S2E10: Decoding Twitter: A Look into Machine Learning, Social Media Moderation, and Workplace Dynamics with Melissa

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Nicola (00:00):
We are starting this episode today with a bit of a
call out for our listeners, sowe are pretty much halfway
through season two now.
We finished recording forseason two as well, which is
really exciting.
However, we don't know if weshould do a season three or not,

(00:22):
because it is a lot of fuckingwork and we don't get paid for
it.
I say a lot of work, it's a lotof work Like this, this camera
here just shows up.

Gina (00:35):
Yeah, nikola has done most of the heavy lifting in terms
of editing Instagram, linkedin,all of those things.
I would be absolutely uselessat doing that, but Nikola has
really taken on the nuts andbolts of getting this up and
running.
Every time we talk to someoneor I tell someone about this

(00:59):
podcast, they're like that'ssuch great work.
We only want to help people.
We're not here to make moneyfrom this, but at this point we
both have full-time jobs.
We're both single moms.
We're not sure the amount ofmoney we're putting into the
podcast.
We're not even breaking even.
So we're upside down in termsof whatever we put into it.

(01:22):
That's not even it.
We just spend so much time onthis and we love doing it, but
at some point it's unsustainable.
It's unsustainable.
It would mean so much to us.
If you guys can rate, you don'thave to write the review.
We would love it if you writethe review.
Just hit five stars or fourstars or whatever.

(01:45):
Just rate us.

Nicola (01:49):
Share on LinkedIn.

Gina (01:53):
Repost our Instagram stuff .
We just need more people tolisten, we need more subscribers
.
We need we just need it all.
So it would mean a lot to us ifyou guys especially guys who've
been with us since thebeginning could potentially help
us and share, rate, review Allthat good stuff.

(02:16):
That way, it tells us that ourwork does matter and that it
makes sense for us to keep going.
We are at a little bit of acrossroads because we're not
sure if we should continue onwith season three just because
it's just up in the air, becauseof all the things that we've
just mentioned.

(02:37):
Nicola, do you want to talkabout anything else?

Nicola (02:41):
I think it's so important.
You know, word of mouth is socritical and there's just so
many opportunities or funnythings that we've shared inside
a lot of these podcasts that Ithink people or these episodes
that people don't even realizeare there because it hasn't
really been shared far and wideenough.
We've got a really great basefollowing, which is amazing.

(03:05):
We're so grateful for that andso thankful for that.
We never you know, it's farmore than the three we
originally assumed would bethere, and that's amazing.
And we've got listeners fromall around the world again
amazing.
But we're kind of at theircrossroads now where it's kind
of like, okay, well, do we keepdoing this?

(03:25):
And it's hard because we'vereceived emails and messages and
texts saying, hey, listen toyour podcast was really
impactful I'm in a similarsituation or messages on
LinkedIn even where people havecommented and saying, oh, my God
, I've been in the samesituation.
It's so good to hear thatthere's kind of light at the end
of the tunnel and you know,that kind of key has kept us

(03:49):
going until now.
But now it's kind of like, okay, we need to do a big shift, or
it's just not as sustainable aswe would hope.

Gina (03:59):
I also think we didn't realize just how much work it
was going to be, because when Iapproached Nicola with the idea,
we were like, oh my God, it'sgoing to be so much fun, and it
is fun.
It is fun, but I don't thinkbecause neither one of us had
ever done it before.
We just didn't have theknowledge to be like okay, yeah,
it's fun, but then you have todo all this other stuff, all the

(04:21):
editing, all you know, all ofthese other back end things that
unfortunately fall on Nicola,because I would be useless.

Nicola (04:30):
Absolutely useless.

Gina (04:32):
We'd have literally have one follower and it would be
like my mom, who would neveractually listen to podcasts, but
because my editing would be sobad.
But we digress.
So anything you guys can do tohelp us, I think that I've tried
not to interrupt a lot of ourguests.
Season two that was big on mypriority list.

Nicola (04:54):
We listen to feedback.
We take it on board.

Gina (04:57):
We might give you a little bit of a harsh time, of hard
time about it at first, but it'sour way of showing love.
It's part of my love language.
Verbal abuse is my lovelanguage.
Just kidding, I'm just kidding,although I feel like there
might be a little bit of truthto that with my really close
friends, but um, yeah, soanything you guys can do just to

(05:18):
let us know that you know we'rebeing heard and we should keep
doing this would be so amazing,so so amazing.

Nicola (05:26):
Absolutely.

Gina (05:27):
Before we get into this episode.

Nicola (05:29):
No, I think that's it from me.
I'm so excited about thisepisode though.
Oh my God, it's wild.

Gina (05:37):
Okay, just a little.
Should we give them a littleteaser?
It has everything.
Okay, oh, it's got everything.
It has pass-all takeovers.
It has machine speaking nuance.

Nicola (05:49):
Language learning.

Gina (05:51):
It's just it's got a little bit of everything.
If anyone knows Stefan, it'slike a Stefan episode.

Nicola (05:59):
Yeah, it's pretty cool and we learned so much.
Oh, my God, so much I was likewait, I didn't even like, I
didn't even subscribe for like amaster's class in machine
learning, but we got it and itwas awesome.

Gina (06:14):
Yeah, we got it.
So now I still cannot domachine learning, but we know at
least that exists, because Idon't even think I realized it
exists before that.
So please buckle up, get readyfor this.
There's Elon.

Nicola (06:27):
Musk there's.

Gina (06:28):
Sinks.
There's right, there's Sinksinvolved.
There's contracts, there'shostage some hostage holding,
hostages being held.
So in a very broad sense.
But yes, we're interested inthis episode to see how it goes,
so please buckle up.

Nicola (06:50):
All right.

Gina (06:51):
Let's go, let's do it.

Melissa (06:53):
Then you want to be interviewed, maybe by the person
who's going to be your boss,your boss's boss, maybe your
co-workers and maybe the head ofthe department or the head of
the wherever.
So it's yeah, that's.
It's a fairly common thing.

Nicola (07:07):
So it sounds pretty robust.
Yes, Okay, because again, weknow nothing of data science.
I'm like this is fascinating.

Gina (07:17):
It's like I was like if someone said natural language
learning, I would be like oh, soyou're a speech pathologist,
like, yeah, right, what Like I?
It's so interesting, though,like because you're right, like
so.
So when you're doing thescience behind that, what does
that look like?
Are you like monitoring forslurs or are you writing

(07:38):
algorithm?
Is it like a data algorithm?
Like what?

Nicola (07:42):
is it?

Gina (07:43):
Tell us exactly what it is , because I have a feeling.
A lot of our listeners probablydon't know either Are as stupid
as we are Lulis says we are yes, no, I think it's also very
niche.
It's a very niche area, right?
So it's like we probably takeit for granted.
We just get on these sites andit is what it is, but there are

(08:04):
people behind that, you knowlike.
Alyssa, who are doing it, soexplain.
Maybe it dumb it down for meplease.

Melissa (08:13):
Of course, and it's you're not dumb at all.
This is a really, reallycomplex, as you say, really
niche topic, so no one shouldfeel unintelligent because they
don't understand something.
I spent 10 years learning,right, so, yeah, so at a place
like Twitter, when I was workingthere, there was 30 million

(08:35):
tweets every hour, so 500million tweets a day, every day,
coming at you.
So some of your users may befamiliar with content moderation
.
On smaller, like a Reddit boardor a Discord forum or something
like that, where maybesomething somebody's acting up,
they're harassing somebody, theycan they just moderate, they

(08:55):
just ban that person, right?
So there's this need of humanlevel interaction.
That type of human levelinteraction is just not possible
on a site with as large andactive a user base as Twitter.
So there's this need formachine learning processes, for
data science.
So I write these algorithms thatcontinually search through

(09:19):
Twitter and that's for keywords,like if somebody uses I'm not
going to say them, but ifsomebody uses a public slur that
we all recognize as a slur,that person can be banned
immediately.
But it's more complex than that, because there's all kinds of
extremism, all kinds of speechthat you may not want on your

(09:41):
site, so you can't just bankeywords, you have to understand
context, and this is this fieldof natural language processing
it's teaching a machine tounderstand human speech, which
is really, really fascinating.
So let's say, you feed amachine language processing
algorithm war and peace and yousay what's this book about, and

(10:04):
it can summarize that fairlywell.
The real challenge is when youhave 280 character tweet, a very
, very short tweet.
So what you do is you look, yousee a word that's potentially
troubling and you look outsideof that word, you expand to try
to understand the context aroundthat word and let's say doofus

(10:28):
is a slur word, you know, justsimple word.

Nicola (10:31):
So if I were to tweet at Gina and I'm like you're a
doofus, that would get flagged.
And then Gina tweets I gotcalled a doofus and I'm sad.
Essentially, that shouldn't beflag right.

Melissa (10:47):
You are.
You are right.
You've honed in right there toone of the major problems
Because, let's say, you havethis statement, as I said, you
know, let's.
I'm just going to use a figure.
Hillary Clinton is a doofus.
She drinks blood, she of thechildren child vampire Right,
she's not.

Nicola (11:05):
She's not.
I'm just saying I'm sorry.

Melissa (11:10):
I'm sorry, that's just normal.

Nicola (11:11):
We're normal people.
We know she's not.

Melissa (11:14):
Of course all your listeners are lovely people,
they understand this.
But you could say, somebodycould say that that would be
definitely a red flag.
But a journalist says such andsuch, a politician says, and
then they repeat the same thing.
You don't want to flag that.

Gina (11:33):
Right or so?
That's what I can't believe, soand so says this you don't want
to essentially, you're teachinga computer to understand the
context, to then flag what isinappropriate and let the non
inappropriate like a repetitionof the slur, like the repetition
of doofus.
So that's so.

Nicola (11:55):
I don't bet that is like the math behind that, I feel,
is like my brain doesn't evenmess.

Gina (12:01):
that's the math, don't mess for Nicola.
It's such an interesting like.
It's such an interesting sortof marriage between linguistics
and like computers, like youknow.
So it from your standpoint, islike AI is great, but it's like
you have to teach right thecomputer to understand, like

(12:24):
what we're asking.
So you're probably high indemand now.
Wow, you're awesome.

Melissa (12:30):
To me and feel and there's even if you'll allow me,
there's even another wrinklebecause I worked in, as I said,
political misinformation.
So, despite gathering a lot of,garnering a lot of headlines,
really as a subset of like allspeech, all the tweets on
Twitter, the pretty smallcategory.

(12:50):
So first of all, you have tosay you have to identify the
tweet as political, so youbucket it into this being a
political speech and then youhave to search for it being
misinformation and it was very,very uncommon.
Again, we get big headlines,but it's really not that common,
so you really need it's.

(13:13):
Can I talk about how the wholeprocess worked, or would you
like to get fascinated?

Gina (13:20):
I don't even know this existed.

Melissa (13:22):
Great.

Gina (13:24):
But it does right, because it's like we take for granted
what you're doing, because wejust go on these sites and type
away you know, and we don't evenknow the half of it, so carry
on.

Melissa (13:33):
It's really amazing.
Yeah, so there's really three,as I say, this three legged
stool, this three legged stool.
So, number one, you've got notin any order, they're all
equally important.
Number one you have the datascientists, the machine learning
engineers, the people who arewriting these algorithms,
because, as I said, 500 milliontweets a day, every day, they

(13:56):
never stop.
No human workforce in the worldis large enough to monitor all
those tweets.
That's the first stool does legof the stool.
The second leg are the experts.
We were expected to.
We Twitter operated in the US,in Brazil, in India, australia,

(14:20):
across the world.
These, you know large countries, both large and small, and we
were expected to keep thosesites clean of.
Now, think about slang andpolitical misinformation.
And I'm I follow US politics,but I'm certainly would not
claim to be an expert on USpolitics and I don't really.

(14:41):
I don't really know much ofanything about Indian politics
or Japanese politics, but Japanis our second biggest market
outside of the United States andit's huge, twitter's huge there
, and so you have to haveexperts who understand.
Hey, I'm seeing imagine in 2017.

(15:01):
Hey, I'm seeing this, thispolitical trend called QAnon.
You should look out for this.
They're saying these things sothey can help us to write our
algorithms to better catch allthese emerging things, all these
code words that you can neverknow right, and so you need
that's that site.
Say that safety council, whichyou may or may not have heard of

(15:23):
, but there's about 100 or sovolunteers who were academics,
who were law enforcement people,who were really experts in
misinformation and hate speech,who would advise us somewhat to
look for.
And the third leg of this stoolwas, in fact, the human
moderators, because, as good asI would love to say my algorithm

(15:45):
is, it's going to miss thingsand it's going to miss classify
things.
If you ever tried to interactwith an AI and it gets something
wrong and you try to correct it, you know, let's say, on a
helpline.

Gina (15:58):
It ends up in a loop.

Nicola (16:02):
Trust me, I had a fight I actually swore it chat GPT the
other day where I was just likefuck off, and it was like we
don't use language like that.
I was like, wow, okay, becauseI asked it to do something.
Gpt, don't tell me what to do,don't tell me, I asked it to do
like a simple task.
I was like, can you take thisbank of text and turn it into

(16:23):
145 characters?
And then it gave me 700characters.
I was like, can you count?
And then it gave me 300characters.
I was like that's still not 145.
Oh, yeah, yeah.

Melissa (16:37):
That's brutal so okay.

Gina (16:40):
So that is honestly, Melissa, that's the coolest
thing.
Like, yeah, it's very, very.

Nicola (16:47):
This is like one of the funnest jobs you could have.

Melissa (16:50):
I really enjoy it.

Gina (16:52):
Do you I?

Melissa (16:53):
really love it.

Gina (16:55):
Because whenever you're data science, I'm like ooh, like
I think there were data andthen I think of like data entry
and I'm like mm-hmm, but that'snot at all what it is, so, so
cool, okay.
So, now that we understandbetter what it is, you do, yeah.
So my question is, though howdo you handle those types of

(17:15):
things?
Like and I think that this is aprecursor to what we're going
to end up talking about how doyou handle, like you know, in
our Constitution, in America, inour Constitution, it's freedom
of speech.
So how do you reconcile thatwith what it is that you're
doing?
And you would have to know,probably and is Twitter governed

(17:37):
, governed by the United StatesConstitution?
Like, how does that work?
Because, right, it's in allthese different other countries.
So how does that work?

Melissa (17:47):
In general, these tech companies follow the laws of the
countries in which they areoperating.
So you definitely have to havelawyers involved.
You have to have people whoknow the, the local laws and you
know kind of customs and howthe laws are interpreted.
You can't feed that into analgorithm yet and have it
understand.
So that is definitely true.

(18:09):
But at the same time, ingeneral, we followed, you know,
whatever company in the UnitedStates follows diversity.
You know DEI practices.
You want to make sure yourspeech is not hate speech, it's
not slandering somebody in termsof gender or, you know, sexual

(18:33):
orientation or race or any ofthese protected characteristics.
And this was not, you know, wewere not a, you know, run by the
government, so we were free to,you know, allow any kind of
speech on the site that wewanted.

(18:54):
I understand the argument.
I've had a lot of conservativessince come up to me and say,
hey, I think you should have letthis speech on because of and
they would lay out their caseand I said I can understand why
you're saying that, but Twitterwas not some sort of outlier.
We were just following, if youlooked in, any kind of corporate

(19:16):
culture.
That's.
That's what we were followingBasically the kind of company
best practices.

Gina (19:21):
Okay, so that makes sense.

Nicola (19:24):
So if you were working in political misinformation and
please, you know, you, you, yousay you read a bit of the
American.
I read a little bit of theAmerican news, but I live in New
Zealand, so and we're about togo through vote like a voting
thing now and it's not great.
So I do not claim to be anyexpert in any political.

(19:48):
I don't even know how thepolitical system works in
America.
All I know is what I read inthe news, right?

Melissa (19:53):
So how do?

Nicola (19:54):
you.
You know how did leading up tothe election.
I'm assuming that that made itincredibly difficult to you know
on the three-legged stool.
Now you've got more tweets,you've got more users using it
to kind of spread the stuff.

(20:14):
How do you capture, or how doyou?
How do you get themisinformation before it's
spread?

Melissa (20:23):
Yeah, another.
Yeah, I mean, that's the bigthing, right?
So, definitely, during before amidterm election or before an
election, in general, politicalactivity would tend to spike in
political, hence politicalmisinformation would spike, and
there were I'm trying to thinkof this example, you know.

(20:46):
So, if I can mention somepolitical topics, no judgment on
these topics.
This is just some things thatwe flag.
Yeah, absolutely so, there was.
So there's a conflict in Ukrainegoing on right now, a big and a
big one.
We would see, oh, this such andsuch a bridge was blown up and

(21:08):
you're depriving all thesechildren of their food, and this
would sort of go viral.
And these monsters, they'vedone this.
A little bit of investigationwas saying actually, this image,
this video, was taken in 2018in a different area of the world
.
So you very, very quickly flagsomething like that.

(21:28):
Other times, there wereoutright calls to incitement of
hate or violence.
There was in another countryagain, I don't want to cast
dispersions the mixed faithwedding was proposed of two
prominent people.
There was a lot of hate speecharound that, people who were
sort of inciting or calling toviolence.

(21:48):
Once those go viral, you needto be very, very quick on them
to tamp them down.
So when your algorithm flagsthings, you're just extra on the
ball for these issues.
You're like okay, here'ssomething is so you really
really need to be checking allthese fires at once?
Everybody just works harder.

Nicola (22:09):
Okay now coming back to like the workplace environment
at this point.
So let's say you are in chargeof the political flags.
Are you on call 24 seven Like?
Is your phone like?
Ping, flag alert, flag alert.

Melissa (22:26):
I mean yes and no.
The secret of working in techis that you're basically always
on call.
Hey, I work from home, you cansee, but that also means I work
in front of my computer.
I'm in front of my computer, Ican solve issues all the time.
So, definitely, this is athat's a big issue in tech.

(22:49):
So basically, I could be askedat any time of the day.
But we did have a globalworkforce.
We had people in Japan, in thePhilippines, we had people
across the world, the actualdata scientists working on.
There's only about 30 of us forthe whole company which can.
This is a whole other issue,but there's only 30 of us for

(23:12):
the whole company.
So already kind of a smallworkforce, but we were dispersed
fairly globally in an attemptto combat this issue.
So there's always people active.

Gina (23:22):
Wow, okay, that's interesting.

Nicola (23:24):
Who did you guys use slack?

Melissa (23:27):
You slack a lot.

Gina (23:31):
We've got, we've got a big hate relationship with.

Melissa (23:34):
Yeah, really.

Nicola (23:39):
We came from.
Slack was used, obviously a lot, and there's just so many
conversations that can go overpeople's heads or misconstrued,
or you know, and then there's somany sub channels and it's just
like.

Melissa (23:56):
I know it's.

Nicola (23:58):
It's been a over a year since we were at this place and
I still have struggles on mynotifications.

Gina (24:05):
Okay so can you take us to give us a little idea of what
the company culture was beforeit was purchased by Elon Musk,
like?
Tell us what it was likeworking there for the, I guess,
the majority of your career,while you're at Twitter, right?

Melissa (24:22):
Sure.
So um, so I came on and I wasthis new data scientist.
I've not new to data scientists, but I didn't understand all
these.
You know funky terms, you knowimpressions and you know
articulations and annotationsand all this kind of stuff and
what what they all meant.

(24:43):
And so the people who werethere were very, very generous
in volunteering their time tohelp me understand and to get me
up to speed.
And you know, you've got towrite these complex algorithms
and multiple languages Pythonand SQL and high spark and all
these different languages andyou know, helping me understand
that the code that existed, youcan imagine, a company's been

(25:04):
there for 15 years and some ofthat code has been around a long
time and I'm hard to interpret,and so people really very
generous and helping meunderstand things.
I would say it's fairly afairly generous and open,
collegial environment.
You don't get to work in datascientists and, you know,

(25:25):
certainly at Twitter unless youhave multiple degrees.
So the everybody there is kindof has a certain level of
understanding of the issues thatare going around.
So I found it to be very openand supportive environment.

Nicola (25:41):
Okay.

Gina (25:42):
Did you?

Nicola (25:42):
get?
I'm just curious to knowbecause I'm the nosey neighbor
Were there any cool benefitslike did you get extra leave?

Melissa (25:54):
I worked at Twitter as a contractor, so I technically
worked for a different company.
This is very common in tech andthis is something I found out
also once I left Twitter thatthis maybe works differently in
different fields in a differentcountry.
So in the US, if you work as acontractor in the tech field,
look, I had a Twitter emailaddress.

(26:16):
Let's angle it.
Twittercom had a company slack.
I went to all the meetings.
I just was treated as this kindof disposable workforce.
I didn't get that, I didn't gethealthcare, I didn't get all
these benefits.
So, yeah, that's basically.
So no, I didn't have no, no coolperks for me, but I think you
know, other people had certainthings.

Gina (26:38):
Yeah, was the office like really cushy, though was it nice
.
Like that was very nice, I wentin maybe once a week or so.

Melissa (26:44):
But I mean, I work in.
I live in San Francisco.
You know you could.
It was easy for me to go in,but yeah, yeah, but it's also
easy to stay here.

Gina (26:53):
Yeah with your cute little puppy over there.
Okay, so you were pretty happywith with Twitter the way it
felt meaningful to work in thatenvironment, to work in
political misinformation.

Melissa (27:04):
Feel like you were going to make a difference.

Nicola (27:08):
Yeah, because you know, because I know nothing, were you
a Twitter during that big likeJanuary six stuff.

Gina (27:18):
Oh, the insurgency here.

Nicola (27:21):
Yeah, because I feel like that would have been a
firestorm.

Melissa (27:26):
Yeah, that actually was right before I started, so I
did not deal with.

Gina (27:30):
Oh, you missed it by like.

Melissa (27:31):
I can't comment on that .
I was still feeling the effectsof that.
A lot of policies came intoplace because of that and, of
course, our then President,donald Trump, was banned from
Twitter two days after that.
Whatever, after Jan six, inthat space he tweeted something

(27:53):
like 145 148 separate violationsof our terms of service.
It's not just excitement toviolence, political
misinformation, etc.
Etc.

Nicola (28:02):
But I you know when there was happening and I read
in the news that he'd beenbanned.
In the back of my mind I sawand please correct me if I'm
wrong here, but in my mind All Isaw was, like all these Twitter
people just like running around, like headless to things go oh,
my God, find us a cat.
Delete, delete, delete.

Melissa (28:23):
That you?
That's really true, becauseanother thing about Twitter is
that if an account had over10,000 followers, it was
considered a political events toban it, so my algorithm could
not ban could not action personhas to actually add the account.

(28:45):
It goes up above my head topeople who make a decision, and
so you can see in the Twitterfiles I don't know if any of
your listeners went through andread the Twitter files, but you
can see people like Jack Dorsey,who was then the CEO, and
you'll rock like agonizing overshould we ban Trump?

(29:05):
Should we ban, you know, otherUS politicians like what do we
do here?
So yeah, it was a really bigdecision.

Gina (29:13):
Yeah, I can't even imagine .
And then, okay, so I am stillfuzzy.
When did the whole, when didTwitter even go like be put on
sale or whatever?
Like when did they say, whendid the current owners say we
want to sell it?
Or did Elon go approach Twitter?

Melissa (29:32):
If I can give us sort of abbreviated history of what
happened.

Gina (29:36):
Yeah.

Melissa (29:41):
So this I know some of this stuff because of the what
this recent Wall Street Journalarticle that came out, but I
knew most of it just because ofworking there at the time.
So in January, elon Muskcompleted the sale of a lot of
his Tesla stock and so we hadall this cash and he was looking
to invest it and he startedbuying up Twitter stock.

(30:04):
So in January, he purchasedabout 9% of available Twitter
stock and he was a significantinvestor at that point in
Twitter, which was then a publiccompany.
And so he, being a very, very,very public personality, was
invited.
Like the board said hey, youwant to be on the board, this
would be, you know, great.
And so he was at first on theboard and he had some

(30:31):
personality clashes with aproject or see pro agro wall and
some other people who werethere at the time and so he made
the decision he wanted to buyit.

Gina (30:47):
Nobody was actually selling it.

Melissa (30:50):
Nobody was selling it.
This is going to be a hostileOkay makes the decision he wants
to buy it, although it's yeah,he made a decision, he wants to
buy it.
He offers he has to puttogether all this money.
So put together this $44billion offer, this staggering
amount of money, for that's acompany that, frankly, probably

(31:11):
was not worth that much even atthe time.

Gina (31:14):
Right, it's worth a lot, but 44.
That's a shit ton of money.

Nicola (31:19):
That's like.
That's like a couple of smallcountries.

Melissa (31:23):
Yes, yeah that's the ggb.

Gina (31:26):
It's like a number you say when you're like exaggerating
oh my God, I don't know.
44 billion, right like I.
Just I called you for a billiontimes and you know if it's like
that much money to throw awayon Twitter.

Nicola (31:42):
Can he not just bump me like just just half a mill, like
I'm not even asking for a lot,like half a mill, why?

Gina (31:48):
don't we all just get like half a mill?
It's a hostile takeover, she's.
That's crazy.
I don't think I realized that.
I thought that he just gavethem like an offer they couldn't
refuse with that 44 billion.

Melissa (32:03):
Yeah, so yeah, essentially he, so he offers
them this money.
There's some resistance in theboard, but again it's.
It's something he's able topush through because of who he
is.
Even Elon Musk doesn't have $44billion.
Three, because most of it'stied up in stock and all this
stuff that you know so putstogether this loan involving

(32:26):
lots of different people LarryEllison, saudi investors, people
around the world.
He sort of pools his money.
When you're rich, people giveyou money.
So he sort of pools his moneyand he makes this offer and
almost immediately starts toback out and he doesn't want to

(32:48):
see this offer anymore.
But he's already signed, theink is already dry, is already
signed.
He's already signed thiscontract that he sort of ironed
out this agreement with theboard that he's going to buy it.
Um, they say no, this is thecontract You've made this
contract.
You need to buy it.
And so from about late Januaryall the way through the summer,

(33:14):
he's saying he wants to back outof it.
All of us workers there werelike what the hell is going on
here?

Gina (33:22):
That was my next question how did all of that back and
forth?

Nicola (33:28):
Because I feel like that would have felt so unsettled,
like what a horrible environment.

Melissa (33:35):
Yeah, it was really tough.
It was really tough.
It was really hard to know whatthe future was gonna be.
And, specifically, I worked incontent moderation.
We all felt like we had targetson our back, even before we
purchased it.
He was saying I don't like theway you do content moderation, I
don't like the way you dopolitical misinformation.
We banned one of his favoritesites, this Babylon B.

(33:56):
It's this politicalconservative humor site.
We banned it.
He didn't like that, but theyviolated our terms of service
and so we all felt like, oh,this is really bad, and so
you're right.
It was really unsettled.
Nobody really knew what thefuture was gonna be.
You can imagine how upsettingthat would be to your morale if

(34:17):
you're trying to work in thatenvironment.

Gina (34:20):
That would be horrible.
And everything's up in the air.

Nicola (34:22):
Yeah, I mean, that's like the worst, like you're just
feeling like shit, Like why areyou even bothering moderating?
Like why are you even bothering?
But the moderation when no onegives a shit now?

Gina (34:32):
Yeah Well, that's not true , because it's still Twitter's
terms and conditions.
From the old guard, right,nothing's actually changed hands
, so-.

Melissa (34:41):
Nothing has changed.

Gina (34:42):
Right, you still have to do what you've always done,
knowing that the potential newowner of this company doesn't
like what you're doing.
That's gotta be the worstfucking feeling.

Melissa (34:53):
It was really bad and there was a lot of internal sort
of rearrangement and we wentthrough a major reorg.
At that time, in anticipationof the very scant information we
had on what he wanted and howhe would like, we tried to
consolidate all of our servicesto sort of better please this
person, who may or may not beour eventual CEO.

Gina (35:17):
And that's nuts, because it's like then you're catering
to one person.

Melissa (35:21):
Yeah, exactly.

Gina (35:22):
It was very like tyrannical, almost it's like
okay, so he ends up not backingout of the deal, obviously,
because now we know.

Melissa (35:31):
Yeah.
So in mid to late October, very, very rapidly, everything
switched.
She says no, I'm gonna go in.
His lawyers advised him youcan't get out of this, they're
gonna sue you, you're gonna lose.
There's a $1 billion back outclause that he would have to pay
out of pocket.
So you can't back out of this.
So he goes to find I'll get it.
So he goes through and I thinkthat just typifies his style of

(35:56):
leadership ever since just sortof this impulsive sort of
mercurial style of leadership.
Besides, he's gonna buy itreally quickly.
Panic, panic sets in at thecompany.
People are freaking out.
I've been through takeovers,I've been through mergers, I've
been through companies justdying on the vine, these tech

(36:18):
startups, and I've never seenanything like it where people
are openly disparaging.
I don't trust the new owner, Idon't trust he's going to do
what's right for the company.
I've just never.
I've never seen that before,and people really fearful for
their jobs.

Gina (36:36):
Yeah, I mean also I go ahead.

Nicola (36:38):
Neville, I was gonna say how did you feel Like?
What were you feeling at thetime?
Cause you're a contractor,right, so you did you have some
kind of like okay, this is gonnaget really ugly, but at least
I've got the opportunity toeject eject if needed.

Melissa (36:56):
Yeah, I did sort of start to.
I was terrified.
I mean, I was personallyterrified.
I did sort of start to look foranother job.
But it's hard to job searchwhen you're I keep in mind the.
US Like you're doing, full-timework Coming up, yeah, exactly,

(37:16):
and I just, you know, my thenpartner and I would just sort of
he was just like sort ofcommiserating with me and sort
of like trying to distract mefrom this awfulness that was
happening and yeah, I mean justreally terrible.
And then, you know, Muskpurchases the company on, I

(37:37):
believe, October 28th.
He goes in.
There's this photo of himholding a sink which I guess one
of his poor aides must havescratched up for him, and he
says he tweets a picture out.
I just bought Twitter, let thatsink in, so he had one of his
assistant partners.
That is so corny.
Yeah, it's so corny.

(37:57):
That's so corny.
I can't stick.

Gina (37:59):
Also, we just did an episode with an ex-celebrity
assistant and what you said?
That his poor aide probably hadto get him the sink.

Nicola (38:07):
Oh my God, he just wanted to get a sink, like
immediately, like he justdecides for the moment he needs
to find me a sink and the pooraide is like OK, I'm a flatter,
like I'm fladding with myflatmates, you know what?
I'm just going to pull it outfrom my own home.

Gina (38:24):
And I'll like drag it over , or like it's like in the
middle of the night and like hecalls up his aide and he's like
I need a fucking sink.

Nicola (38:35):
Oh my God this makes me think of.
I don't know if you've seen themovie the consultant.
I mean the TV show theConsultant Melissa.

Melissa (38:42):
Yep, yes, definitely.

Nicola (38:43):
You did.

Melissa (38:45):
Yes.

Gina (38:47):
Is he not Elon Musk?

Melissa (38:49):
Oh yeah, and inside there put that together.

Gina (38:53):
Absolutely.
He is Elon Musk, because he'sjust like oh that's so funny.
He's like I need a pistachiochurro at 2.30 in the morning.

Nicola (39:01):
Oh no, give me my pistachio sink.
It's going to be pistachiocolored.
It's going to come from like anart deco vibe.

Gina (39:10):
Yes, Vintage sink.
Ok, so when that happened, whenyou saw the let that sink, I
truthfully have never seen that,because that's unbelievable,
but it doesn't surprise me.
Famous people are fuckinginsane.
Yeah, definitely Most of them.

Nicola (39:29):
Well, OK, let's dial that back for a second.
I don't feel like he's famous.
Let's just make that clear.
He's just got a fuck ton ofmoney.
That has made him special.

Gina (39:38):
But he's well known.
I would say he's famous.
He's like not famous after he'sa famous business person, like
Lea Aikoko was famous businessperson, like there's tons of
people.

Nicola (39:48):
And I'm just restating that South Africa does not claim
him.
We're not claiming him.
He can be someone else's issue.
He's not us.

Gina (39:58):
I guess he's.
He's ours now.
He's America.

Melissa (40:01):
He's all American now your problem.

Gina (40:04):
Ok, so when you found that out, how?
First of all, how was theinformation being disseminated
to you?
Was it that nobody even reallyknew what was going on, and then
it would come trickling in?
Or how was the chain?
Well, we know that the changemanagement was not existing.
It was not an existing changemanagement.
But when you did get fedinformation, how are you

(40:27):
learning it when the rest of theworld was learning it?
How was that happening?

Melissa (40:31):
There was very little that was coming through to us.
So there's panic, there'sscrambling, there's let's write
up our business processes somaybe we can communicate to this
guy what we're doing, what'shappening.
And so that went on for about aweek.
Where we were, everybody wastrying to get me I was not

(40:52):
important enough to get ameeting with him, believe me,
but there were people who wereahead of my department who were
trying to get meetings with them, just trying to get FaceTime
with them, to explain like, hey,this is what we're doing, this
is why our job is important, andI imagine that was happening
across the company.
Sure, yeah.
So for about a week thathappened, he had one quote

(41:13):
unquote town hall with everybodywhere he just talked about how
hard he was going to haveeverybody work and we asked
about layoffs and he said hefelt the company was over
staffed at that point.
And I just I know there's a lot.

(41:38):
Well, thankfully, you all areprobably not aware of this, but
there's like a lot of argumentson Twitter, outside of Twitter,
about like, hey, he owns thecompany, he can do whatever he
wants, he can make his employeeswork hard.
Yes, he can, but we're alladults, we all I have two
children.
We all have lives.
What about this work-lifebalance?
Hey, you're going to work 60,70 hours a week and you might

(42:02):
get fired.
Just imagine what that would dothe morale of your company.
Not good things.
And we were already highlyfractured from this year almost
year long now tension of sort ofreal he won't he.

Nicola (42:14):
Yes, you know what?
Ok, so that actually really isinteresting for me, right?
Because we talk a lot abouttoxic workplaces and I find that
fascinating that now you've gotnot only a toxic environment
unfolding literally in front ofyour eyes you can see it coming
like a freight train but nowyou've got people online pushing

(42:37):
that toxic workplace narrativeas well, where it's like, oh,
tough luck, just fucking pullyour socks up and work.
Any way, you slackers.

Gina (42:46):
Yeah, that's exactly true and people who idolize Elon Musk
, like, oh, he's going to letTrump back on blah, blah blah,
and just want to say I don'thave any say on any of that.
It's above my pay grade, Idon't know whatever, but the big
thing that we heard about wasthat one of the impotences for

(43:07):
Elon Musk to purchase it was toget Trump back on Twitter and to
really sort of uphold theconstitutional right of free
speech, so that basically thatendangers what you do on so many
levels, like personally.

Melissa (43:26):
Yeah, exactly, it was something that we were all very,
very much worried about,especially after the events of
Jan 6, that he's going to letback on these dangerous
political actors who are goingto continue to undermine the
election integrity and to inciteviolence and to perform this
just really awful amounts ofspeech and hate that they were

(43:50):
spewing onto the site pre beingkicked off.

Nicola (43:54):
Wild, but it just undermines all of the work that
you and the rest of the datapeople have done.
You've got this years you said15 years of algorithm and work
and shit that had blood, sweatand tears, that has gone into
creating something that is safeenough or manageable enough to

(44:18):
make things not bad.
And here you have someonecoming in and going eh, the
wanky name, Meh.

Gina (44:25):
It's like what is I really like this.
We're just going to throw itout.

Nicola (44:30):
I'd rather like it in pistachio.

Gina (44:34):
I'm sure it must be pistachio.
I agree.
Yeah, go ahead, melissa,because that must have made you
feel were you angry?
I would have been fuckingfurious if I were you like in
your position.

Melissa (44:48):
I was terrified.
I was terrified, I wanted tocheck out.
I wanted to curl up on my bed,which I worked right in front of
, and just cry.
But I didn't know how to handlethis stress, and so I am a tech
worker I handled it the veryunhealthy way of just working

(45:08):
harder.
I didn't know what else to do.
And again this is like October28.
Us midterms are November 8.
And I thought, look, the onlything I have controlled over is
trying to keep this platformsafe of bullshit Right Before
this midterm.
It's the only thing I canreally do at this point.

(45:30):
And we just safely shepherdedthrough the Brazilian election.
If your listeners recall, thiswas a really big election.

Nicola (45:38):
Oh, I remember that in the news.
That was weird.

Melissa (45:41):
Yeah, yeah, yeah, and there was like a runoff election
.
It was a huge thing for us todeal with because we're very,
very big in Brazil as well.
So, yeah, wild, ok, put my nosedown and work harder.

Nicola (45:57):
I didn't know what else to do.
Now I mean I can understandthat Now I'm going to take a
slight left turn for a second,ok, and so I have got a couple
of glass door.
What glass door reviews?

Gina (46:14):
Please, I cannot wait, I can't believe you and I didn't
think to do this earlier.

Nicola (46:19):
Oh, I had already thought this through buddy.
Ok, good, let's hear it, I wasright with my glass door reviews
.

Gina (46:24):
Let's hear them please.

Nicola (46:27):
Ok, these are not old reviews.
This is August 23.
August 2023.

Gina (46:36):
Like last week, possibly so 27th of.

Nicola (46:37):
August yeah, 2023.
Worst place on Earth.
Pros Nothing, not a singlething.
Cons Elon Musk is a horrendousexcuse for a person putting
people through hell all the time.
The treatment of employees, thetreatment employees are

(46:58):
subjected to, is equivalent totorture.
Also, a rich moron can have hisfun.
Do your mental health a favor,quit now.
That's amazing.
This is not just like a techperson or a person at the bottom
, you know, down at the frontlines.
No, this is management.

(47:19):
Wow, all right.
Oh, there, we have another onestar.
The title is Eh Social mediaspecialist.
Ok, pros Cool team.
Very fun people.
Cons it's not heading in theright direction.

(47:42):
I'm going to quit.
Oh my God, this is so good.

Gina (47:46):
So what is overall on Glassdoor, what's the rating?
It's not great, it's 3.2.

Nicola (47:53):
I really thought it would be lower.
That's probably because youknow why.

Gina (47:57):
Previously to Elon Musk it was a good.
Previously they're pretty goodright.

Nicola (48:03):
So that balances it out.

Gina (48:04):
We should check back in like a year and see what happens
.

Nicola (48:07):
Yes, if the overall so, with that in mind, so now we're
in the process of transitioninginto this new empire.
Oh my God, We've got so many TVreferences here.
It's like we've got Star Wars,We've got, you know, the
rebellion.
Were there people within theorganization that were like

(48:30):
actually something needs to bedone about this?
Or was there people in theorganization that were like this
is this is not okay, Like whatdid HR do?
Yeah, I think.

Melissa (48:47):
Yeah, I think there definitely were so people at the
top.
So all the entire executivecouncil was fired immediately.
Jack Dorsey had left about Idon't know five months earlier
to be replaced with PragueAgrawal.
Yol Roth was there, I thinkYazmina oh my gosh, yazmina Reza

(49:09):
, I think, was still there.
So I know just from listeningto Yol Roth interviewed I didn't
have contact with these people,but I know from hearing him
interviewed and reading otherpeople they all pushed back
against this, against hisdecisions to do these things,
quite vocally and I know thatpeople on my team one person did

(49:30):
was able to speak with himPersonally.
He'd been there for many, manyyears.
Highly respected data scientist, just a brilliant guy, and he
was fired after offering hissuggestions.
People who spoke out on Slackwere critically.
He had Musk had people gothrough Slack channels to see if

(49:53):
people were talking badly abouthim and he fired engineers and
other people.
I know the engineers who werefired if they spoke poorly of
him.
It's just.
You know, that's just insane todo that.

Gina (50:06):
It's insane.
Because, like any boss, you'rea boss for a reason, right, or
you're a leader for a reason andyou're not gonna be able to
please everyone all of the time.
And he had to have some kind ofmodicum of understanding that
when it's this like social andit's being written about in the
news and you know there's gonnabe people with feelings Like he

(50:28):
just wanted a blind loyalty, itsounds like yeah, that's.

Melissa (50:33):
Exactly right.

Gina (50:34):
Yeah, that's terrible Okay so when the executive team
started firing or getting firedor leaving or whatever.
Obviously, I would imagine themorale of everyone is in the
garbage.
It's in the gutter, right, likenobody wants.
You know you're eitheroverworking because you're
terrified or you're probablylike fuck this place, I can't
wait to quit, like our yeah,dumpster fire and also like our

(50:59):
glass door review.
What did they say it's likeit's akin to torture?

Nicola (51:07):
Yeah, I'll see if I can find some more James.

Gina (51:12):
How did you feel Like?
You're like, oh shit, now allof the people who were in
leadership positions, like superleadership positions,
executives aren't being pickedoff Like.
Did you just know it was like amatter of time before you were?

Melissa (51:26):
Yeah, who's gonna advocate for me?
Who's gonna advocate for mydepartment?
Who's going to say, hey, thisis from 30,000 feet, this is
what all the departments do?
Who's gonna help him understandthe company?
And, as it turned out, therewas nobody there, because the

(51:47):
way the layoffs I'm sorry tojump ahead a little bit the way
the layoffs were conducted isbasically he just took a giant
sword and split and cut thecompany in half.
Because on my team I worked ina subset of a subset of
political misinformation Two ofthe three of us were let go and

(52:08):
everybody I talked to thedepartment just seemed to be cut
in half, irrespective of yourjob, your specialty, what you
were actually doing.
And then he ordered a completereorganization of the company at
that exact same time.
No, oh, I'm gonna walk youthrough.
I'm gonna have a gradual change, nothing like that.

(52:29):
It's just gonna happen in ablink of an eye and you're gonna
get aboard my new regimeimmediately.

Gina (52:35):
And if you don't, I'm gonna find out and slack and
fire you.

Melissa (52:38):
I will fire you.
Yeah, you have to perform highto high standards A few, and I
was like I can get into this,but I was let go after two weeks
after he purchased the company.
But people who stuck around hadto sign a loyalty pledge that
they wouldn't work hard.
Yeah, they had to sign aloyalty pledge.
He promised them that theywould have to work hard and that

(53:00):
they would be highly rewardedif they did so, but they had to
pledge that they would fulfillthese terms.

Nicola (53:07):
And were they unrealistic?
The expectations just shitty.

Melissa (53:13):
Yeah, the expectations, the stated expectations, were
really incredibly onerous.
But I just again, I just wantpeople to what if your boss said
, hey, for the next year we'regonna have to put in extra long
hours you might be workingnights and weekends how do you

(53:34):
feel you probably wouldn't.
That's not a place you'd wannawork at, would you?
I wouldn't.
I work hard, by the way, I workreally hard.
I have to believe I believe it.
I have to work hard throughoutmy life.
But, like Absolutely, Iabsolutely also need some
balance in my life.

Gina (53:54):
And we all do.
Of course, nobody can be thebest version of themselves,
working all the time, you knowwhat.

Nicola (54:00):
Let's use our toxic workplace as an example.
Right when Gina met me, I wasworking like maybe 3 am my time
to 11 pm my time, so there wasvery little time for work-life
balance or sleep.
And when she met me, I was afucking asshole.

Gina (54:20):
This is a different person you see before you, melissa,
than the person I met, and soit's similar to that.
Right, it's like if you'regonna be working nights,
weekends, you're like eating,sleeping at your desk or at work
or whatever.
It's like you're not the bestversion of yourself.
There's no way you can be.

(54:41):
Like, if you have to do it likefor one week because you're
like your deadline is likecoming up and maybe you have
like a bunch of last minutethings to do, fine, I think
anyone can get behind that, butlike literally for a year, no,
and did anyone have anything tosay about like, was there any
offer to and this might not havebeen something that you were

(55:02):
offered because you werecontractively, but yeah but like
any therapy or mental healthoffers for like people, the
people who are staying there togo through the transition was
there anything like that,provided?

Melissa (55:17):
No, there was actually nothing.

Gina (55:20):
There's actually nothing.
Oh my God.

Melissa (55:23):
Yeah, nothing like that at all.
Are these poor people?

Gina (55:26):
Yeah, I just feel bad for everyone.
So, out of the people who werethere when you were working, how
many do you think still exist?
And you're like, we know like amajority of them got fired.
But of those people who didn'tget fired, who were there when
you were there, what percentagedo you think is still there?

Melissa (55:47):
Well, my department of 30 data scientists I specialized
not easy to replace who wereworking in political
misinformation.
Of those 30, there were eightthat were left after the various
firings.

Nicola (56:01):
But then that doesn't that then just cause like a
ripple effect of shit.
Like now you've got lowresourcing, you've got people
that are stressed out.
They're gonna miss thingsbecause they're human.
You're not going to have thebest algorithms.
Spotting bullshit comingthrough Like this is just like a
ripple effect of doom.

Melissa (56:21):
This is what I'm saying because these algorithms we
have this idea that machinelearning you, just it's magical,
it's understanding you, butreally that's not true.
You need to continually beupdating these algorithms Like
teach it, keep teaching, keepteaching your accuracy,
especially as the nature of inmy field I know about it's the

(56:42):
nature of political discoursechanges.
You need to make sure you know.
Let's say, like six years agowhen this pizza gate scandal,
there was a scandal in theUnited States called Pizza Gate,
this conspiracy theory.
If you did not update youralgorithm since Pizza Gate and
you were still looking for PizzaGate today, you're not going to
catch all the new things, allthis new misinformation that's

(57:02):
out now.
So you need people there tocontinually update.
Automation is a nice goal Iactually think it's a really
good goal and we probably wereover sapped.
There's a lot of layoffs thathave gone on in the tech space
since then, but I've been partof change management.
Before you have a goal, youhave a plan, you transition

(57:24):
people, you make sure roles arefilled.
He cut staff in half, ended upeven being more than half, but
who quit and were firedlong-term he cut services.
So all this cafeteria that wasclosed, all these other things
were closed, all these extraswere taken away from people and

(57:47):
you're just.
It's like this austerity plan.
You know like you have to workso hard to succeed at this
company, where you're givennothing.

Gina (57:57):
So I just like I don't understand how he, I also don't
understand how he thinks thiscan now be successful.

Melissa (58:07):
I don't either, and I believe he you know a lot of the
people who've left.
They don't have an option.
They are people who are here onH1B visas or they are people
who he has locked into.
You know, like, who are thepeople he has some amount of
power over, and not everybody,of course.

(58:27):
But there's a photo of theTwitter workers that he posted
about a month after I left.
It was posted at 2 am on aSaturday and he's bragging about
how they're all working so hardand if you look there, all
people who are foreign bornworkers, people who are here on

(58:48):
H1B visas, people who are stuck,who are here because they have
them and they leave, they losetheir visa.
Yeah, they lose their visa, theygotta go home and I know plenty
of people who were cut off likethis.
You got, you know, I think, 60days, 90 days I'm sorry, I'm not
super familiar in the US andyou have to leave and you have
to go home if you can't find ajob in that time.

Gina (59:09):
Right, it's just cruel.
That is cruel and even, like Iknow, when we first briefly
talked, we were talking about,like you know it does.
It trickles down to everyone,even the cafeteria workers, like
okay, so now you don't haveon-site cafeteria, and all of
those people are now without ajob and like what's the purpose

(59:30):
of that, Like don't you wantyour employees to be fed and
healthy?

Melissa (59:35):
Exactly.

Gina (59:36):
Like I'm a mess when I don't eat enough.
Like I am a fucking mess.

Melissa (59:40):
I make terrible decisions.

Nicola (59:42):
Oh my God, you guys can do that.
I get, I get.
I'm trying to still come upwith a word for thirst hangry,
because I get thirst hangry.

Gina (59:50):
Okay.
Thirst hang Thirst hang oh myGod, I get so angry, Like if I'm
thirsty, I start like, like no,for me it's food, unlike if I'm
not being fed, if I'm notfeeding my face in, like you
know, a normal fashion.
It's no good.

Nicola (01:00:08):
Totally unrelated.
But when you do eat, do you do?
When you eat something reallyyummy, do you do the food wiggle
?

Gina (01:00:18):
Not usually, unless, like I'm out at like a really fancy
restaurant, because then I waslike, like, like, like.
But my daughter does the food,the food, toe wiggle when she's
eating something she likes, shelike, wiggles her toes, she's
like, yeah, she's so cute.

Nicola (01:00:35):
Do you do the food wiggle, Melissa?

Melissa (01:00:39):
I do, sometimes Absolutely.
Right sometimes it helps.

Nicola (01:00:43):
It has to be like exceptionally delicious.
Oh no, I could do it for like asandwich.

Melissa (01:00:48):
Oh, I love it.

Nicola (01:00:50):
I love it.
I'm like, ooh, that's so good.

Gina (01:00:56):
Well, you know I get it.
You know food does bring joy.
But like, how fucked up.
Back to Melissa's story.
How fucked up is that?
It's like now you got to go outand like Find a sandwich.

Nicola (01:01:07):
Pay for a sandwich with your location.
Get a sandwich.
Where do you go home?

Gina (01:01:10):
But you're never home because you're fucking working
24 seven, so you're just eatingair and ice and water.
I mean, I don't understand this, how I still, even with the
limited amount of information Ihad because I don't really get
too involved with much in termsof, like, I hear about it, I
don't form an opinion, but I waslike I just thought, like in

(01:01:32):
the back of my mind how does hethink that this is going to be
successful?

Nicola (01:01:35):
Like there's already all what's sustainable.

Gina (01:01:37):
Right, and there's already all this rumbling of exactly
what Melissa is now confirmingand what the glass-drawered
views say that are like it isintolerable to work here.
It is intolerable to be underElon Musk, like how does he
think this is going to get areturn on his investment?

Melissa (01:01:53):
I don't understand, because even he did not even
really reward the people whoended up being loyal to him.
There was a woman named EstherCrawford and I don't mean to put
her on blast.
She was a brilliant, lovelywoman.
I'm not criticizing her.
She very publicly put her outon Twitter that she was bringing
her sleeping bag to work.

(01:02:14):
She was sleeping in the office.
She was so dedicated.
This is the thing she'sadvertising publicly, everybody
saying so.
Some people mocked her at thetime.
Well, three months later, shewas fired by Elon.
He just doesn't reward peoplemaking sacrifices, people trying
to work under his regime.

(01:02:35):
There's nothing that you'regetting.

Gina (01:02:38):
What does he actually want ?

Nicola (01:02:42):
He wants results, he wants results Of people just to
suck his dick every morning Likethis is where we're at.
I just don't understand.

Gina (01:02:49):
Okay, so he comes on, he gets rid of some of whatever
it's.
Now Trump is reinstated Freedomof speech.
It's probably a lot, messier onTwitter, which is also now
called X.

Nicola (01:03:02):
I can't with the X.
Actually, to be honest, theminute I woke up okay, I don't
really use Twitter that much,it's not really my pet views,
but like cool, cool, cool,usually had it in the background
.
The minute I woke up in themorning and my Apple phone had
updated and I had a big black Xon my phone, I was like you're
done, delete, we're done.

(01:03:24):
Thank you very much.

Gina (01:03:27):
And also, like when we first started chatting with you,
melissa, we said, like it'spart of like the lexicon, like
it's like I tweeted something,or did you read this tweet?
What the fuck are you gonna saynow, did you?
I asked something, I mean.

Nicola (01:03:40):
I'm sorry, what are?

Gina (01:03:41):
you saying, like, where's the forethought?
Does he have like people whohandle branding and marketing
helping him, or is he justcoming up with shit and, like
word, vomiting it out?
Cause, that's what it soundslike he is the toxic visionary.

Melissa (01:03:57):
You know, when he originally, you know he had this
company called this paymentscompany that he ended up merging
with PayPal Peter Thiel'scompany, paypal.
So they merged together hewanted to call it X.
This is like in 1999.
So he wanted to call it Xcom.
So he's had this in his brainfor a long time 24 years, 25

(01:04:21):
years.

Nicola (01:04:22):
Laying dormant, waiting for the right time to strike.

Gina (01:04:26):
To exit Wait.
So his kid is also named,something ridiculous.

Nicola (01:04:32):
X, black x x, a lot, a lot it's like with like weird,
like outer space letters thatdon't really fit.
Give me a second.
I won't Google it, Because it'sright.

Melissa (01:04:42):
It's pretty much like Ash 12 or something like that
it's ridiculous.

Gina (01:04:47):
So based on that, he should be.
He should be like stripped ofall of his name giving power.

Melissa (01:04:53):
I think you're right Like.

Gina (01:04:55):
I just like and I keep saying it, but I just can't
understand, from a businessperspective, why he thinks what
he's doing is going to end upbeing successful.

Melissa (01:05:05):
It's really wild.

Nicola (01:05:07):
It's X, a E A 12.
So X S to 12.

Melissa (01:05:16):
Rolls off the tongue.

Nicola (01:05:18):
Just really rolls off the tongue, doesn't it?
Oh, okay, hold on, fast forward.
I went to take you all in yourminds to a moment in time.

Melissa (01:05:29):
We're in.

Nicola (01:05:32):
This child is 25.

Gina (01:05:35):
He's at a pub.

Nicola (01:05:37):
He's, he's picking up some girls or guys.
We're you know, we don't judgewhatever he's picking up.
They get home and he's like,yeah, baby.
And the guy or girl is like,say, my name bitch.
And he's like I don't know how.

Gina (01:05:55):
I've been pronouncing your name, or like what do you think
he does when he goes intoStarbucks?
For me, just give me a sec Ineed the pronunciation the
Google translation on that.

Nicola (01:06:06):
Hold on a second.
I'm sorry, what was it againHold on.
Can we go through that?

Gina (01:06:10):
one one time.
I just I can't, and he likeokay.
So there's just, this is a hotfucking mess, melissa, so tell
us how you got, let go.

Melissa (01:06:19):
Yeah, this is quite a story in and of itself, so I'm
here for it.

Gina (01:06:23):
Let's buckle it.

Melissa (01:06:24):
I mean just, a little bit of time to talk about this.
So, as I said, I was acontractor, so all of the
full-time employees andeverybody in my department
except me and one other personwere full-time, so all the
full-time employees got an emailon Thursday.
I guess this would have beenNovember 2nd or something.

Nicola (01:06:44):
I don't have the exact date, but and now we're also on
stress mode because we're liketwo minutes away from your
elections.

Melissa (01:06:51):
Yes, exactly, exactly so we've got double, double
stress okay.
The email says tomorrow morningthere's going to be a
reorganization.
Tomorrow morning you're goingto get an email in your inbox at
9 am and it's going to haveyour instructions for your
future here at the company.
If you are fired, it'll come toyour personal email because you

(01:07:15):
won't have a company email.
If you still have a job, it'sgoing to come to your company
email and it's going to containyour new, your instructions for
your new, your new role,basically your new department.
This is awful.
It's awful, it's terrible.

Nicola (01:07:29):
What kind of change management is that?
None, the answer is none.
There is no change management.

Gina (01:07:34):
This is I'm picturing this right out of that the
consultant TV show.
Oh my god this is absolutely.
If you're not here in 10minutes, you're done.

Nicola (01:07:44):
No, he gave them an hour .
If you're not here in an hour,you're fired, okay so they get
this email.

Melissa (01:07:53):
And so all of us formed this WhatsApp group outside of
Twitter, so that we could talkin case we were fired.
You know, to commiserate, we'reall.
We all feel awful and peoplehere on visas, you know, people
have kids.
It's just traumatic to do yourjob anyway.
And so we all from thisWhatsApp group we're talking.
We're supposed to get.

(01:08:14):
They were supposed to get firedat 9am.
Now I don't know what's goingto happen to me.
I didn't get this email.
I don't know, no idea.
So I'm there, but I'm there tosupport them into whatever.
And so what actually happened isat 11pm local time, people
started losing access to Slackand email and being cut off,

(01:08:37):
being booted out of thecomputers.
So that'll tell you how hard wewere working.
I'm on this WhatsApp people onthe east coast in New York City
hey, I just got kicked off Slack.
Hey, I just got kicked off.
I just got booted off on mycomputer while I was running
code, I just got kicked off, andyou know, by half the people.

(01:08:57):
And then it goes to Chicago,which is in our central time
zone.
I just got kicked off, and, asit turns to be 11 in the
mountain time zone, I got kickedoff.
So it's this slow motion tidalwave that's coming across the
United States and I'm like, am Igoing to be high enough to not
ground?
Like, am I high enough groundhere to?

Nicola (01:09:17):
escape or not.
The tsunami is coming for you.

Gina (01:09:21):
Yes, so they didn't even honor the email thing, it was
just you lost access.

Melissa (01:09:28):
We got kicked out.

Gina (01:09:30):
Okay, just wanted to make sure, okay.
So then what happens?

Melissa (01:09:33):
My boss was fired and my boss's boss was fired as well
.
Somehow I escaped, I don't knowwhy.
I guess because I was acontractor.
They just kicked out.
My boss texted me, I got kickedout.
I guess I'm fired, and he has atwo-year-old daughter and so of

(01:09:53):
course that's awful, and helived in Seattle and so it's
terrible, and so that's likeFriday.
That's Thursday night, and soFriday is a mess.
Everybody's scrambling around.
The weekend happens On Monday.
I'm still there, but I have noofficial role.

(01:10:15):
I don't know my roles.
I didn't get this email, but Istill have access to all the
systems.
So again, I just nose down.
I'm working hard.
I was like who's my boss?
Who do I report to?
What am I supposed to be doing?
I emailed the HR rep and Istill have this because she

(01:10:38):
filed a ticket in Jira, which islike a ticketing system, and it
has Elon Musk's name on it andit says who does this employee
work for?
Who does this employee reportto?
It has Elon's name on it and Iwas like this is insane.

Gina (01:10:55):
That's so right.
You're employed, but you haveno idea who you're reporting to.

Melissa (01:11:00):
You have no idea who you're reporting to who?

Gina (01:11:01):
does he know who I report to?
Oh my god.
So, nicola, to answer yourquestion, there was no change
management, none whatsoever,zero.
It was fear mongering andsurprise retaliation.
It sounds like okay, so you'rethere, you don't know what the
hell you're doing.
Then what happens?

Melissa (01:11:19):
So because Twitter is this public company, it's in the
spotlight, people start saying,hey, what about the US election
that's coming up?
This is outside pressure,shouldn't we have some people
here?
And he's like fine, nobody elsein political misinformation is
going to be fired for this time.
Oh, great, I get like a week,yeah, okay great.

(01:11:43):
I don't know.
So this week goes by, theelection happens, then we do.
You know, there's always likecollection to mop up and clean
up and all this kind of stuffhappens.
I make it through the weeksomehow.
I go that.
That Saturday I was at the mallwith my daughter, who at that
time was 11, and her friend.
I'm at the mall, I looked at myphone and I get a pop-up that

(01:12:05):
says one or more of your accesstokens has been removed about 7
pm and I open it up and I try toget into Slack on my phone and
I try to get an email on myphone and I couldn't.
And that's how I was fired onmy phone at the mall at 7 pm on
a Saturday.

Gina (01:12:24):
So wait, did you, just so.
You never got an email.
You never did anyone ever reachout to you.
Did the?
I know you were working througha staff.

Melissa (01:12:33):
They did later, but not till after I was already fired.

Gina (01:12:38):
Right, but did anyone say like?
So what did they say to youwhen they reached out to you?

Melissa (01:12:43):
The email said your position at Twitter and it was
just.

Gina (01:12:46):
And you're like so I know I don't have one.
Yeah, like I'm not an idiot,like you could have started with
just you're fired.

Nicola (01:12:53):
Thank you, oh my god.
You know, what would have beenreally fun at least they should
have taken a bit of piss out ofthis is they should have gotten
Trump to just do a you're fired.
You're fired, melissa.
That's so funny.

Gina (01:13:08):
Maybe you want to laugh at that.

Nicola (01:13:10):
I would have had a chuckle at least, yeah.

Melissa (01:13:13):
I wanted to say also is like for that week of of
uncertainty.
As a contractor, I have tosubmit a time card every week
and get it signed.
So I didn't even know who wasgoing to sign my time card,
approve my hours.
I contacted my employmentcompany and I'm like, hey,
what's going on?
Am I going to get paid for thisweek?

(01:13:33):
And they said well, I don'tknow if you get paid.
If we get paid, you'll get paid.
And I was like, then you'dbetter make sure that you get
paid.

Gina (01:13:44):
Yeah, did you get paid I?

Melissa (01:13:47):
did end up getting paid .
I will say Okay good, but Ididn't.
I didn't know, I didn't knowwhat was happening.

Gina (01:13:54):
Yeah, and like most Americans, regardless of their
salary are still living,especially like an hour
generation.
Are still living paycheck topaycheck.
Absolutely Like in San Franciscois more expensive than
Manhattan to live, and I mean,I'm sure you're making a good
amount of money, but when you'reliving in like the most
expensive place it doesn'tmatter.
I mean, I remember right beforeI moved to Florida, when I was

(01:14:16):
in New York, still in ManhattanI was probably making close to
200,000 and I was still prettypoor, like I was still like
living paycheck to paycheck, youknow, yes, so it's difficult.
And do you know if the fulltime employees were given any
kind of severance, or were theyjust fired and that's it?

Melissa (01:14:36):
So they were given three months of severance and
although there were some issueswith some people not receiving
it on time, to my knowledge mostof them have received them,
have received that three monthsseverance by now, although some
haven't.
But I, as a contractor, I gotnothing Right.

(01:14:58):
I got no severance.
My company sent me a box ofGeardelli chocolate, which
Geardelli's nice, but it's notthree months of severance.

Gina (01:15:07):
You're in San Francisco.
I mean Geardelli's everywherethey should have imported
something different, right, likethat you wouldn't have access
to.

Nicola (01:15:14):
To kids from New.
Zealand.

Gina (01:15:17):
Or something like even I don't know something else I feel
like yeah, or like even I'mtrying to think is there like a
New York chocolate place thateveryone loves?
I can't think of one, but um,yeah, that's kind of tacky, I
mean.

Nicola (01:15:31):
I'm sure you ate it.
It's not tacky.
Okay, so we've gone throughthis whole shitty change
management process.
It sucks.
Do you still have friends there?
Do you still have people youchat to?

Melissa (01:15:44):
Yeah, so we still have this WhatsApp group.
There's still people who arethere.
We don't chat as much, but youknow we still do chat.
People post updates every oncein a while but yeah, there's
still people there, they're.
I mean again, of the eightpeople who are left, six of them
are here on H1D visas.

(01:16:05):
So they they don't really havea choice of like, hey, maybe I
should look for another job.
It's extra difficult if you'rehere on an immigrant visa.

Nicola (01:16:13):
Yeah and what about you?
Are you doing okay now?

Melissa (01:16:19):
Yeah, so you know, like you said, it's an just
incredibly expensive place tolive.
Most of my adult life I'veworked two jobs.
I teach part-time on the side.
So, um, yeah, I mean, for youknow, this is like November 12th
that I was let go and I'm likethe holidays are coming, nobody
hired the end of year anyway.
I'm like panicking, I'm tryingto take.

(01:16:42):
I took my kids to see somemovie, a movie and I just
remember having a panic attackin the movie theater, like I
can't, I'm not gonna be able toafford to continue to pay rent
for this place, let alone toprovide my children a nice
Christmas.
Sure that's terrifying, I'm justlike I'm like panicking, and
then like you don't want to showyour children.

(01:17:02):
Of course you know you want tobe that good, strong parent for
them and and it was a brutaltime to just to be frank, but
like, fortunately I did get ajob.
My skills are in demand, so Iwas able to get a job.
I'm employed now.
I'm doing great at my new job.

Nicola (01:17:19):
Great.
Are there, are, there, are.
Is there any chance of thembeing taken over by Elon?
Because I'm like now it's likea like a toxic, like bacteria.
It's just coming, it's justspreading.

Melissa (01:17:31):
Just spreading everywhere?
Well, I hope not.

Gina (01:17:36):
We work with medical universities.

Melissa (01:17:37):
So I don't think he's interested in that.

Gina (01:17:39):
He's probably not, and I kind of can't wait for him to
lose 44 billion dollars when heruns Twitter or X, I should say,
into the ground, because that'sthe only way I can see this
ending.

Melissa (01:17:50):
He just announced that his ad revenue was down 60
percent.

Nicola (01:17:55):
He gave away blue ticks for like two, two dollars.

Melissa (01:17:59):
Yeah, yeah, yeah, yeah, it's not surprising.

Gina (01:18:03):
And is he unilaterally making decisions for X slash
Twitter?

Melissa (01:18:08):
Yeah, people who talk about.
Of course there's maybe not ofcourse, but there's a new CEO,
linda Yoccarino.
She comes from NBC, she has adexperience.
He brought her on because she'sall these relationships with
advertisers and he wanted tobolster up somebody, but he very
much remains in charge.
It's very obvious.
He's the public base of thecompany.

(01:18:29):
He's tweeting out directivesand people who again, people who
do work there talk about.
You know, he's just thisincredibly mercurial director
and owner and he can beincredibly charming if he wants
to be, but he can all.
He turns on a dime and he'she's dark Elon and that's the
worst kind of leader to have.

Gina (01:18:47):
Because you're not, you never feel safe in that
environment because you're like,oh, he likes me today, and then
you're like I wonder if he'sgonna like me tomorrow or you
know, on a bigger scale ofcourse, but you know it's.
I grew up with someone who wasvery much like that, like would
switch on a dime, and it's so.
It's like it's traumatic.
It's traumatic to be aroundsomeone who you have to kind of

(01:19:10):
try to guess their mood.
Guess what you can say, what,guess what you can't.
Guess what you do, what youdon't do.
It is, I feel, bad for all thepeople who stay there.

Melissa (01:19:20):
It is.
I really do yeah, so do youthink?
He's obviously a brilliant guy,but just no.

Gina (01:19:25):
No, I know, but maybe he's just not a good leader, right?

Nicola (01:19:28):
I don't think he's a good leader at all.
I think he has been pended tobecause he's got a whole fuck
ton of money to.
Oh my god, he is literally theguy from the consultant.

Gina (01:19:39):
No, he is he is.

Nicola (01:19:41):
They are one and the same and he's not a good leader.
That's not gonna you're notgonna get.
You're not gonna get the bestout of people.
You're not gonna get the bestout of anything.
And now you're in a situ whereyou know you're just gonna see a
decline in in morale.
You're gonna see a decline.
You see more staff quitting.

Gina (01:19:59):
But that's why I'm like how does he think this is going
to be successful?
Like it's just like it's likethe company's like spiraling the
dream hole, you know, it's likeI don't know.
The sink, the sink, let thatsink in.

Melissa (01:20:14):
The sink.
That's so funny.

Gina (01:20:19):
So now that you're on the other side of it and you went
through that painful time butyou're now gainfully employed,
do you think you, knowing whatyou know now, looking back,
would you have done anythingdifferently?
Would you have quit of your ownaccord or like what do you?
Is there anything you wouldhave done differently?

Melissa (01:20:39):
You know, I'm not really close to my coworkers
there and I think again, I stillbelieve in what we're doing.
I still believe in keeping itsafe from political
misinformation.
I think that's a good thingthat we were doing.
So it's hard to say that Iwould have given that away, but
I think from my own sanity andmy own safety, I would have just

(01:21:01):
hey, this person doesn't careabout me.
He clearly doesn't care aboutmy job security.
I'm gonna devote myself fulltime to looking for work.
I would have just tried harderto get another job.

Gina (01:21:11):
Yeah, yeah.

Melissa (01:21:13):
Instead of doubling down and working harder.

Gina (01:21:15):
Yeah, because you don't.
But in that period you're sostressed You're like I don't
know what's happening.
I'm being told new things everyday, like nobody knows who's
coming and going.
Who knows how you would handleit, you know?
Like unless you're in it, butyeah.
But looking back I could seelike, yeah, you probably should
have devoted your entire time tofinding a new job.

(01:21:37):
Yeah, but are you happy thatyou had the experience of
working at Twitter?

Melissa (01:21:44):
I really love working.
Having worked at Twitter, I'veused the platform since like
2009,.
A long time user, I can hardlystand to be on there anymore.
I used to use it a lot.
It's painful to be on now, butI'm glad have you moved over to
Threads.
I've been on Threads, I've beenon Blue Sky Post Tribal, all

(01:22:07):
the ones that are out, yeah,they're just not.
It's not.
None of them is quite there yet.

Gina (01:22:12):
Is that the same yeah?

Melissa (01:22:13):
No, we'll see.

Gina (01:22:14):
That's because there's not data scientists like yourself.

Melissa (01:22:18):
There's not data scientists.
They're not you?

Gina (01:22:19):
Yeah, that's true, they're not you, okay, well, I feel
like we just I feel like we hada school.

Nicola (01:22:26):
We've had another school lesson today where we're like
we came into this not knowingwhat data science was.
I feel like we've lived stillnot knowing, but having some
understanding.

Gina (01:22:37):
Well.

Nicola (01:22:38):
I mean.

Gina (01:22:38):
I knew what data science was.
I did like I actually recentlydid a course on it, just so I
wasn't antiquated with certainthings, but I would not have
known that super niche that youare in, Melissa.
It's so interesting and it'slike, like.
I said previously, like we go onthese sites and you know things
are being cleaned up and wedon't even know that that's

(01:22:59):
happening behind the scenes, youknow, and it's like there's a
whole team of people who aremaking sure that it's safe for
you to post, you know, yourthoughts on Taylor Swift or
whatever.
And then she's a doofus, youknow, because doofus is now our
flag, our slurrward yeah, but Imean, it's true, it's like I
wouldn't have even thought of,like oh, this is just safe, I

(01:23:21):
can say what I want to say aboutwhoever you know, or what I did
today or whatever, and youdon't think about what really
goes behind it.
So it's been very insightfulfor me and I thank you for
explaining it so likebeautifully, because I totally
got what you were doing once youyeah, once you got there.

Melissa (01:23:39):
Great.
Well, yeah, I mean, I love totalk about data science and I'll
get enough opportunities too,so thank you for being great
listeners.

Gina (01:23:48):
You're welcome.

Nicola (01:23:49):
Well, we really yeah, go ahead.
Nicholas.
I was gonna say we reallyappreciate your time today
because it's just, I feel likethis has been a great chat.

Gina (01:23:57):
It has.
I feel like it was fun.
Yeah, it was fun.
Is there anything you want tosay to anyone or how can people
find you if they want to hireyou or have you come?

Melissa (01:24:11):
in and consult and not be a terrible consultant.
Yeah, what.
Yeah, I'm still occasionally onTwitter at mingle Melissa Engel
mingle74.
Oh, my God Talbot yes, okay,funny, you can find me there.
I'm go find me at LinkedIn.

(01:24:32):
I'm happy to take on consultingprojects and yeah, I'm just
sort of generally available tooffer my expertise as needed.
Yeah.

Gina (01:24:44):
Okay, great.
Well, we will put all of thatinformation in this show notes
so people who might hear yourlovely tweet, tweet, tweet.
No, your voice will can reachout.
Yeah, so this has beenwonderful.

Melissa (01:25:01):
Thank you so much.

Gina (01:25:03):
Thank you, I'm not going to ever use Twitter again.
I used it for like months.
Yeah, we probably should.
No, not we probably shouldn't,it'll be interesting.
Maybe in like a year from now,if we're still doing the podcast
and Twitter acts whatever is ina different like, well, maybe
we have to invite you back onand see what happened.

Melissa (01:25:24):
I think it can still be saved.
I really do, yeah, yeah.

Gina (01:25:28):
We'll see.

Nicola (01:25:28):
We'll see It'll be saved if you know my skits on a on
one of his rocker If he gets outof his own way it'll probably
be saved, yeah, I think, if heextracts himself and let's and
hires appropriate people.

Gina (01:25:41):
it's like exactly what we had in our toxic workplace If
the owner got out of her own way, the business would be totally
successful.
Yeah, yeah Just as long aswe're able to do that, just as
long as you're hiring peoplewith the correct experience and
understanding of how to get yourbusiness to where you think it
should be, and if you don't dothat.
You're dead in the water.

Melissa (01:26:01):
Right Actual people not like friends and family.

Gina (01:26:04):
But yeah, if you're not doing that, you're dead in the
water.
So let's see what happens.

Nicola (01:26:08):
What's the?

Gina (01:26:09):
space.

Nicola (01:26:10):
Yeah.

Gina (01:26:10):
Yeah, all right.
Well, thank you so much,melissa it was such a joy and so
insightful, and we will chatwith you later.

All Episodes

Episode Transcript

Popular Podcasts

Amy Robach & T.J. Holmes present: Aubrey O’Day, Covering the Diddy Trial

Dateline NBC

The Breakfast Club

.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;}S2E10: Decoding Twitter: A Look into Machine Learning, Social Media Moderation, and Workplace Dynamics with Melissa

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;}Amy Robach & T.J. Holmes present: Aubrey O’Day, Covering the Diddy Trial

Dateline NBC

The Breakfast Club

All Episodes

S2E10: Decoding Twitter: A Look into Machine Learning, Social Media Moderation, and Workplace Dynamics with Melissa

Amy Robach & T.J. Holmes present: Aubrey O’Day, Covering the Diddy Trial