe had a government that was
racist against Jews and decidedto erase the Holocaust from
their AI and to have Holocaustdenial or other forms of racism
(32:49):
against Jews in Mizrahi,sephardi community denial?
Could that happen?
Speaker 3 (32:54):
Yes, it could happen and we've run the risk of it,
but right now it's either theAustralian government or a long
mask and it's like, unlessthere's a grassroots movement,
perhaps the worldwide Jewishcommunity could create their own
(33:14):
AI, but then again that couldprobably potentially be accused
for being biased.
Let's say, maybe for our peoplewe would have to.
Really, when I say governments,I shouldn't, I don't have the
intellectual capacity or theskills capacity to advise on the
solution, but what I'm sayingis that someone that's
(33:40):
responsible, that has acollective well-being as humans
maybe the United Nations, unesco, I'm not sure who will create
an AI tool that is collective,that has a contribution from
different diversities, and itwill, you know, and it accepts
(34:01):
you know.
I think the disclaimer issomething that I would like to
see.
Despite our best efforts, it ispossible that some perspectives
were missing, and this isimportant to note when reading
this information on track tptokay, don't talk yourself down.
Speaker 1 (34:20):
I think you do have the capacity to be an advisor on
this committee.
I'm thinking maybe it shouldnot go through the united
nations, because then theremight be a bit of october 7
denial.
Speaker 2 (34:32):
Oh, she threw shade on the united nations guido, I'm
interested in what you make ofhow quickly things are changing
and if we were possibly to fastforward five years, what is a
positive case scenario with AIand its generation of images?
(34:54):
If we were to be optimistic,what could we see, maybe in five
years' time?
Speaker 1 (34:58):
Good question, Dash it is a great question.
Speaker 3 (35:01):
There's this collective in New Zealand who
uses AI to recreate Maorilanguage and find, because
there's words that disappearedand then they find ways, because
they fed the AI with Maorilanguage.
They've recreated, they'verebirthed this language.
So AI is great and I think itwill improve our productivity.
(35:27):
We're still going to have to becreative, you know we still
have to do, but AI can be usefulfor us, but with more
contribution from diverse voices.
Speaker 2 (35:36):
In the worst-case scenario, the stereotypes, the
racist imagery just continues toget worse and worse.
Speaker 1 (35:43):
I don't know if I want to know worst case scenario
.
It's too much.
Speaker 3 (35:47):
No, I think it's just like.
The worst case scenario is whathappened in the beginning of
January this year, 2025.
Like Facebook, filters wentdown, you know, and then now
anyone can say anything.
So the worst case scenario isone we can have sponsorship
research, so people, you chargeon, charge TPT, but people get
(36:12):
sponsored AI answers and youcould drop the filters because
right now there's filtersmitigating racism, mitigating
anti-semitism.
You drop those.
That's the worst case scenario.
I don't want to put fear intopeople.
I think we'll survive all theautocrats out there and we'll
(36:35):
have Shabbat.
We'll eat challah five yearsfrom now.
Ai is just going to be a tool.
What was a tool?
Microsoft is a tool.
Photoshop is a tool.
Remember in the beginning ofPhotoshop?
Oh my God, people are makingwhatever, someone is skinny
whatever.
There was this panic and thingssettled in and people
(36:56):
understand that these tools cometrue, you know, for marketing
purposes.
So I'm not an AI iconoclast.
Speaker 2 (37:05):
No.
Speaker 3 (37:06):
I'm an AI encourager, but what I'm saying is that we
have to have critical thinkingwe have to have mitigation
strategies Bad boundaries.
Speaker 2 (37:16):
I'm envisaging a future, Guido, where I can go to
AI and say create me an imageof an Afro-Brazilian working in
an internet cafe in Coco CabanaBeach and this Afro-Brazilian is
Jewish.
I'd like to see what you comeup with, AI, and I get something
(37:37):
approximating you, Guido.
Ai doesn't kind of go no, can'tcompute, that doesn't make
sense.
A future scenario where you cancreate something that
approximates the diversity, themultiplicity of what it means to
be Jewish yeah.
Speaker 3 (37:53):
And for me, like it's a privilege, you know, to be
Jewish, it's reconnecting wellto my soul, as my rabbis say, or
a connection to who I want tobe and who I am, and it's
beautiful to be diverse.
I have these beautiful kids who, you know, have African
ancestry, brazilian indigenousancestry, portuguese, possibly
(38:15):
Jewish ancestry and born inAustralia, and I think we should
be grateful for it and I hopethat the community is embracing
and you know I feel welcome inthe synagogues I've been in, I
hope I continue to be and I hopethat more people from other
(38:36):
racial groups can be part of theJewish community in.
Speaker 1 (38:37):
Australia.
Guido Mello, thank you so muchfor joining us on A Shame to
Admit.
Speaker 3 (38:43):
Thank you, it was my pleasure.
Speaker 2 (38:45):
That's it for another week.
Speaker 1 (38:47):
You've been listening to.
A Shame to Admit with me TammySussman and executive director
of TJI, dr Darsha Lawrence.
Speaker 2 (38:54):
This episode was mixed and edited by Nick King,
with theme music by DonovanJenks.
Speaker 1 (38:59):
If you like the podcast, leave a positive review
, tell your people or encourageyour third cousin's cousin to
advertise on the show.
Speaker 2 (39:08):
You can tell us what you're ashamed to admit via the
contact form on the JewishIndependent website or by
emailing ashamed atthejewishindependentcomau.
Speaker 1 (39:18):
As always.
Thank you so much for yoursupport and look out for us next
week.
Thank you.
Are you interested in issues affecting Jews in real
life and online?
But a little bit ashamed thatyou're barely keeping up to date
.
Speaker 2 (00:15):
Well, you've come to the right place.
I'm Dash Lawrence and in thispodcast series, your spicy third
cousin, Tammy Sussman, and Icall on experts and each other
to address all the ignorantquestions that you might be too
ashamed to ask.
Speaker 1 (00:31):
Join us as we have a go at cutting through some
seriously chewy and dewy topics.
Speaker 2 (00:37):
Welcome to the Jewish Independent Podcast.
Ashamed to Admit, a shame toadmit.
Hello everyone, I'm Dash.
Speaker 1 (00:54):
Lawrence, executive Director here at the Jewish
Independent, and I'm Tammychronically ill, but make it
cute, sussman.
I've spoken about my arthritison this show before, but I
haven't spoken about theimmunosuppression, have I?
It's a little bit boring.
Speaker 2 (01:09):
No, I'm interested to learn more.
Speaker 1 (01:12):
Are you actually?
Speaker 2 (01:14):
Does that sound disingenuous?
No, definitely.
I want to hear what's happeningfor you this week.
Speaker 1 (01:20):
Members of the Jewish community have autoimmune
issues, and mine is autoimmunearthritis, inflammation, pain,
fatigue, and so I takemedication for that.
And the great thing about themedication is it helps my pain
and my fatigue.
And the bad thing about themedication is that it suppresses
(01:44):
my immune system.
So I catch everything Every dayis do I want to be in pain and
exhausted or do I want to have apersistent cough?
This week I've chosenpersistent cough.
Luckily for you, Dash, it meansthat I'm not really able to
talk much today.
(02:05):
I've prepared something for youto read.
I wanted it to be a surprise.
Speaker 2 (02:12):
Okay.
Speaker 1 (02:13):
Okay, so I'm sending it to you now.
I'd like you to read the bodyof the email in its entirety.
Speaker 2 (02:22):
So email's arrived in , tammy.
Thank you for sending it.
So email reads hello Dash, Iwanted to honour your Irish
ancestry and the Jewish familyyou've created with your Jewish
partner, so I asked AI to writea limerick about you.
Can you please read it now?
Thanks so much, tammy.
Speaker 1 (02:44):
Okay, before you read the limerick, I did put partner
in capital letters, because youalways correct me when I say
wife, you say partner.
Speaker 2 (02:55):
It's correct, we're not married.
Speaker 1 (02:56):
Because you're super cool, you live in North
Melbourne.
People don't get married,they're just partners.
Speaker 2 (03:04):
So the limerick goes.
There was once a fellow namedDash whose love for the Jews was
no clash.
He'd donate his cash and joineach festive bash and stand with
them, bold in a flash.
Tammy, that is terrible.
I mean who, what, where, why,why?
Speaker 1 (03:27):
Why did I get ChatGPT to write a limerick about you?
Speaker 2 (03:37):
Yes, why do you think Today's?
Speaker 1 (03:38):
episode is all about AI Mm-hmm, Specifically how AI
imagines Jews.
Speaker 2 (03:45):
There are many great advantages to AI, I think both
of us also see some of itslimitations and some of its
problems.
In fact, I'm now recalling that, by way of example, in season
two, when we had David Baddielon the program, you shared with
(04:08):
David and the listeners and ifyou haven't heard that episode,
you must do One of our greatepisodes to date but at the end
of that conversation you sharedwith David an image that you
asked ChachiBT to create of himgolf and bringing in Jewish
elements.
And what did AI spit out?
Speaker 1 (04:30):
A few golf balls and some cats.
And to represent his Jewishness, there was no menorah, there
were no Shabbat candles, therewas no Sidor or Talit prayer
shawls, it was just moneycandles.
Speaker 2 (04:45):
There was no Sidor or Talit prayer shawls.
It was just money, just money,just money.
Astounding.
So clearly there are some realproblems with AI.
It's the way that it sourcesits information and then the
types of stereotypes that itperpetuates.
This is a problem because inyears to come, we're going to
see an increasing number of AIimages used across our news
(05:11):
media, across the internet, Imean it will be ubiquitous.
Speaker 1 (05:14):
That's right.
In fact, after I published thatreel where I exposed the
problematic image of DavidBaddiel that AI had created, one
of our listeners, Guido Mello,commented and said that this is
what his research is zoominginto.
Speaker 2 (05:34):
Guido has recently published an article on the
Jewish Independent this is howAI Imagines Jews, and in that
article Guido shares hisresearch not only on the way
that AI generates stereotypical,anti-semitic images and ideas
of Jewish people, but also theway that people respond to those
(06:00):
images and the ways that theimagery reinforces some of the
biases and the perceptions thatpeople have.
So really encourage ourlisteners to take a look at
Guido's article, but for thetime being, have a listen to our
(06:20):
conversation with Guido Mello,who is a master's researcher
from Victoria University here inMelbourne.
He's a published author and amultilingual columnist in
Portuguese, spanish and English.
A Jewish community membercompleting his conversion, guido
focuses on AI erasure andbiases in artificial
(06:41):
intelligence.
His writings can be found acrossmultiple countries, including
Australia, the Americas andAfrica and, of course, in our
humble little publication, theJewish Independent.
Enjoy this fascinatingconversation about the scary,
dark but ultimately hopefulworld of AI.
(07:03):
Guido, we're going to talkabout a recent article of yours
(07:29):
in the Jewish Independent, whereyou shared with us some really
fascinating research looking atstereotypes and prejudice and
anti-Semitism in AI, a newemerging field.
But before we get there, tellTammy and I and our listeners
about your unconventional pathto life in the Australian Jewish
community.
Speaker 3 (07:48):
I was born in Brazil and I was born in a city called
Salvador.
Not like the country, there'sno L, so it's just Salvador, and
it's in the northeast of Braziland it's the most African city
outside of Africa.
You know most African cityoutside of Africa.
You know like 75% of thepopulation is African
(08:08):
descendants.
And for people who doesn't know, all Brazilians, africans or
black people really came becauseof the transatlantic slavery
trade.
And another thing that peopledon't know is that most of the
humans that came from Africaenslaved to the Americas, came
(08:30):
to Brazil.
The population of the UnitedStates of Africans is about 50
million.
In Brazil it's about 120million.
Speaker 1 (08:38):
Wow, I didn't know that.
Speaker 3 (08:39):
Yeah, so if Brazil was a country in Africa, it
would be the second largest,just behind Nigeria.
Speaker 2 (08:46):
Do you know much about the story of your
ancestors and when and how theyfound their way to Brazil?
Speaker 3 (08:52):
No, absolutely not.
So what happened is there wasan event which is kind of a
history erasure, which issimilar to what I studied in my
research.
History erasure which issimilar to what I studied in my
research.
This Brazilian presidentdecided to eliminate all of the
traces of slavery to show thatit was important for community
(09:15):
to move on.
So he burned all the documents.
Then we can all start like akind of an year zero, like Pol
Pot tried to do in Cambodia.
And he did that.
So it burned out the documents.
On my father's side, which isentirely West Africa, on my
mother's side, funnily enough,my grandfather, who is a white
(09:36):
Portuguese guy, apparently hadblue eyes.
So my mom says, and his name isOliveira, 50% of the Oliveiras
are Jewish converts who, afterthe Inquisition, were forced
into conversion to Catholicism.
So you know, maybe I'll startthere.
Speaker 2 (09:56):
This is going to become particularly relevant
when you tell us about how it isthat you came into the Jewish
community and into Judaism,because you weren't born Jewish,
as we'll soon discuss.
Tell us about your upbringing.
Speaker 3 (10:09):
This is a shame to admit moment.
Recently, the Brazilian filmindustry won an Oscar for a
movie called I Still here, orAinda Estou Aqui in Portuguese,
and the movie depicts adictatorship in Brazil that
began in 1964 and lasted until1985.
(10:29):
So I lived, actually livedthrough a dictatorship, you know
, a military dictatorship, andmy shape to the mid-moment is,
paradoxically, my father, who isa very left-wing Marxist guy
who taught me like when I was 13, he gave me Karl Marx to read
because he's like oh, you needto read the manifesto, the
(10:51):
communist manifesto, and I did.
He was from the Air Force, myfather was from the Air Force,
so the military dictatorshipthat oppressed Brazil and
oppressed people like me wasalso what defended me or
protected me because, as I said,my father was in the Air Force
(11:14):
during the dictatorship of themilitary.
It's important to note for mepersonally and for the record
because this is a record thatmight last forever that my
father often told me how he didnot participate in the
dictatorship and he refused toengage and he was very left-wing
, and I think that's importantfor me, you know, for his record
(11:35):
.
But we were part of thedictatorship, which is so
connected to what Israel istoday.
For me, because things are socomplicated and people want an
easy answer, it's like oh,you're part of the dictatorship,
you're the bad guy, and you'renot part of the dictatorship,
you're the good guy.
And for me, that's why maybefor me it's so easy to
(11:57):
comprehend Even with someonethat didn't grow up with, you
know Israel in the forefrontthat yes, there's things that we
can criticize, but within thatsystem that clog, there's humans
, and all these humans havethese particular stories.
And, yeah, I don't see thingsas bad and as good, and I often
(12:21):
tell my kids there's no good andthere's no bad.
You have to remember that, thatno one is really good.
And I often tell my kidsthere's no good and there's no
bad.
You have to remember that, thatno one is really good, no one
is bad.
People live circumstantiallyand they will do bad things.
Speaker 1 (12:34):
Guido, that's very wise words, but we've skipped
over a little bit.
We still don't know how youcame to Australia and found
yourself in Melbourne's Jewishcommunity.
Speaker 3 (12:47):
Talk us through that so when I lived in Rio, I used
to work in a cyber cafe fixingcomputer printers and plugging
people to the internet, and Imet this young, well, ashkenazi
girl.
At the time we were both veryyoung and, yeah, we got together
(13:11):
.
Speaker 1 (13:12):
So she was in Brazil.
Speaker 3 (13:13):
She was visiting Brazil.
She was, you know, doing awalking holiday from London,
from England.
She came to Rio, met me.
I convinced her to stay.
She lived there for almost ayear, over nine months, and then
we decided to get engaged, youknow, and we chose, you know
like, we could live here, wecould go back to Australia.
(13:35):
She said I have to finish myuniversity.
We came back to Australia andfrom that moment on I was a
member of a Jewish family.
You know, and I don't know,like, maybe a few months later I
was going to a bar mitzvah withone of the cousins and I went
to that.
There's a beautiful synagoguein Melbourne, Turak, I think
(13:56):
it's called the MelbourneSynagogue.
It's like a very big dome.
Speaker 2 (13:59):
This is Turak Shul.
I think it's the second oldestsynagogue in Melbourne.
Speaker 3 (14:04):
yeah, yeah, on Turak Shul, I think it's the second
oldest synagogue in Melbourne.
Yeah, on Turak Road, and I wentto a Bahamian for that and I
was like, oh, I like this and Ireally you know, I can't explain
it.
Well, I think the simplest wayis that my father, he was really
an admirer of Jewish thinking.
We read Kafka, or you know Isaid Marx, and he always talked
(14:27):
about Jewish composers and wewatched many Holocaust films and
he was very interested.
There was very a big connectionbetween.
He could see parallels betweenthe African, black suffering and
the Jewish suffering, so he wasvery interested.
Speaker 2 (14:45):
You're in Melbourne.
This is in the 1990s, I imagine.
Speaker 3 (14:48):
No, no, no, early 2000s, so 2003.
Speaker 2 (14:52):
Okay, the early 2000s , but certainly I don't imagine
there were a lot of newlyarrived Afro-Brazilians in the
Jewish community at that time.
Speaker 3 (15:02):
No, there wasn't and there isn't today.
Speaker 2 (15:04):
Yeah.
Speaker 3 (15:05):
I'm like the single African guy in my congregation,
which is fine.
Speaker 2 (15:11):
Sometimes the community is accused by itself
of being exclusionary, beingdiscriminatory to some members
of the diverse Jewish communitythat we have in Melbourne and in
Sydney.
Was that your experience in theearly 2000s?
Speaker 3 (15:26):
Well, the way I see it is, racism always exists in
all aspects of my life, but Idon't have any case where I was
singled out in my congregationor even every congregation I've
been because I was black.
But I know, you know I can'tcontrol individuals.
You know, like I remember therewas this guy that was looking
(15:47):
at me once at shoe Not the shoeI currently go under, one of the
shoes.
I've been to many and he waslooking at me.
I think I said Shabbat Shalomor whatever, and he's like
didn't reply and I was like,okay, you know, whatever, you
know like, you know like it's,but I it's, it's just one person
.
But yeah, so I I reallyconnected to my wife's ex-wife
(16:08):
she's at next now, but we'resomewhat friendly because we
shared three kids, but yeah, Ilove the Shabbos.
I was like everyone was yellingat each other.
She was yelling at her mom andshe was like, wow, this is, I
wanted that.
I like that.
I tell people I joined Judaismfor the challah and for the
latkes and then I got October7th, you know, like as a side.
Speaker 1 (16:34):
As a side dish.
Speaker 3 (16:35):
And I always felt connected.
I realized later on I had a lotof Jewish friends.
I basically was Jewish already,without you know, having done
the conversion, and I found thisyou know, great rabbi, and
that's sort of that's thejourney I'm living now, but it's
like I really felt that wasalways part of my life.
(16:56):
And my rabbi says you know.
Rabbi Allison says you know,you're just returning.
You're just returning and Ifeel like that you know.
Speaker 1 (17:07):
So we've heard a little bit about your context
and your relationship to Judaismand Jewishness, which is
important for our listeners toknow, especially when it comes
to the work that you're doingnow, and that is in academia.
So what originally led you toacademia?
Speaker 3 (17:31):
Well, I was always a reader.
I loved nonfiction.
I'm really interested in theways the word works and the
system works and power anddynamics and gender and race and
history.
And when I moved to Australia,I only had year 12.
So I completed year 12.
(17:53):
And my ex-wife and I, we openeda business and I started to make
money out of this businessselling women's clothes and
women's jewelry from SouthAmerica, mostly from Brazil, but
also from Argentina, peru,chile and I remember talking to
my father.
He passed away in 2017, butthat's like you know, 2005 or
2006.
(18:13):
And I said to him you know, Imoved to Australia.
Now I could study, but thewhole point of study for me was
to make money.
And I said to him I'm alreadymaking money, but, yeah, so in
2019, as my relationship reallywent south, like I realized we
were divorcing, I thought youknow what I always wanted to do?
(18:37):
This?
I'm not really into fashionanymore.
I want to do academia, but Istill was thinking I should do
one degree, one for me and onefor the business.
So I did a diploma ofprofessional writing and editing
, which I could use forcopywriting for my website at
the time, and then I did thebachelor's of digital media and
(18:57):
writing, and digital media iswhere I started to use, you know
, the tools that I could use formy business and I realized
really digital media was veryEurocentric.
Like I remember, I went to aclass and this woman was
teaching us about film and shesaid the first films were made
(19:19):
da, da, da, da da.
And she was telling somethingabout Europe and she never
mentioned Bollywood.
And it's like really, you gotpaid to teach film and you're
not even going to mentionBollywood, not even going to
mention polio, not mention, noteven sure, not teach, but not
mention.
So this is eurocentric.
And I was typing my tiktok andI saw the, the filter, and the
(19:40):
filter showed me my features butlike it sort of thinned my nose
and made my eyes green and mademy head straight and and I was
like here is the most beautifulversion of you.
And I was like no, that's notbeautiful, you know for me, and
that's what led me to theresearch.
(20:02):
I was like what else AI istelling people?
That's the most beautiful?
It's right or wrong.
Maybe even two years ago Ityped Jewish people.
Two years ago it was likepeople in grey with sad faces.
That was Jewish people.
Speaker 1 (20:19):
So you're saying, if you typed into something like
chat GPT, show me an image of aJewish person, it would feed
back.
Something to you that lookedlike it would feed back
something to you that lookedlike Two years ago.
Speaker 3 (20:33):
It looked like Holocaust figures and sad people
, people wearing grey and skinny, and it was really horrible.
Now, two years later, that wasin the beginning.
Now, two years later, itimproved, but then it became
(20:56):
ultra-stereotype into, like youknow, bearded men.
So basically all Jews areAshkenazis.
Speaker 1 (21:00):
According to AI, according to AI.
Speaker 3 (21:03):
Most Jews are men, like basically they don't have
women.
And it's funny because if youtype family, they would put two
guys.
They don't have women.
And it's funny because if youtype family, they would put two
guys.
They would put a guy andanother guy, and a girl and a
boy or a guy and a guy and a boy.
But they wouldn't put a woman,like they couldn't.
The AI cannot see a Jewishwoman unless if you say Jewish
(21:25):
woman, if you say Jewish family,it thinks it's all men.
And if you put Zionists, youhave this army looking, people
looking very bad at the camera.
Speaker 2 (21:37):
They look aggressive.
We encourage the listener totake a look at some of the
photos that Guido unearthedthrough his research, and we'll
put the link to that in the shownotes.
But, guido, can you just help?
Tammy and I understand themechanism that sits underneath
AI, so if you were to look underthe hood, what exactly is
running it that generates theimages that we see?
(22:01):
I use AI frequently, but I'veactually got really very little
understanding and knowledge ofwhat it draws on.
Speaker 3 (22:10):
Yeah, it's really not transparent.
Exactly what goes inside ofthat jush?
It's basically a big cooking ofinformation that is collected,
scavenged, really, over theinternet, but what we know based
on whistleblower here orcomments there or kind of paper
(22:32):
here and there is that a lot ofthe data that was trained by
ChatGPT or Midjourney, which isthe visual software that I use
in my research, is informationaround the internet.
What happens is, because it'sanglophone, so it uses mostly
English-speaking information andit's mostly focused on Europe
(22:57):
and the North sort of globalNorth, which I don't like that
term, but you know by likepeople that are European and
European descendants.
Most of the information comesfrom those places.
It also uses, like even thingslike Reddit and obscure blogs.
So what happens is you'regetting all these male-centric,
(23:18):
chauvinistic, at timesEuropeanized views of the world.
Again, it's very important forme personally not to say AI is
bad or even that Europeans arebad or whites are bad.
That's not what I'm saying.
I'm saying that the AI toolsthey utilize, they center those
(23:43):
people despite the existence of,you know, despite Europeans and
really white people being aminority in numbers in the
global, even though thesecountries are the leading
countries in the world in manyaspects, in the industrial
aspects, every search orsomewhere saying most of our
(24:15):
data is Eurocentric, it wastrained by males and those
buyers will appear.
If you want to get somethingdifferent, you have to be more
specific.
And just a disclaimer so peopleunderstand.
Speaker 1 (24:27):
In your article you wrote that this idea of a
homogenized Jewish people isdangerous and, in your view,
anti-Semitic.
Can you speak more to that andthe negative repercussions of
presenting Jews in that way?
Speaker 3 (24:47):
Yeah, so the reason this idea is anti-Semitic is
because I can't prove it, butit's very likely that, with the
information we have, that the,the diaspora, are the main
drivers of the narrative aboutJewish people.
Again, there's not a problemwith that per se.
(25:27):
I'm not combating that.
That could be a differentconversation.
I'm more saying because that'sthe information out there,
that's what these tools areutilizing to create the big Jews
, and also because, historically, jewish men have had more
opportunities to have theirvoice heard in many aspects.
(25:48):
And I think the dangers in thatis that it shows us as Jewish
people, as one specific type,with one specific gender, with
one specific idea.
And if there's one thing that Iknow about being connected to
the Jewish community for overtwo decades is that we are so
different, we look so different,you know, not just physically.
(26:09):
You know I met Black Jews, Imet Indian Jews, I met Asian
Jews.
You know I met Asian Jews, Imet disabled Jews, I met queer
Jews and tall short, all sortsof people.
And if that's the image that'sbeing shared to the world,
especially with Gen Zs, ifthere's a study that says 46% of
Gen Zs rely on AI to tell themwhat to do or what to say.
(26:33):
And if that's the image thatthey get, homogenization and
stereotype can what to say?
And if that's the image thatthey get, homogenization and
stereotype can lead to violenceand violence can lead to harm to
us People if they see us asthese people that are Zionists,
men, beaded, whatever it's, justlike it kind of would lose our
(26:56):
own self-identity, our idea ofsecrecy, and it becomes easier
to harm us, to attack us.
Speaker 1 (27:01):
I think and we've seen that happening recently,
when there's been a lot ofdenial about Jewish people being
Indigenous to the land ofIsrael and that's because many
people just see Jews asAshkenazi inverted commas white
and they don't actually see thefull picture.
Speaker 2 (27:25):
Guido, we should mention that the research you've
done has been conducted throughshowing these images to
research participants.
Showing these images toresearch participants.
So tell us about how theparticipants responded to the
(27:46):
images of these stereotypicalAshkenazi bearded Jewish men.
What was the response?
Because this is the reallyintriguing part of your research
.
Speaker 3 (27:56):
What my research found is, as the bias embedded
on AI tools, such as generatedimage systems, increases, so
does the bias on the usersincrease.
And I think this is a bigdanger because if people believe
(28:18):
what they've seen, becausethere's an underperformization
of AI tools, because they speakto me I have an AI, I talk to
Chachi Pichu often, you know,had a proper conversation with
it and unless, if you are reallyswitched on and analytically
(28:38):
thinking critical, you willbelieve it's a human that's
telling you information orthere's an anthropomorphization
of it and people believe inthose informations.
And for me, what my researchshows showing images to people
is that they became more racistand more stereotypical and
accepting.
(28:59):
I don't see even a mitigationof it because the companies have
no intention of retrainingtheir materials and when they
try to do mitigation, they mightremove entire words that need
to exist for historical context.
For example, the New York Times,this African-American artist.
(29:19):
She was doing a work withenslaved ships and she couldn't
Google the word enslaved becausethe word enslaved got disabled.
So if you disable the wordHolocaust, for example, because
you wanted to protect Jews,let's say you want to do, for
example, because you wanted toprotect Jews, let's say you want
to do a good thing, you want toprotect Jews against people
(29:41):
utilizing that, but then if youdisable that word, you erase the
entire history, and that'swhere AI erasure comes from.
The mitigation won't solveanything.
Disclaimer I think ourgovernment should take AI.
As you know, we build roads.
We should be building AIregulated by government and with
(30:06):
support of institutions,universities, like we should
have AI.
That's an Australian AI,because it's like it could have,
for example, people saying AIwould say Australia began in
1788.
And it's like sure, but youshould add European Australia
started in 1788.
Speaker 1 (30:24):
I'm glad you touched on that, because I wanted to
talk about something that youmentioned in another interview
you did on I think it was theABC when you mentioned the
Alicia Keys Superbow Bowlincident.
Speaker 3 (30:36):
Yes.
Speaker 1 (30:37):
So just to enlighten listeners who aren't aware, it
was last year's Super Bowl.
Speaker 3 (30:42):
Last year's Super Bowl 2024.
Speaker 1 (30:45):
2024 Super Bowl.
Alicia Keys was the halftimeperformer.
Those who were listening livewould have heard that there was
this lovely human moment whereone of her notes was slightly
off.
It's fine.
Everyone moved on.
Then, after the event, thehalftime performance was
replayed.
(31:05):
It was perfect.
So whoever reconstructs thathad used AI to fix her off note
and then you spoke about that inthe interview as some kind of
AI erasure.
Speaker 3 (31:21):
Yes.
So AI erasure is a conceptthat's tangential for my
research that's going to come ina paper that I'm writing
currently and I've been writingarticles in different places
which is the idea that, becauseAI is so pervasive, the
individuals, institutions,governments can use AI tools and
(31:42):
generative images generated bywant to erase their
contributions to the RomanEmpire.
Stalin removed adversaries fromphotographs because he wants to
(32:04):
erase those contributions tothe adversaries, so that always
existed.
So AI erasure what happenedwith Elissa Keyes is that she
missed a note and then used AItools to recreate her voice, and
what happens in thesesituations is that the public
can't go to history.
It's a lie, it never happened,and what would that be for the
(32:26):
future?
Speaker 1 (32:27):
But you mentioned that AI should be regulated by
governments.
My Jewish fear is that what ifwe had a government that was
racist against Jews and decidedto erase the Holocaust from
their AI and to have Holocaustdenial or other forms of racism
(32:49):
against Jews in Mizrahi,sephardi community denial?
Could that happen?
Speaker 3 (32:54):
Yes, it could happen and we've run the risk of it,
but right now it's either theAustralian government or a long
mask and it's like, unlessthere's a grassroots movement,
perhaps the worldwide Jewishcommunity could create their own
(33:14):
AI, but then again that couldprobably potentially be accused
for being biased.
Let's say, maybe for our peoplewe would have to.
Really, when I say governments,I shouldn't, I don't have the
intellectual capacity or theskills capacity to advise on the
solution, but what I'm sayingis that someone that's
(33:40):
responsible, that has acollective well-being as humans
maybe the United Nations, unesco, I'm not sure who will create
an AI tool that is collective,that has a contribution from
different diversities, and itwill, you know, and it accepts
(34:01):
you know.
I think the disclaimer issomething that I would like to
see.
Despite our best efforts, it ispossible that some perspectives
were missing, and this isimportant to note when reading
this information on track tptokay, don't talk yourself down.
Speaker 1 (34:20):
I think you do have the capacity to be an advisor on
this committee.
I'm thinking maybe it shouldnot go through the united
nations, because then theremight be a bit of october 7
denial.
Speaker 2 (34:32):
Oh, she threw shade on the united nations guido, I'm
interested in what you make ofhow quickly things are changing
and if we were possibly to fastforward five years, what is a
positive case scenario with AIand its generation of images?
(34:54):
If we were to be optimistic,what could we see, maybe in five
years' time?
Speaker 1 (34:58):
Good question, Dash it is a great question.
Speaker 3 (35:01):
There's this collective in New Zealand who
uses AI to recreate Maorilanguage and find, because
there's words that disappearedand then they find ways, because
they fed the AI with Maorilanguage.
They've recreated, they'verebirthed this language.
So AI is great and I think itwill improve our productivity.
(35:27):
We're still going to have to becreative, you know we still
have to do, but AI can be usefulfor us, but with more
contribution from diverse voices.
Speaker 2 (35:36):
In the worst-case scenario, the stereotypes, the
racist imagery just continues toget worse and worse.
Speaker 1 (35:43):
I don't know if I want to know worst case scenario
.
It's too much.
Speaker 3 (35:47):
No, I think it's just like.
The worst case scenario is whathappened in the beginning of
January this year, 2025.
Like Facebook, filters wentdown, you know, and then now
anyone can say anything.
So the worst case scenario isone we can have sponsorship
research, so people, you chargeon, charge TPT, but people get
(36:12):
sponsored AI answers and youcould drop the filters because
right now there's filtersmitigating racism, mitigating
anti-semitism.
You drop those.
That's the worst case scenario.
I don't want to put fear intopeople.
I think we'll survive all theautocrats out there and we'll
(36:35):
have Shabbat.
We'll eat challah five yearsfrom now.
Ai is just going to be a tool.
What was a tool?
Microsoft is a tool.
Photoshop is a tool.
Remember in the beginning ofPhotoshop?
Oh my God, people are makingwhatever, someone is skinny
whatever.
There was this panic and thingssettled in and people
(36:56):
understand that these tools cometrue, you know, for marketing
purposes.
So I'm not an AI iconoclast.
Speaker 2 (37:05):
No.
Speaker 3 (37:06):
I'm an AI encourager, but what I'm saying is that we
have to have critical thinkingwe have to have mitigation
strategies Bad boundaries.
Speaker 2 (37:16):
I'm envisaging a future, Guido, where I can go to
AI and say create me an imageof an Afro-Brazilian working in
an internet cafe in Coco CabanaBeach and this Afro-Brazilian is
Jewish.
I'd like to see what you comeup with, AI, and I get something
(37:37):
approximating you, Guido.
Ai doesn't kind of go no, can'tcompute, that doesn't make
sense.
A future scenario where you cancreate something that
approximates the diversity, themultiplicity of what it means to
be Jewish yeah.
Speaker 3 (37:53):
And for me, like it's a privilege, you know, to be
Jewish, it's reconnecting wellto my soul, as my rabbis say, or
a connection to who I want tobe and who I am, and it's
beautiful to be diverse.
I have these beautiful kids who, you know, have African
ancestry, brazilian indigenousancestry, portuguese, possibly
(38:15):
Jewish ancestry and born inAustralia, and I think we should
be grateful for it and I hopethat the community is embracing
and you know I feel welcome inthe synagogues I've been in, I
hope I continue to be and I hopethat more people from other
(38:36):
racial groups can be part of theJewish community in.
Speaker 1 (38:37):
Australia.
Guido Mello, thank you so muchfor joining us on A Shame to
Admit.
Speaker 3 (38:43):
Thank you, it was my pleasure.
Speaker 2 (38:45):
That's it for another week.
Speaker 1 (38:47):
You've been listening to.
A Shame to Admit with me TammySussman and executive director
of TJI, dr Darsha Lawrence.
Speaker 2 (38:54):
This episode was mixed and edited by Nick King,
with theme music by DonovanJenks.
Speaker 1 (38:59):
If you like the podcast, leave a positive review
, tell your people or encourageyour third cousin's cousin to
advertise on the show.
Speaker 2 (39:08):
You can tell us what you're ashamed to admit via the
contact form on the JewishIndependent website or by
emailing ashamed atthejewishindependentcomau.
Speaker 1 (39:18):
As always.
Thank you so much for yoursupport and look out for us next
week.
Thank you.
Popular Podcasts
Amy Robach & T.J. Holmes present: Aubrey O’Day, Covering the Diddy Trial
Introducing… Aubrey O’Day Diddy’s former protege, television personality, platinum selling music artist, Danity Kane alum Aubrey O’Day joins veteran journalists Amy Robach and TJ Holmes to provide a unique perspective on the trial that has captivated the attention of the nation. Join them throughout the trial as they discuss, debate, and dissect every detail, every aspect of the proceedings. Aubrey will offer her opinions and expertise, as only she is qualified to do given her first-hand knowledge. From her days on Making the Band, as she emerged as the breakout star, the truth of the situation would be the opposite of the glitz and glamour. Listen throughout every minute of the trial, for this exclusive coverage. Amy Robach and TJ Holmes present Aubrey O’Day, Covering the Diddy Trial, an iHeartRadio podcast.
Betrayal: Season 4
Karoline Borega married a man of honor – a respected Colorado Springs Police officer. She knew there would be sacrifices to accommodate her husband’s career. But she had no idea that he was using his badge to fool everyone. This season, we expose a man who swore two sacred oaths—one to his badge, one to his bride—and broke them both. We follow Karoline as she questions everything she thought she knew about her partner of over 20 years. And make sure to check out Seasons 1-3 of Betrayal, along with Betrayal Weekly Season 1.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.