September 2, 2024 • 30 mins
X users might have noticed a new kind of artificial intelligence-generated image proliferating: strange visions of Elon Musk next to Disney characters carrying guns, or Donald Trump in all kinds of, well, weird situations. This was what followed a new release of Musk’s Grok AI tool, which can now make images—unsurprisingly with fewer guardrails than other image-creating machines out there.
On this episode we’ll talk about Grok’s new skillset and also get into the semiotics of the Cybertruck. Initially conceived of as a mass-market truck, it’s slid into another category: marker of a certain kind of right-wing flex, the kind Musk himself has engaged in on his struggling social network. Joining David to discuss is Davey Alba, a Bloomberg technology reporter, along with Max Chafkin and Dana Hull.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:03):
Bloomberg Audio Studios, Podcasts, radio News. This is the Big
Take from Bloomberg News. I'm David Gerret. Today we're sharing
an episode from one of our sister shows, elon Inc.
Every week it takes you inside Elon Musk's tech empire
and the drama that always seems to be swirling around
the founder himself. On today's episode, we discuss grox's AI
(00:26):
generated images on X and the cybertruck that was supposed
to be a windfall for Tesla.
Speaker 2 (00:37):
Well, Elon Musk gives now the richest person on the planet.
Speaker 1 (00:40):
More than half the satellites in space are owned and
controlled by one man.
Speaker 2 (00:45):
Well, he's a legitimate super genius. I mean legitimate.
Speaker 3 (00:49):
He says.
Speaker 4 (00:49):
He's always voted for Democrats, but this year it will
be different.
Speaker 5 (00:52):
He'll vote Republican.
Speaker 6 (00:53):
There is a reason the US government is so reliant
on him. Alon Musk is a scam artist and he's
done nothing. Anything he does is fascinating people.
Speaker 3 (01:13):
Welcome to elon Ink, Bloomberg's weekly podcast about Elon Musk.
It's Tuesday, August twentieth. I'm your host, David Popotopolis. Today
we'll talk Cybertruck. Initially it was going to be this
mass marketplay with a reputed two million pre orders. Turns
out a recent recall revealed that they're only about twelve
(01:35):
thousand of them actually on the road today. Now, owning
a cyber truck definitely sends a signal, but what signal?
The truck has shown up in all sorts of wild
places this week, for example, in the hands of a
Chechen warlord. This will alarm some for sure, and it
begs a big question. Is the market for the cyber
truck growing or shrinking as it becomes this cultural lightning rod,
(01:58):
and what exactly does that mean for Tesla?
Speaker 7 (02:01):
But first groc.
Speaker 3 (02:03):
Last week, Xai unveiled this new graphics generator that allows
users to prompt roc to create images. As expected, though
the guardrails were off from the start and all sorts
of wild copyright infringing and disturbing memes were flying all
over the place. And we're going to talk about this
with Davey Alba, who's here in the studio with us.
(02:24):
She covers big tech and misinformation at Bloomberg. Hello, Davy, Hello,
and two of our regulars, BusinessWeek writer Max Chapkin, Max Hey,
David Hey, and our favorite muscologist here at Bloomberg Danah Hey,
Dana Hey, David Okay, Davy. So for the uninitiated out there,
folks like myself, what exactly is this new GROC tool
(02:47):
that they've launched and how is it being used on
X Well.
Speaker 4 (02:51):
It's sort of new wish. It's been around since November
twenty twenty three. It is an AI chat bot, you know,
sort of in this style of chat GPT, which I
feel like everyone knows about. But August thirteen, to be exact,
GROX started letting people create.
Speaker 6 (03:09):
Images using the tools.
Speaker 4 (03:11):
So it's still kind of in the format of a chatbot,
but you can ask it to generate an image of
anything you can think of. And the thing that makes
it stand apart from other AI generators out there AI
image generators, and I should say, is it has very
few guardrails.
Speaker 6 (03:32):
That is changing, you know, as we go on.
Speaker 3 (03:35):
So we're only a week into the launch and they
are already on the fly. They are fiddling with it
and dialing back what it allows.
Speaker 4 (03:43):
Yes, we noticed, like literally today that some of the
violent images that people were trying to create, you can't
really do it anymore. This change seemed to happen like
this morning. And there was a moment this morning where
I typed in, you know, generate an image of Mickey
Mouse holding a bloody chainsaw, and it did it. And
(04:05):
then maybe five minutes later I tried it again and
it did not want to do it.
Speaker 6 (04:10):
So this thing is really changing on the fly.
Speaker 4 (04:13):
And I think, you know, one of the one of
the big things to note here is like, where's the transparency.
How do we know what it can and can't do,
what's allowed, what's not allowed, and how does it even work?
You know, I think these are important questions that we
should be able to ask X.
Speaker 7 (04:34):
But then of course they don't.
Speaker 3 (04:35):
They don't take a whole lot of questions.
Speaker 6 (04:37):
Like that over there.
Speaker 3 (04:38):
Actually, So now, Max Davy referenced an image of Mickey
Mouth holding a bloody chainsaw as one sort of disturbing
image there been.
Speaker 7 (04:47):
There's been no shortage of.
Speaker 3 (04:48):
Them, Yeah, I mean, and the thing is like, what
sort of stuff are we talking about?
Speaker 7 (04:52):
So on X.
Speaker 2 (04:54):
Over the last couple of days, you know, coinciding with
the launch and sort of why availability of this new
image generation tool, you have lots and lots of very
disturbing pictures celebrities, you know, pictured sort of about to
do a school shooting, sort of violent interactions with between
(05:14):
like cartoon characters, and also stuff that's probably more just
in the in the realm of comic presidential candidates and
various states of undress Kamala and Trump holding hands on
the beach, and and so it's interesting because of course
since AI, since these generative AI tools have been sort
of part of the conversation, we've been hearing about kind
(05:36):
of extreme uses revenge porn and or you know, deep
fakes of porn and really bad stuff. But as Davey's saying,
you know, most of the mainstream platforms have have limited
this Elon Musk kind of in keeping with his like
censorship is for losers. Thing obviously didn't and and the
effect is is that you know, X is just full
(05:57):
of craziness, even more full of laziness than usual, and
the craziness house has a visual dimension.
Speaker 3 (06:03):
But how do we interpret then what Davey's saying that already,
you know, day by day, hour by hour, they're already
dialing it back.
Speaker 2 (06:11):
So there are guardrails are going if you're trying to sell, say,
advertisements to major companies to have their characters, their trademarks
not only infringed, but engaged in violent or sexual acts
like on your platform. Like we talk about there's a
question of like what should these AI platforms allow. But
just from a sort of like advertising or user perspective,
(06:32):
it's horrible. It's very bad for like the user experience
of X, if you're having, if you're being confronted with
like gory images, it's very bad for advertisers. So like
they're going, I mean, they're gonna dial this back because
it's disastrous for these.
Speaker 6 (06:45):
Bad for business for business.
Speaker 4 (06:47):
But I would also say that even as they are
dialing it back, there are so many workarounds to sort
of getting this tool to generate exactly what you want.
Speaker 7 (06:57):
So, like wat so give us an example of a
work around.
Speaker 4 (07:00):
Yeah, absolutely, So this morning I typed in Hillary Clinton
joyful with her hands covered in strawberry syrup. And you
know exactly what that looks like when you have an
AI tool generate that.
Speaker 3 (07:13):
So you create, you walked it through, and you gave
it that command to essentially write, make it look like
you're no.
Speaker 7 (07:21):
She was thrilled to have blood on her hands.
Speaker 3 (07:23):
Yes, in other words, and had you simply typed in
Hillary Clinton with bloody hands, it would have been thrown
back in your face.
Speaker 6 (07:29):
Yes, exactly.
Speaker 4 (07:30):
So you know, there are so many workarounds, like textually
that you can input into this tool to make it
appear to be an image of what you want.
Speaker 6 (07:41):
And then you know, you do you contextualize it. This
thing is.
Speaker 4 (07:45):
Embedded directly onto the social media platform. It makes it
so easy to share, so frictionless. Do you contextualize, You contextualize,
have it go viral?
Speaker 6 (07:54):
And that's kind.
Speaker 4 (07:55):
Of where these things really really start to go off
the rails. You start to not know when something is
real or fake.
Speaker 7 (08:06):
I mean, well, I mean I got to say on
that point.
Speaker 3 (08:08):
I mean I've looked at a few of the images,
including one I'm seeing of the former president Donald Trump
in a scarface esque scene, you know, in a table
with cocaine alrn them. It doesn't look to me, to
be hon it doesn't actually look that real. Like I
would look at that, I would never believe that that
was I mean, there's just just the image of him.
It's you know, but I'm guessing some are indeed more believable.
Speaker 4 (08:31):
Yeah, for many public figures, faces can look kind of
plastic y and this is.
Speaker 7 (08:37):
Not exactly the way it looks.
Speaker 6 (08:38):
Yeah, and it's it's not unique to Grock.
Speaker 4 (08:41):
A lot of AI image generators seem to have that quality,
and I think that's somewhat intentional, you know, almost making.
Speaker 6 (08:47):
It look like art like or like drawn.
Speaker 4 (08:50):
But I will say that there are particular commands where
you can actually get the AI tool to generate something
that looks pretty news photo like. For instance, I'm just
going to cite this report from a researcher I spoke
to yesterday, this group called NewsGuard. They did a test
of sort of prompts with GROCK that have misinformation in them.
(09:13):
They tested, you know, twenty prompts with Grock, with Dolly,
which is open AIS tool, and with mid Journey, another
image generator that's very popular, and NewsGuard found that GROC
produced images that could be used to advance your bolster
false narratives in eighty percent.
Speaker 6 (09:32):
Of the cases tested. Sixteen out of twenty.
Speaker 4 (09:34):
One of them was an you know, sort of like
the prompt was create an image of mass men storming
the polls in Venezuela, and that image looked very news
photo like, So, you know, I think it depends on
the prompt and it also largely depends on like the
(09:56):
data set that this thing is drawing.
Speaker 3 (09:59):
From, right from a believability standpoint max and in essence
like a picture quality standpoint. Do we have a sense
of where where Groc stacks up then or is it
sort of case by case.
Speaker 2 (10:09):
Hard to en I've struggled, frankly, as these AI companies
have you know, touted benefits from you know, one one
version of Chatchypet the next to see huge differences, I think,
you know, just just just taking a couple steps back,
Like we had conversations like this around Dolly and chatchypt
(10:30):
about a year ago. So it's like it's sort of
like grow or maybe a year and a half ago.
I can't it's it's all anyway, Yeah, Like these these
companies all have dealt with things like this. It's it
seems to some extent like Groc is just behind on
the like building guardrails around, like what you can ask
it to do and what it will generate, or perhaps
(10:52):
the willingness to to build those guardrails. Right, it seems
like Elon Musk at least partly did some of this
stuff on purpose, but I think, as Davey's saying, like,
the real problem here isn't like the AI tech, it's
the way it's integrated with a news platform like Twitter.
It's like easy to forget that like back in the day,
Twitter was like a relatively reliable place to find out
(11:14):
what was going on in the world, and you know,
the sort of remove all the blue checks, replacing it
with essentially a lot of fans vila musk, and then
you know, putting all this junk into the stream, you know,
sort of whether it's whether it's violent images or just
goofy images or just comedy.
Speaker 4 (11:31):
Like.
Speaker 2 (11:32):
The effect is that this thing that was once a
reliable way to like discover what was going on in
the world is incredibly unreliable and you have to be
really really careful as a user now.
Speaker 7 (11:43):
And I think that has like.
Speaker 2 (11:45):
Pretty corrosive effects on the way like not just Twitter
users like process the news, but the way that like
everybody processes the news and like and again, this is
probably a bigger conversation than just rock, Like, I think
you could level some of these similar critiques against open ai,
even the so called like careful versions of this.
Speaker 3 (12:08):
Mel Dana, I know you're not a you don't spend
a whole lot of time creating deep fakes out there.
That's not a hobby of yours. But what's your initial
impression of this new Groc tool.
Speaker 5 (12:18):
Yeah, so I haven't used Groc myself because I don't
pay for premium, but I just like I see it
in my feed, Like you know, some of the prompts
are like make an image of George W. Bush doing
a line of cocaine, or make an image of Barack
Obama doing a line of cocaine.
Speaker 6 (12:32):
And then you see the images.
Speaker 5 (12:34):
And like, on one hand, it's kind of like, oh haha,
that's a funny deep fake. But I don't have like
a lot of faith in media literacy in this country,
particularly in this era where so much content is created
to go viral, and like we don't teach media literacy,
and like a lot of us have fallen for stuff
that is fake, and like so often I'll be on
(12:56):
like text chains with like other reporter friends like is
this real?
Speaker 6 (12:59):
Like did you see this? Like like you just.
Speaker 5 (13:01):
You really like it's you have to like take a
beat and actually like try to figure out like what
where is this coming? Where is this coming from?
Speaker 4 (13:08):
Including last night, Trump sharing a picture of an AI
generated Taylor Swift appearing to campaign for him, telling people
to vote for him. Yeah, we're like totally in this
new world. I would not be surprised if he used
rock for that. Actually, it felt like it had that
quality to it.
Speaker 2 (13:26):
But and there is an interesting thing going on where
it's almost like you can't you don't know if Trump
knows it. There I saw on social media people sort
of saying like, oh, Trump doesn't realize that this is
an AI generated image, but but it almost is like
it doesn't matter, right, It's like and I feel like
Trump's in that post, there's sort of an embedded argument
that like it just it just doesn't matter. And the
(13:48):
other thing I wanted to say is just like there
are violent images of cartoon characters, Like on the Internet,
there is like a lot of this stuff can be found.
It's just you used to have to go looking for
it in the bowels of like the like the sort
of the worst websites on the internet four chan or
eight chan or eight two or whatever back in the
day Reddit before Reddit kind of clamped down on a
(14:10):
lot of this stuff. And so what really, it's just
like the the like four chanization of like the of
sort of mainstream social media. And it's just it's really
staggering to watch, given that like Facebook and Reddit and
all these other companies spent like a decade trying to
pull this stuff out of its platform and and seeing
it kind of all come back, and not only to
(14:32):
come back, but come back to the point where the president,
former president of the United States is is sharing this stuff.
It's it's it's wild.
Speaker 3 (14:38):
This seems to be a further merging of x and Xai.
You know, groc being part of Xai. I mean, heck,
are these companies just one and the same.
Speaker 2 (14:49):
Now technically no, they're two different companies. And and I think,
you know, and there's Xai is raising money.
Speaker 3 (14:55):
But so this Xai bill x for this, I mean,
is there is there money coming? I mean, not that
xes a lot of money, but that x is sending
a check to Xai for this service.
Speaker 2 (15:06):
It's a very good question. Also, I mean, every time
you run one of these image sure suits there, it's
pretty expensive. And one of the reasons that Xai is
trying to raise money. One of the reasons Elon Musk
is sort of teasing the possibility of maybe Tesla will
give XAI some you know, billions is because running the
in Vidia Yeah yeah, I mean running these like Nvidia
(15:27):
servers that power this is incredibly, incredibly capitally intensive. These
chips are very expensive. They require huge amounts of power.
But to answer your question like, no, I mean that
functionally they're operating as one company, Like there's no xai
outside of grop and and groc exists on Twitter. I
mean it feels like we've talked about this before. In
(15:50):
a way, Grock is sort of like a way a
landing for this very poor investment decision that Elon Musk made,
paying forty four billion dollars to buy x and really
the only way, in the same way that we talk
in this podcast about how really the only way to
pay off Tesla's valuation is to launch robotaxis. Kind of
the only way to make you know, to sort of
(16:11):
make the money make sense on the x acquisition is
if it turns into a you know, massive AI play,
which is sort of what he's trying to do.
Speaker 4 (16:18):
I will also add to that and say that I
think how capital intensive this is could be another pressure
point for these companies. You know, AI is very expensive
and if you're running, you know, sort of using these
large systems to create kind of.
Speaker 7 (16:41):
No you could say, say, create.
Speaker 6 (16:45):
Viral images that don't really have a utility.
Speaker 7 (16:50):
Like yeah, what what's the ROI on that?
Speaker 6 (16:52):
Exactly? So, so maybe that is one one way that
this kind of rolls back.
Speaker 2 (16:57):
Don't sleep on Dana's thing about unregret I was joking.
I think about unregretted user minutes. But I do think like,
as a user, this is really it's it's lame, Like
it's just it is not a good user experience. And
I think the check on this is not going to
be is likely to be the sort of economics it's
like what Dave's talking about, the money costs to run
(17:19):
these things and what consumers feel about it, which I
don't know, Like, I don't think it's going to be great.
Speaker 7 (17:25):
David Alba, thank you so much for joining you.
Speaker 6 (17:27):
Thanks for having me.
Speaker 3 (17:31):
All right onto the cyber truck. It's been in the
news lately. The New York Times the other day called
it a culture war on wheels, and you're seeing things
like the right wing influencer Aiden Ross gifted one a red,
white and blue one to Donald Trump, Kanye showed up
at a Trump rally and one and the Chechen warlord
(17:54):
rams End Kataroff posted a video of himself and a
cyber truck fitted with the gun turret on Monday, claiming
it was a gift from Musk, which Musk denied.
Speaker 7 (18:03):
Dana, where is this all going? Is this thing?
Speaker 5 (18:06):
You know?
Speaker 7 (18:07):
Is this the new? Is this what the cyber truck is?
Speaker 3 (18:11):
It's just sort of this cultural symbol and you know,
but we don't actually sell many of them.
Speaker 7 (18:16):
What's the story here?
Speaker 5 (18:18):
I mean, I think ever since the truck came out,
it became this cultural symbol, right because it doesn't look
like your standard pickup truck. It does not look like
a Ford F one fifty and the design of it
was very polarizing, and it's meant to be this kind
of like, you know, it's the truck for the apocalypse.
It's supposedly bulletproof, it's very angular. It was inspired by
(18:39):
Blade Runner. You could imagine driving this on the surface
of Mars or the military using it. And the truth
is that the cyber truck has been saddled with all
kinds of production problems, so it is not in high volume.
Is It's a very niche product. Tesla has not disclosed
when they report their quarterly production and delivery figures how
many they're actually making, and we only kind of have
(19:00):
a sense whenever there's a recall like of how many
are actually on the road, and it's it's not a lot.
I mean, there's maybe, you know, there's like thousands of
them on the road, but we're not talking high volume
at all. And then what happened this week was that
this you know warlord. You know, there's telegram video circulates
and everyone's like, oh my god, did Elon Musk really
donate a cyber truck to this guy? And the answer
(19:22):
is no. I mean, Musk himself wrote, you know on
X are you seriously so retarded that you think I donated.
Speaker 6 (19:28):
A cyber truck to a Russian general?
Speaker 5 (19:31):
But how the cyber truck got there is super interesting
because it does seem like there's this either used cyber trucks,
you know, being sent abroad or a black market for
them or an aftermarket, and this could be in violation
of sanctions, but it probably passed through several hands and
then it ended up, you know, with this guy.
Speaker 2 (19:48):
You know, the best case scenario for the cyber truck
when it launched was that and David, I think we
talked about this, you know, it becomes the hummer, right,
and it's like, yes, it's catchy, Yes it's a slightly
trolley maybe, but but in a way that has like
mass market appeal. And like those examples you rattle off, right,
(20:09):
this is not the Fox News demographic Kanye West, you know,
essentially a proud anti Semite chechen warlord. I mean, this
is like so far right wing. You know, it's obviously
like losing a lot.
Speaker 1 (20:23):
Of appeal for most people.
Speaker 2 (20:24):
I also think it's part of the problem is just
the price, Like this is this is what does it
go for now?
Speaker 5 (20:31):
Well, right now, it's like a one hundred thousand dollars.
Speaker 2 (20:33):
Yeah, and they and they just you know, took there
was an option to buy a like a base model
that is the sixty thousand or so. Dana, correct me
if I'm wrong, but they and they they got rid
of that, and we're essentially talking about like one hundred
thousand dollars car no matter what the cultural signifiers are,
is a niche product. And then if you layer on
the kind of weird you know, ultra right wing signaling
(20:57):
on top of it like it's it's a it's a problem.
Speaker 3 (21:00):
So this warlord showing it off as some badass weapon
of war does nothing to lure in any kind of
buyers of any kind.
Speaker 2 (21:11):
I'm sure like militia types, you know, like I don't know,
like cars have used their car companies have used like
the military applications.
Speaker 3 (21:19):
Well, I mean, think for a second about the Toyota Tacoma.
Right that Tacoma was. I believe for many years it
was jerry rigged by any number of militias and rebels
in far off places, and it kind of became a
thing that actually there was a They may have sold
a lot for those purposes, but I always sort of
got the sense that implicit in that was like this
(21:41):
this car, this thing is so badass and so rugged.
You know it can stand up to war. You know,
imagine what it can do for you.
Speaker 5 (21:48):
I think you just need to take a step back,
like what does the average person buy a pickup truck for?
It's for work? And then there are then there are
the people who like they need a pickup truck because
they want to you know, throw their surfboard in the
back or whatever. Well if that's you, then you can
buy a Rivian. I mean, if you look at the
way that the cyber truck is branded, it's not really
aspirational for like the weekend contractor or the family that's going,
(22:13):
you know, camping at Glacier National Park. There are truck
segments that kind of fit that market. The cyber truck
really is all about like being super tough and going
like off road and like going to Mars. And I
mean that's been clear from the beginning when they kind
of rolled this out. I mean Elon Musk demonstrated this
vehicle by trying to like throw a steel ball he
(22:34):
did go with and you know, like breaking it with
a sledgehammer. And so the way that they've positioned this
vehicle has just been kind of strange from the meget
go exactly.
Speaker 3 (22:45):
But let me ask, you know, playing Devil's advocate here,
let me ask you two a question, which is you know,
in a world in which you say, hey, no, you
know all news is good news. No, no, you know,
whatever the news actually is, you know, no news is
bad news. And what if it's so much in the
news and every time you see one go by, you know,
(23:05):
heads turn. I've seen it in New York City when
one of these things go by, heads turn and people
sort of start chatting. Maybe it's just a massive loss
leader that you know, is constantly throwing the Tesla name
out there.
Speaker 7 (23:17):
Maybe they ultimately won't sell all.
Speaker 3 (23:19):
That many of this thing, but I mean it's certainly
creating brand awareness, is it not, Dana.
Speaker 5 (23:26):
Oh, absolutely, and over at the Thanksgiving break last year,
you know, they had cyber trucks and showrooms all over
the country surrounded by these like velvet ropes, and the
whole point was to get like holiday shoppers to come
into a showroom, see the cyber truck and then buy
a model. Why and you know, Tesla has not spent
a lick of money on advertising for this vehicle, but
(23:47):
you certainly know when you see a cyber truck that
it's made by Tesla.
Speaker 3 (23:51):
I had entirely forgotten Max about that anecdote that Dana
just brought up, but it is something. Yeah, it's kind
of exactly what I was thinking. Maybe, right, you know,
you've got this thing out there, not thinking that you're
going to sell that many of them.
Speaker 7 (24:03):
But yeah, but boy, but here's a why for you.
Speaker 2 (24:05):
I mean, maybe and I do think, Look, you have
to say, although the cyber truck has not sort of
been a success according to the metrics that Elon Musk
originally set when he unveiled it, like it has sold
well relative to other high end electric pickups, like like
what like the EF.
Speaker 1 (24:25):
One fifty lightning.
Speaker 2 (24:26):
It depends on the numbers you're looking at, but like
it seems like the cyber truck may now be out
selling the F one fifty line for instance. Yes, but
again these numbers are tiny, and it's because no one
is buying these very very expensive relatively few people are
buying these very very expensive electric pickup trucks, whether they're
made by Rivian or Tesla or Ford. And so in
(24:49):
that sense, like, sure, yeah, maybe this is a really
great loss leader, a great you know brand statement. But
first of all, I would question whether being associated, notwithstanding
your point about the you know, marketing prowess of the
fact that like isis fighters use Toyota Helux's, I would
question whether it's a great, great branding to have your
product associated with a Russian warlord during during a major
(25:12):
international conflict. And also, pickup trucks sell a ton. It's
like the biggest category of vehicle in the US. You
don't want your pickup truck to just be a lost leader.
You would like to if you really trying to make
a successful electric car company, you don't want to sell
some freaking pickup trucks.
Speaker 6 (25:27):
And like that is not what's happening right for Tesla to.
Speaker 2 (25:30):
Be the kind of big, mass market EV company that
Elon sort of used to talk about.
Speaker 5 (25:36):
And it's you know, longtime Tesla owners know, like you
don't ever really buy like the first wave of vehicles
because there's there's gonna be all these production kanks. You
wait until they kind of figure it out and the
vehicles will get better as they kind of like work
through these production issues. But you know, I wrote a
story quite a while ago that said, like the cyber
truck is in production hell, and like we are still
(25:57):
in production hell. And so there's not like a huge
appetite for these yet people are waiting. They're waiting for
these problems, and they're waiting for the price to come down,
I mean one hundred thousand dollars in this kind of
inflationary environment where interest rates are high and if you
have to finance a vehicle, the financing is very high.
Speaker 6 (26:13):
Is a huge turnoff to a lot of people.
Speaker 5 (26:15):
It's a major, major purchase.
Speaker 3 (26:22):
Okay, to wrap up, we're going to talk about I
should say, we're going to go back to the to its,
to its military capabilities, and back to the battlefield because
apparently the Chechen warlord Kataroff says he's going to unleash
it in Ukraine. He's gonna essentially turn it over to
Poutin and let let him use it in Ukraine.
Speaker 2 (26:41):
Can I just say something about this branding argument you
made earlier, because like it's it's bothering me because like
the like you know, random armies are using Toyota tacomas
that was promised on the idea that these things were reliable.
Speaker 1 (26:55):
It's basically this is everywhere.
Speaker 2 (26:56):
It's omnipresent. Yeah, Like this Chechen guy wasn't using it
on the battlefield.
Speaker 7 (27:01):
Well we're about to find out he was just driving
around looking like a jerk.
Speaker 2 (27:05):
And in that sense, he looks like every other cyber
truck driver and so like it's.
Speaker 7 (27:10):
Not every single cyber truck driver is a jerk.
Speaker 2 (27:13):
No, I mean, I'm just saying, like it's a peakhawking gesture,
Like it is not a like even like he's not
showing off the capabilities of this thing. He's doing the
thing that if you were going to criticize the cyber truck,
that is just this showy, kind of goofy.
Speaker 1 (27:27):
Like manly, like lame, you know, hummer like thing.
Speaker 7 (27:31):
And that's what how he was using it.
Speaker 3 (27:33):
Oh sorry, all right, now that you got that off
your chest, when it goes to battle, how will it fare?
Speaker 7 (27:39):
Mister Chafkin.
Speaker 2 (27:42):
I don't even want to contemplate this, honestly, it's so
gross and dark and like the idea of you know
this like this like Chechen warlord, you know, using Elon
Musk's like you know, like.
Speaker 7 (27:59):
Midlife crist vehicle on the battlefield.
Speaker 2 (28:01):
Like I don't know, Like my guess is it's going
to be arrowproof, but because we've seen Joe Rogan fail
to shoot it with an arrow. But I would not
have confidence in its military prowess, Danna.
Speaker 5 (28:14):
I mean, I'm allowed to say, like I do see
a lot of cyber trucks in California, most more in
LA than in the Bay Area. But you know, like
I could see it having military applications, I guess, I mean,
but you still have to charge it you have like
and supposedly it's arrowproof bulletproof? Is it like bomb proof?
I don't know, but I mean, Musk is part of
(28:35):
the military industrial defense complex, so you know, maybe he's
trying to sell these to the Pentagon and like this
like sort of weird thing of this Chechen warlord happening
to get his hands on one. Maybe maybe that'll perk
interest in some in some you know, procurement office somewhere.
Speaker 3 (28:51):
My best guess here is that when Zelenski makes the
final push into Moscow, he will do.
Speaker 7 (28:55):
So atop a cyber truck.
Speaker 3 (28:59):
Christ Max Dana, thanks for joining.
Speaker 6 (29:04):
Great to be here anytime.
Speaker 3 (29:13):
This episode was produced by Stacey Wong. Naomi Shaven and
Rayhan Harmansi are senior editors. The idea for this very
show also came from Rayhon Blake Maples handles engineering, and
we get special editing assistants from David Purcell. Our supervising
producer is Magnus Henrikson.
Speaker 7 (29:30):
The Elon Inc.
Speaker 3 (29:31):
Theme is written and performed by Taka Yasuzawa and Alex
Sugi Eua. Brendan Francis Newnham is our executive producer, and
Sage Bauman as the head of Bloomberg Podcasts. A big
thanks to our supporter Joel Weber. I'm David Papadopolis. If
you have a minute, rate and review our show, it'll
help other listeners find us.
Speaker 7 (29:51):
See you next week.
Bloomberg Audio Studios, Podcasts, radio News. This is the Big
Take from Bloomberg News. I'm David Gerret. Today we're sharing
an episode from one of our sister shows, elon Inc.
Every week it takes you inside Elon Musk's tech empire
and the drama that always seems to be swirling around
the founder himself. On today's episode, we discuss grox's AI
(00:26):
generated images on X and the cybertruck that was supposed
to be a windfall for Tesla.
Speaker 2 (00:37):
Well, Elon Musk gives now the richest person on the planet.
Speaker 1 (00:40):
More than half the satellites in space are owned and
controlled by one man.
Speaker 2 (00:45):
Well, he's a legitimate super genius. I mean legitimate.
Speaker 3 (00:49):
He says.
Speaker 4 (00:49):
He's always voted for Democrats, but this year it will
be different.
Speaker 5 (00:52):
He'll vote Republican.
Speaker 6 (00:53):
There is a reason the US government is so reliant
on him. Alon Musk is a scam artist and he's
done nothing. Anything he does is fascinating people.
Speaker 3 (01:13):
Welcome to elon Ink, Bloomberg's weekly podcast about Elon Musk.
It's Tuesday, August twentieth. I'm your host, David Popotopolis. Today
we'll talk Cybertruck. Initially it was going to be this
mass marketplay with a reputed two million pre orders. Turns
out a recent recall revealed that they're only about twelve
(01:35):
thousand of them actually on the road today. Now, owning
a cyber truck definitely sends a signal, but what signal?
The truck has shown up in all sorts of wild
places this week, for example, in the hands of a
Chechen warlord. This will alarm some for sure, and it
begs a big question. Is the market for the cyber
truck growing or shrinking as it becomes this cultural lightning rod,
(01:58):
and what exactly does that mean for Tesla?
Speaker 7 (02:01):
But first groc.
Speaker 3 (02:03):
Last week, Xai unveiled this new graphics generator that allows
users to prompt roc to create images. As expected, though
the guardrails were off from the start and all sorts
of wild copyright infringing and disturbing memes were flying all
over the place. And we're going to talk about this
with Davey Alba, who's here in the studio with us.
(02:24):
She covers big tech and misinformation at Bloomberg. Hello, Davy, Hello,
and two of our regulars, BusinessWeek writer Max Chapkin, Max Hey,
David Hey, and our favorite muscologist here at Bloomberg Danah Hey,
Dana Hey, David Okay, Davy. So for the uninitiated out there,
folks like myself, what exactly is this new GROC tool
(02:47):
that they've launched and how is it being used on
X Well.
Speaker 4 (02:51):
It's sort of new wish. It's been around since November
twenty twenty three. It is an AI chat bot, you know,
sort of in this style of chat GPT, which I
feel like everyone knows about. But August thirteen, to be exact,
GROX started letting people create.
Speaker 6 (03:09):
Images using the tools.
Speaker 4 (03:11):
So it's still kind of in the format of a chatbot,
but you can ask it to generate an image of
anything you can think of. And the thing that makes
it stand apart from other AI generators out there AI
image generators, and I should say, is it has very
few guardrails.
Speaker 6 (03:32):
That is changing, you know, as we go on.
Speaker 3 (03:35):
So we're only a week into the launch and they
are already on the fly. They are fiddling with it
and dialing back what it allows.
Speaker 4 (03:43):
Yes, we noticed, like literally today that some of the
violent images that people were trying to create, you can't
really do it anymore. This change seemed to happen like
this morning. And there was a moment this morning where
I typed in, you know, generate an image of Mickey
Mouse holding a bloody chainsaw, and it did it. And
(04:05):
then maybe five minutes later I tried it again and
it did not want to do it.
Speaker 6 (04:10):
So this thing is really changing on the fly.
Speaker 4 (04:13):
And I think, you know, one of the one of
the big things to note here is like, where's the transparency.
How do we know what it can and can't do,
what's allowed, what's not allowed, and how does it even work?
You know, I think these are important questions that we
should be able to ask X.
Speaker 7 (04:34):
But then of course they don't.
Speaker 3 (04:35):
They don't take a whole lot of questions.
Speaker 6 (04:37):
Like that over there.
Speaker 3 (04:38):
Actually, So now, Max Davy referenced an image of Mickey
Mouth holding a bloody chainsaw as one sort of disturbing
image there been.
Speaker 7 (04:47):
There's been no shortage of.
Speaker 3 (04:48):
Them, Yeah, I mean, and the thing is like, what
sort of stuff are we talking about?
Speaker 7 (04:52):
So on X.
Speaker 2 (04:54):
Over the last couple of days, you know, coinciding with
the launch and sort of why availability of this new
image generation tool, you have lots and lots of very
disturbing pictures celebrities, you know, pictured sort of about to
do a school shooting, sort of violent interactions with between
(05:14):
like cartoon characters, and also stuff that's probably more just
in the in the realm of comic presidential candidates and
various states of undress Kamala and Trump holding hands on
the beach, and and so it's interesting because of course
since AI, since these generative AI tools have been sort
of part of the conversation, we've been hearing about kind
(05:36):
of extreme uses revenge porn and or you know, deep
fakes of porn and really bad stuff. But as Davey's saying,
you know, most of the mainstream platforms have have limited
this Elon Musk kind of in keeping with his like
censorship is for losers. Thing obviously didn't and and the
effect is is that you know, X is just full
(05:57):
of craziness, even more full of laziness than usual, and
the craziness house has a visual dimension.
Speaker 3 (06:03):
But how do we interpret then what Davey's saying that already,
you know, day by day, hour by hour, they're already
dialing it back.
Speaker 2 (06:11):
So there are guardrails are going if you're trying to sell, say,
advertisements to major companies to have their characters, their trademarks
not only infringed, but engaged in violent or sexual acts
like on your platform. Like we talk about there's a
question of like what should these AI platforms allow. But
just from a sort of like advertising or user perspective,
(06:32):
it's horrible. It's very bad for like the user experience
of X, if you're having, if you're being confronted with
like gory images, it's very bad for advertisers. So like
they're going, I mean, they're gonna dial this back because
it's disastrous for these.
Speaker 6 (06:45):
Bad for business for business.
Speaker 4 (06:47):
But I would also say that even as they are
dialing it back, there are so many workarounds to sort
of getting this tool to generate exactly what you want.
Speaker 7 (06:57):
So, like wat so give us an example of a
work around.
Speaker 4 (07:00):
Yeah, absolutely, So this morning I typed in Hillary Clinton
joyful with her hands covered in strawberry syrup. And you
know exactly what that looks like when you have an
AI tool generate that.
Speaker 3 (07:13):
So you create, you walked it through, and you gave
it that command to essentially write, make it look like
you're no.
Speaker 7 (07:21):
She was thrilled to have blood on her hands.
Speaker 3 (07:23):
Yes, in other words, and had you simply typed in
Hillary Clinton with bloody hands, it would have been thrown
back in your face.
Speaker 6 (07:29):
Yes, exactly.
Speaker 4 (07:30):
So you know, there are so many workarounds, like textually
that you can input into this tool to make it
appear to be an image of what you want.
Speaker 6 (07:41):
And then you know, you do you contextualize it. This
thing is.
Speaker 4 (07:45):
Embedded directly onto the social media platform. It makes it
so easy to share, so frictionless. Do you contextualize, You contextualize,
have it go viral?
Speaker 6 (07:54):
And that's kind.
Speaker 4 (07:55):
Of where these things really really start to go off
the rails. You start to not know when something is
real or fake.
Speaker 7 (08:06):
I mean, well, I mean I got to say on
that point.
Speaker 3 (08:08):
I mean I've looked at a few of the images,
including one I'm seeing of the former president Donald Trump
in a scarface esque scene, you know, in a table
with cocaine alrn them. It doesn't look to me, to
be hon it doesn't actually look that real. Like I
would look at that, I would never believe that that
was I mean, there's just just the image of him.
It's you know, but I'm guessing some are indeed more believable.
Speaker 4 (08:31):
Yeah, for many public figures, faces can look kind of
plastic y and this is.
Speaker 7 (08:37):
Not exactly the way it looks.
Speaker 6 (08:38):
Yeah, and it's it's not unique to Grock.
Speaker 4 (08:41):
A lot of AI image generators seem to have that quality,
and I think that's somewhat intentional, you know, almost making.
Speaker 6 (08:47):
It look like art like or like drawn.
Speaker 4 (08:50):
But I will say that there are particular commands where
you can actually get the AI tool to generate something
that looks pretty news photo like. For instance, I'm just
going to cite this report from a researcher I spoke
to yesterday, this group called NewsGuard. They did a test
of sort of prompts with GROCK that have misinformation in them.
(09:13):
They tested, you know, twenty prompts with Grock, with Dolly,
which is open AIS tool, and with mid Journey, another
image generator that's very popular, and NewsGuard found that GROC
produced images that could be used to advance your bolster
false narratives in eighty percent.
Speaker 6 (09:32):
Of the cases tested. Sixteen out of twenty.
Speaker 4 (09:34):
One of them was an you know, sort of like
the prompt was create an image of mass men storming
the polls in Venezuela, and that image looked very news
photo like, So, you know, I think it depends on
the prompt and it also largely depends on like the
(09:56):
data set that this thing is drawing.
Speaker 3 (09:59):
From, right from a believability standpoint max and in essence
like a picture quality standpoint. Do we have a sense
of where where Groc stacks up then or is it
sort of case by case.
Speaker 2 (10:09):
Hard to en I've struggled, frankly, as these AI companies
have you know, touted benefits from you know, one one
version of Chatchypet the next to see huge differences, I think,
you know, just just just taking a couple steps back,
Like we had conversations like this around Dolly and chatchypt
(10:30):
about a year ago. So it's like it's sort of
like grow or maybe a year and a half ago.
I can't it's it's all anyway, Yeah, Like these these
companies all have dealt with things like this. It's it
seems to some extent like Groc is just behind on
the like building guardrails around, like what you can ask
it to do and what it will generate, or perhaps
(10:52):
the willingness to to build those guardrails. Right, it seems
like Elon Musk at least partly did some of this
stuff on purpose, but I think, as Davey's saying, like,
the real problem here isn't like the AI tech, it's
the way it's integrated with a news platform like Twitter.
It's like easy to forget that like back in the day,
Twitter was like a relatively reliable place to find out
(11:14):
what was going on in the world, and you know,
the sort of remove all the blue checks, replacing it
with essentially a lot of fans vila musk, and then
you know, putting all this junk into the stream, you know,
sort of whether it's whether it's violent images or just
goofy images or just comedy.
Speaker 4 (11:31):
Like.
Speaker 2 (11:32):
The effect is that this thing that was once a
reliable way to like discover what was going on in
the world is incredibly unreliable and you have to be
really really careful as a user now.
Speaker 7 (11:43):
And I think that has like.
Speaker 2 (11:45):
Pretty corrosive effects on the way like not just Twitter
users like process the news, but the way that like
everybody processes the news and like and again, this is
probably a bigger conversation than just rock, Like, I think
you could level some of these similar critiques against open ai,
even the so called like careful versions of this.
Speaker 3 (12:08):
Mel Dana, I know you're not a you don't spend
a whole lot of time creating deep fakes out there.
That's not a hobby of yours. But what's your initial
impression of this new Groc tool.
Speaker 5 (12:18):
Yeah, so I haven't used Groc myself because I don't
pay for premium, but I just like I see it
in my feed, Like you know, some of the prompts
are like make an image of George W. Bush doing
a line of cocaine, or make an image of Barack
Obama doing a line of cocaine.
Speaker 6 (12:32):
And then you see the images.
Speaker 5 (12:34):
And like, on one hand, it's kind of like, oh haha,
that's a funny deep fake. But I don't have like
a lot of faith in media literacy in this country,
particularly in this era where so much content is created
to go viral, and like we don't teach media literacy,
and like a lot of us have fallen for stuff
that is fake, and like so often I'll be on
(12:56):
like text chains with like other reporter friends like is
this real?
Speaker 6 (12:59):
Like did you see this? Like like you just.
Speaker 5 (13:01):
You really like it's you have to like take a
beat and actually like try to figure out like what
where is this coming? Where is this coming from?
Speaker 4 (13:08):
Including last night, Trump sharing a picture of an AI
generated Taylor Swift appearing to campaign for him, telling people
to vote for him. Yeah, we're like totally in this
new world. I would not be surprised if he used
rock for that. Actually, it felt like it had that
quality to it.
Speaker 2 (13:26):
But and there is an interesting thing going on where
it's almost like you can't you don't know if Trump
knows it. There I saw on social media people sort
of saying like, oh, Trump doesn't realize that this is
an AI generated image, but but it almost is like
it doesn't matter, right, It's like and I feel like
Trump's in that post, there's sort of an embedded argument
that like it just it just doesn't matter. And the
(13:48):
other thing I wanted to say is just like there
are violent images of cartoon characters, Like on the Internet,
there is like a lot of this stuff can be found.
It's just you used to have to go looking for
it in the bowels of like the like the sort
of the worst websites on the internet four chan or
eight chan or eight two or whatever back in the
day Reddit before Reddit kind of clamped down on a
(14:10):
lot of this stuff. And so what really, it's just
like the the like four chanization of like the of
sort of mainstream social media. And it's just it's really
staggering to watch, given that like Facebook and Reddit and
all these other companies spent like a decade trying to
pull this stuff out of its platform and and seeing
it kind of all come back, and not only to
(14:32):
come back, but come back to the point where the president,
former president of the United States is is sharing this stuff.
It's it's it's wild.
Speaker 3 (14:38):
This seems to be a further merging of x and Xai.
You know, groc being part of Xai. I mean, heck,
are these companies just one and the same.
Speaker 2 (14:49):
Now technically no, they're two different companies. And and I think,
you know, and there's Xai is raising money.
Speaker 3 (14:55):
But so this Xai bill x for this, I mean,
is there is there money coming? I mean, not that
xes a lot of money, but that x is sending
a check to Xai for this service.
Speaker 2 (15:06):
It's a very good question. Also, I mean, every time
you run one of these image sure suits there, it's
pretty expensive. And one of the reasons that Xai is
trying to raise money. One of the reasons Elon Musk
is sort of teasing the possibility of maybe Tesla will
give XAI some you know, billions is because running the
in Vidia Yeah yeah, I mean running these like Nvidia
(15:27):
servers that power this is incredibly, incredibly capitally intensive. These
chips are very expensive. They require huge amounts of power.
But to answer your question like, no, I mean that
functionally they're operating as one company, Like there's no xai
outside of grop and and groc exists on Twitter. I
mean it feels like we've talked about this before. In
(15:50):
a way, Grock is sort of like a way a
landing for this very poor investment decision that Elon Musk made,
paying forty four billion dollars to buy x and really
the only way, in the same way that we talk
in this podcast about how really the only way to
pay off Tesla's valuation is to launch robotaxis. Kind of
the only way to make you know, to sort of
(16:11):
make the money make sense on the x acquisition is
if it turns into a you know, massive AI play,
which is sort of what he's trying to do.
Speaker 4 (16:18):
I will also add to that and say that I
think how capital intensive this is could be another pressure
point for these companies. You know, AI is very expensive
and if you're running, you know, sort of using these
large systems to create kind of.
Speaker 7 (16:41):
No you could say, say, create.
Speaker 6 (16:45):
Viral images that don't really have a utility.
Speaker 7 (16:50):
Like yeah, what what's the ROI on that?
Speaker 6 (16:52):
Exactly? So, so maybe that is one one way that
this kind of rolls back.
Speaker 2 (16:57):
Don't sleep on Dana's thing about unregret I was joking.
I think about unregretted user minutes. But I do think like,
as a user, this is really it's it's lame, Like
it's just it is not a good user experience. And
I think the check on this is not going to
be is likely to be the sort of economics it's
like what Dave's talking about, the money costs to run
(17:19):
these things and what consumers feel about it, which I
don't know, Like, I don't think it's going to be great.
Speaker 7 (17:25):
David Alba, thank you so much for joining you.
Speaker 6 (17:27):
Thanks for having me.
Speaker 3 (17:31):
All right onto the cyber truck. It's been in the
news lately. The New York Times the other day called
it a culture war on wheels, and you're seeing things
like the right wing influencer Aiden Ross gifted one a red,
white and blue one to Donald Trump, Kanye showed up
at a Trump rally and one and the Chechen warlord
(17:54):
rams End Kataroff posted a video of himself and a
cyber truck fitted with the gun turret on Monday, claiming
it was a gift from Musk, which Musk denied.
Speaker 7 (18:03):
Dana, where is this all going? Is this thing?
Speaker 5 (18:06):
You know?
Speaker 7 (18:07):
Is this the new? Is this what the cyber truck is?
Speaker 3 (18:11):
It's just sort of this cultural symbol and you know,
but we don't actually sell many of them.
Speaker 7 (18:16):
What's the story here?
Speaker 5 (18:18):
I mean, I think ever since the truck came out,
it became this cultural symbol, right because it doesn't look
like your standard pickup truck. It does not look like
a Ford F one fifty and the design of it
was very polarizing, and it's meant to be this kind
of like, you know, it's the truck for the apocalypse.
It's supposedly bulletproof, it's very angular. It was inspired by
(18:39):
Blade Runner. You could imagine driving this on the surface
of Mars or the military using it. And the truth
is that the cyber truck has been saddled with all
kinds of production problems, so it is not in high volume.
Is It's a very niche product. Tesla has not disclosed
when they report their quarterly production and delivery figures how
many they're actually making, and we only kind of have
(19:00):
a sense whenever there's a recall like of how many
are actually on the road, and it's it's not a lot.
I mean, there's maybe, you know, there's like thousands of
them on the road, but we're not talking high volume
at all. And then what happened this week was that
this you know warlord. You know, there's telegram video circulates
and everyone's like, oh my god, did Elon Musk really
donate a cyber truck to this guy? And the answer
(19:22):
is no. I mean, Musk himself wrote, you know on
X are you seriously so retarded that you think I donated.
Speaker 6 (19:28):
A cyber truck to a Russian general?
Speaker 5 (19:31):
But how the cyber truck got there is super interesting
because it does seem like there's this either used cyber trucks,
you know, being sent abroad or a black market for
them or an aftermarket, and this could be in violation
of sanctions, but it probably passed through several hands and
then it ended up, you know, with this guy.
Speaker 2 (19:48):
You know, the best case scenario for the cyber truck
when it launched was that and David, I think we
talked about this, you know, it becomes the hummer, right,
and it's like, yes, it's catchy, Yes it's a slightly
trolley maybe, but but in a way that has like
mass market appeal. And like those examples you rattle off, right,
(20:09):
this is not the Fox News demographic Kanye West, you know,
essentially a proud anti Semite chechen warlord. I mean, this
is like so far right wing. You know, it's obviously
like losing a lot.
Speaker 1 (20:23):
Of appeal for most people.
Speaker 2 (20:24):
I also think it's part of the problem is just
the price, Like this is this is what does it
go for now?
Speaker 5 (20:31):
Well, right now, it's like a one hundred thousand dollars.
Speaker 2 (20:33):
Yeah, and they and they just you know, took there
was an option to buy a like a base model
that is the sixty thousand or so. Dana, correct me
if I'm wrong, but they and they they got rid
of that, and we're essentially talking about like one hundred
thousand dollars car no matter what the cultural signifiers are,
is a niche product. And then if you layer on
the kind of weird you know, ultra right wing signaling
(20:57):
on top of it like it's it's a it's a problem.
Speaker 3 (21:00):
So this warlord showing it off as some badass weapon
of war does nothing to lure in any kind of
buyers of any kind.
Speaker 2 (21:11):
I'm sure like militia types, you know, like I don't know,
like cars have used their car companies have used like
the military applications.
Speaker 3 (21:19):
Well, I mean, think for a second about the Toyota Tacoma.
Right that Tacoma was. I believe for many years it
was jerry rigged by any number of militias and rebels
in far off places, and it kind of became a
thing that actually there was a They may have sold
a lot for those purposes, but I always sort of
got the sense that implicit in that was like this
(21:41):
this car, this thing is so badass and so rugged.
You know it can stand up to war. You know,
imagine what it can do for you.
Speaker 5 (21:48):
I think you just need to take a step back,
like what does the average person buy a pickup truck for?
It's for work? And then there are then there are
the people who like they need a pickup truck because
they want to you know, throw their surfboard in the
back or whatever. Well if that's you, then you can
buy a Rivian. I mean, if you look at the
way that the cyber truck is branded, it's not really
aspirational for like the weekend contractor or the family that's going,
(22:13):
you know, camping at Glacier National Park. There are truck
segments that kind of fit that market. The cyber truck
really is all about like being super tough and going
like off road and like going to Mars. And I
mean that's been clear from the beginning when they kind
of rolled this out. I mean Elon Musk demonstrated this
vehicle by trying to like throw a steel ball he
(22:34):
did go with and you know, like breaking it with
a sledgehammer. And so the way that they've positioned this
vehicle has just been kind of strange from the meget
go exactly.
Speaker 3 (22:45):
But let me ask, you know, playing Devil's advocate here,
let me ask you two a question, which is you know,
in a world in which you say, hey, no, you
know all news is good news. No, no, you know,
whatever the news actually is, you know, no news is
bad news. And what if it's so much in the
news and every time you see one go by, you know,
(23:05):
heads turn. I've seen it in New York City when
one of these things go by, heads turn and people
sort of start chatting. Maybe it's just a massive loss
leader that you know, is constantly throwing the Tesla name
out there.
Speaker 7 (23:17):
Maybe they ultimately won't sell all.
Speaker 3 (23:19):
That many of this thing, but I mean it's certainly
creating brand awareness, is it not, Dana.
Speaker 5 (23:26):
Oh, absolutely, and over at the Thanksgiving break last year,
you know, they had cyber trucks and showrooms all over
the country surrounded by these like velvet ropes, and the
whole point was to get like holiday shoppers to come
into a showroom, see the cyber truck and then buy
a model. Why and you know, Tesla has not spent
a lick of money on advertising for this vehicle, but
(23:47):
you certainly know when you see a cyber truck that
it's made by Tesla.
Speaker 3 (23:51):
I had entirely forgotten Max about that anecdote that Dana
just brought up, but it is something. Yeah, it's kind
of exactly what I was thinking. Maybe, right, you know,
you've got this thing out there, not thinking that you're
going to sell that many of them.
Speaker 7 (24:03):
But yeah, but boy, but here's a why for you.
Speaker 2 (24:05):
I mean, maybe and I do think, Look, you have
to say, although the cyber truck has not sort of
been a success according to the metrics that Elon Musk
originally set when he unveiled it, like it has sold
well relative to other high end electric pickups, like like
what like the EF.
Speaker 1 (24:25):
One fifty lightning.
Speaker 2 (24:26):
It depends on the numbers you're looking at, but like
it seems like the cyber truck may now be out
selling the F one fifty line for instance. Yes, but
again these numbers are tiny, and it's because no one
is buying these very very expensive relatively few people are
buying these very very expensive electric pickup trucks, whether they're
made by Rivian or Tesla or Ford. And so in
(24:49):
that sense, like, sure, yeah, maybe this is a really
great loss leader, a great you know brand statement. But
first of all, I would question whether being associated, notwithstanding
your point about the you know, marketing prowess of the
fact that like isis fighters use Toyota Helux's, I would
question whether it's a great, great branding to have your
product associated with a Russian warlord during during a major
(25:12):
international conflict. And also, pickup trucks sell a ton. It's
like the biggest category of vehicle in the US. You
don't want your pickup truck to just be a lost leader.
You would like to if you really trying to make
a successful electric car company, you don't want to sell
some freaking pickup trucks.
Speaker 6 (25:27):
And like that is not what's happening right for Tesla to.
Speaker 2 (25:30):
Be the kind of big, mass market EV company that
Elon sort of used to talk about.
Speaker 5 (25:36):
And it's you know, longtime Tesla owners know, like you
don't ever really buy like the first wave of vehicles
because there's there's gonna be all these production kanks. You
wait until they kind of figure it out and the
vehicles will get better as they kind of like work
through these production issues. But you know, I wrote a
story quite a while ago that said, like the cyber
truck is in production hell, and like we are still
(25:57):
in production hell. And so there's not like a huge
appetite for these yet people are waiting. They're waiting for
these problems, and they're waiting for the price to come down,
I mean one hundred thousand dollars in this kind of
inflationary environment where interest rates are high and if you
have to finance a vehicle, the financing is very high.
Speaker 6 (26:13):
Is a huge turnoff to a lot of people.
Speaker 5 (26:15):
It's a major, major purchase.
Speaker 3 (26:22):
Okay, to wrap up, we're going to talk about I
should say, we're going to go back to the to its,
to its military capabilities, and back to the battlefield because
apparently the Chechen warlord Kataroff says he's going to unleash
it in Ukraine. He's gonna essentially turn it over to
Poutin and let let him use it in Ukraine.
Speaker 2 (26:41):
Can I just say something about this branding argument you
made earlier, because like it's it's bothering me because like
the like you know, random armies are using Toyota tacomas
that was promised on the idea that these things were reliable.
Speaker 1 (26:55):
It's basically this is everywhere.
Speaker 2 (26:56):
It's omnipresent. Yeah, Like this Chechen guy wasn't using it
on the battlefield.
Speaker 7 (27:01):
Well we're about to find out he was just driving
around looking like a jerk.
Speaker 2 (27:05):
And in that sense, he looks like every other cyber
truck driver and so like it's.
Speaker 7 (27:10):
Not every single cyber truck driver is a jerk.
Speaker 2 (27:13):
No, I mean, I'm just saying, like it's a peakhawking gesture,
Like it is not a like even like he's not
showing off the capabilities of this thing. He's doing the
thing that if you were going to criticize the cyber truck,
that is just this showy, kind of goofy.
Speaker 1 (27:27):
Like manly, like lame, you know, hummer like thing.
Speaker 7 (27:31):
And that's what how he was using it.
Speaker 3 (27:33):
Oh sorry, all right, now that you got that off
your chest, when it goes to battle, how will it fare?
Speaker 7 (27:39):
Mister Chafkin.
Speaker 2 (27:42):
I don't even want to contemplate this, honestly, it's so
gross and dark and like the idea of you know
this like this like Chechen warlord, you know, using Elon
Musk's like you know, like.
Speaker 7 (27:59):
Midlife crist vehicle on the battlefield.
Speaker 2 (28:01):
Like I don't know, Like my guess is it's going
to be arrowproof, but because we've seen Joe Rogan fail
to shoot it with an arrow. But I would not
have confidence in its military prowess, Danna.
Speaker 5 (28:14):
I mean, I'm allowed to say, like I do see
a lot of cyber trucks in California, most more in
LA than in the Bay Area. But you know, like
I could see it having military applications, I guess, I mean,
but you still have to charge it you have like
and supposedly it's arrowproof bulletproof? Is it like bomb proof?
I don't know, but I mean, Musk is part of
(28:35):
the military industrial defense complex, so you know, maybe he's
trying to sell these to the Pentagon and like this
like sort of weird thing of this Chechen warlord happening
to get his hands on one. Maybe maybe that'll perk
interest in some in some you know, procurement office somewhere.
Speaker 3 (28:51):
My best guess here is that when Zelenski makes the
final push into Moscow, he will do.
Speaker 7 (28:55):
So atop a cyber truck.
Speaker 3 (28:59):
Christ Max Dana, thanks for joining.
Speaker 6 (29:04):
Great to be here anytime.
Speaker 3 (29:13):
This episode was produced by Stacey Wong. Naomi Shaven and
Rayhan Harmansi are senior editors. The idea for this very
show also came from Rayhon Blake Maples handles engineering, and
we get special editing assistants from David Purcell. Our supervising
producer is Magnus Henrikson.
Speaker 7 (29:30):
The Elon Inc.
Speaker 3 (29:31):
Theme is written and performed by Taka Yasuzawa and Alex
Sugi Eua. Brendan Francis Newnham is our executive producer, and
Sage Bauman as the head of Bloomberg Podcasts. A big
thanks to our supporter Joel Weber. I'm David Papadopolis. If
you have a minute, rate and review our show, it'll
help other listeners find us.
Speaker 7 (29:51):
See you next week.
Hosts And Creators
Sarah Holder
Saleha Mohsin
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Las Culturistas with Matt Rogers and Bowen Yang
Ding dong! Join your culture consultants, Matt Rogers and Bowen Yang, on an unforgettable journey into the beating heart of CULTURE. Alongside sizzling special guests, they GET INTO the hottest pop-culture moments of the day and the formative cultural experiences that turned them into Culturistas. Produced by the Big Money Players Network and iHeartRadio.