Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:11):
Taylor, Natasha, really good to see you all. I want
to ask you a day about the sunsetting of ARS
dot com and any any fond farewells you will have
for the artists formerly known as ask Jeeves, who I think,
to my mind kind of scooped the chatbot moment by
almost thirty years before before dying too young.
Speaker 2 (00:30):
So sad.
Speaker 3 (00:31):
I was gonna that was what I was saving for
my worst week in tech. That was gonna be my joke.
Oh man, you stole it from me. Very sad end
of an era.
Speaker 4 (00:40):
I had no idea they were still around, but it
kind of makes me sad when these niche services die.
I want to like talk to the users that are
still using it. You know, I'm curious.
Speaker 3 (00:49):
They could have just merged with SpaceX, Like, I don't
know why they had to shut down.
Speaker 1 (00:52):
That's true, Barry did it didn't have enough vision for
the SpaceX merger of US dot Com.
Speaker 5 (00:57):
Taylor, you missed it, read as an AI engineer are now.
Speaker 4 (01:00):
Oh that's awesome.
Speaker 3 (01:02):
Cong that.
Speaker 2 (01:05):
All are all right?
Speaker 4 (01:05):
Read well, you're not going to be stuck in the
permanent underclass, so take us with you.
Speaker 1 (01:09):
You are flexing. Before we started about how you you
have vibe coding into the early hours last night.
Speaker 3 (01:14):
I was I was vibe coding trying to figure out
how to check my email more efficiently. We also have
a new product out today since you just open the
door for a plug sorry, Semaphore Intelligence, which which is
like came out of me spending an hour vibe coding
this app after our World Economy Summit. It looks at
(01:35):
like every single statement that people made on these two
hundred and fifty panel discussions we had and finds the
you know, quote unquote consensus on different issues. It's not
an I wouldn't call it an AI product that we're
sending out into the world because it was very much
like human edited and human engineered and designed. But it
wouldn't happen without AI. So yes, I'm an AI engineer,
(01:59):
and I think I need a raised something like a
billion a year. That's my final offer.
Speaker 1 (02:06):
Okay, let's get into it. Welcome to Tech Stuff, MS Valsh,
and this is the Week in Tech where I'm joined
by three of the most plugged in reporters to break
down what's really happening in.
Speaker 2 (02:14):
Tech right now.
Speaker 1 (02:15):
Today, we're joined by Reid Albergotti, tech editor for Semophore
and tech founder and attached Tiku, tech reporter for the
Washington Post, and Taylor Lorenz of User mag read. I
want to start with you. There was a story this
week in the New York Times that the Trump administration
wants oversight of AI models before they're released. And this
was kind of a stunning about face because I remember
(02:37):
Vice President of Vance's trip to Europe last February where
he's scolded all of the European leaders about their you know,
the fact they were strangling AI's development with all of
their nasty regulations. What's going on here?
Speaker 3 (02:50):
Yeah, I mean I wouldn't if you're an AI safety
person like I honestly wouldn't get your hopes up too
much based on this report. I think from what I
can tell just docing to some people, you know, in
the White House this week, I think it's like a
there's a bit of like tension between a couple of constituencies,
you know, from the cybersecurity side and the and the
(03:11):
AI side and the policy side. So it's I don't
think we're going to see something like a like a
draconian regulation where you have to get a license to
sell these models. I think it's much more like, hey, uh,
you got to check in with us like this, this
mythos thing kind of freaked people out in DC. You
(03:32):
were there a couple of weeks ago. Odds, I saw
you at our conference that everyone was talking about it.
So I think it's I think it's like, oh, you know,
it's not a it's not like we need to see
these models because they might, you know, become sentient and
take over the world. It's like, no, there's like some
real consequences when these things come out, and we want
to get ahead of it. That's that's sort of my
(03:53):
takeaway from just the conversations that I've had.
Speaker 1 (03:55):
Natasha Tita, What have you been hearing about this?
Speaker 5 (03:58):
Yeah, sure, Like the folks I've been talking to, I mean,
it's really interesting how over the past i don't know,
three years, like the words licensing regime are supposed to
strike fear in the hearts of all, Like there's actually
I mean, the way people have talked about it with me,
it's like way outside the Overton window. It's such an
extreme policy. Even AI safety people are against it because
(04:22):
it would slow down the release of these models.
Speaker 1 (04:24):
So I don't know what the license licensing.
Speaker 5 (04:32):
Okay, let me back up, let me roll back. Basically,
what they're saying is, you know, they want companies want
to be able to release things on their schedule. You know,
we've heard from David Sachs, from Shriram Krishnan, from Mark
Andresen like anything that slows down innovation is bad for
(04:52):
the US, good for China and including by the way,
at least at least where andreason is concerned. And so
you know, voluntary commitments were fine. That's what happened under
the Biden administration, although you know there was pushback against that.
But the idea that you would need to prevat something
with the government has been referred to as a licensing regime.
(05:16):
And those words are the ones that are supposed to
strike hearts, a strike fear in the hearts of all
of all innovators and builders. So I just would be
very surprised if something, you know, like Reid said, there
is it seems like the concern around mythos, the fact
that national security is involved. And now you know, there's
(05:37):
lots of different constituencies, they have actual analysts telling them,
you know, here are the capabilities and here are the risks.
Speaker 1 (05:45):
But it is one of those things. I mean, like
David Sex, who was the AI sar and the kind
of great booster for the AI industry within the White House,
has left his position. Is this one of those moments
where like there's a mega split and the kind of
regulating side or the more anti tay excide or an
ascendency or is this just kind of post mythos noise,
Like how how significant is this?
Speaker 2 (06:05):
Well? I I just don't think.
Speaker 3 (06:06):
I don't see it like that so much because I
think actually shri Ram was sort of maybe moved a
little bit closer to Natasha mentioned Treron Chrissans, who's still
at the White House, was sort of moved I think
a little closer to the to the you know, inner circle.
So he's still there, and I think still very much
has those those views. But I think that there is
(06:29):
not like a it's not so much a maga non
maga split. I think there's just different there's different philosophical
ideas about cybersecurity and how much we should be worried
about mythos. And I think David Sachs tweeted something along
the lines of like this is sort of an the
mythos thing is a little bit over overblown, I think
was the gist of it. So I think it's I
(06:52):
think these differences are more subtle than like maga non maga,
if that, if that makes sense.
Speaker 5 (06:57):
But there is like best at lutnix use you whiles
are definitely stepping into the power vacuum of sex being.
I wouldn't say he's out of the mix, right, I
mean it's still a little bit TBD, but I think
he's still you know, present in the discussions. But yeah,
their concerns about you know, uh Besen has talked about
(07:19):
the need for balance is just different rhetoric too than
what we've heard from from sex.
Speaker 1 (07:23):
Yeah, Tayla.
Speaker 4 (07:24):
I think it's just interesting that like pretty much no
one on any side of the issue seems to like
this idea and this regulation. Like Kevin Bankston at the
Center for Democracy and Technology tweeted, tech lefties, tech wright's
and AI safety heads all agree this is stupid. And
he's quote tweeting a Politico article about this policy. So
I feel like it is misintentioned. I don't know. I mean,
(07:47):
David Sachs is definitely like his sort of I don't
know if it's fair to call it like a full
exit or not. But he's definitely created a power vacuum.
And what I'll say about Streiram is like he has
no juice, he has no real like ability, he has
no lef ability. I would say, to push things the
way that David Sax did. Like, I think David Sax
is a much more aggressive like character. He's a lot
(08:08):
more powerful. Streem is sort of like this former and
recent guy that's very like I don't want to say
like Beta, but like I feel like I'm talking and
like looks Maxinger like clinic ular terms, but he's just
not like I wouldn't.
Speaker 1 (08:21):
Say he's very cool. He's six foot eight, he is.
Speaker 4 (08:24):
This has nothing to do with how he looks. It's
more just like I think David Sax is a real
fighter and has a real perspective on things. Street Ram
sort of notoriously will just cater to anybody who's in power.
Speaker 1 (08:33):
So yeah, well, I mean to talk about the Are
you a tech left detailer?
Speaker 4 (08:38):
I would call myself on the left, and yes, I'm
a tech head, so yeah, fair.
Speaker 1 (08:42):
So why well tech lefties in general? I would say,
like a pro regulation I think of AI like or is.
Speaker 4 (08:49):
That well certainly we're pro smart regulation. I think what's
been quite terrifying for everyone on the left lately is
that there's a lot of really bad tech regulation that's
being sold to try, you know, by the far right
and buy some of these more extremist groups as like, hey,
don't you want to crack down on big tech? We're
going to do that through mass surveillance. And I think
what myself and others on the left would argue is,
(09:11):
you know, we want smart policies, not stupid policies that
aren't going to achieve our ultimate goals.
Speaker 1 (09:17):
What do you think the administration is responding to? Is
is it genuine fear of a myth thoss?
Speaker 2 (09:20):
Is it?
Speaker 1 (09:21):
Is it midterms coming up? And Pure Research released research
that fifty percent of Republicans and fifty one percent of
Democrats are more concerned than excited about AI in their
daily lives. And what's what's is this a political response
or what's do you have any idea read Natasha? What
are people in the administration is saying about why this
is happening?
Speaker 3 (09:39):
I think that I think this idea that they might
need to come out and say we're going to do
something about AI. I think that might be political, Like
there might be because it's definitely not popular, right, the
data centers are hugely unpopular, So it could be that
the administration's thinking like, yeah, we should we should sound
concerned or something about AI. But I think there is
(10:00):
actual genuine concern about MYTHOS, and some of it may
be overblown, but it's like not political, it's it's actual
concern that you know, this could cost us a lot
of money and economy. I think there's even probably concern
on intelligence, like mythos being used defensively, like we the
US has you know, intelligence has the upper hand right
(10:22):
right now because people, you know, they find they buy
and find all these vulnerabilities that allow them to hack
into basically anything they want. And if those if those
back doors close, that's bad for the NSA. I that's
like me, just that's just conjecture on my part, Like
that's not based on any real reporting on my part.
Speaker 2 (10:42):
But you know, I think it's got to be a concern.
Speaker 5 (10:45):
Yeah, I definitely went in as especially I think I've
been looking at like super PACs and midterm elections, went
in thinking, oh, it has to be a response to
what is increasingly feeling like this, you know, populist uprising
or at least it's more of a central issue. And
everyone I spoke to said no, like this administration does
not respond to pulling in that way, like, look, cost
(11:09):
of living is the top issue. You know, people don't
like the ballroom. You don't see them moving on that that.
This is much more like you know, once national security
is brought in, it's a bipartisan thing. It's very hard
to argue against. And yeah, I was. And they also said,
you know, the polls have been bad for a while,
so it's it's not that it's really hard to not
(11:33):
imagine that they want to at least gesture at we're
doing something on AI.
Speaker 1 (11:40):
Yeah, I mean so, And how do the tech industry
for Are they like, oh, this is like this is
a great fig leaf, this is actually a relief rather
than real regulation, or are they actually nervous about this?
Speaker 5 (11:50):
No, I think it's been like, look, this is what
happens when you refuse to have any kind of reasonable
dialogue about oversight. Like I think you hear a lot
of Dems say, like, you know, we would have done
we would have work done a federal bipartisan regulation, but
you tried to do a moratorium for ten years. That's
(12:11):
hugely unpopular. You know, you have the Trump administration going
after Red States for their you know, for their AI
AI laws and I don't know what like federal bipartisan
legislation sort of means like lots of fig leafs or
a lengthy fagi leaf or something true. But yeah, I
think the sentiment and it kind of carries over into
(12:32):
data centers too. It's like total inaction plus public backlash
equals bad regulation.
Speaker 1 (12:41):
Taylor, you said your four pro your favorite smart regulation.
What does that mean to you? And who has pitched
the smartest regulation that you've heard?
Speaker 2 (12:49):
Like?
Speaker 1 (12:49):
What should we be aspiring to?
Speaker 4 (12:51):
Yeah? I think, I mean, I think I think AI
is a scary thing, and I think that we're right
to be concerned. I think a lot of I mean,
I would consider myself like pro technology. I think a
lot of the left is potentially not as pro technology
as I am as a tech reporter and somebody that
loves tech. But I think overwhelmingly the concerns are around
labor and job loss and what do these transitions look like?
(13:13):
And how can AI be implemented in things like healthcare?
You know, can it be used to deny you housing opportunities?
Like where what is it? I think a lot of
us want like what are the boundaries there? And I
think a lot of us also just want transparency. How
are these companies being run? You know, what do these
contracts look like with the government? And also, you know,
when it comes to these data centers, I think a
(13:35):
lot of people on the left are also very concerned
about you know, environment, you know, cost to local communities.
I think some of the water stuff is completely overblown,
but I think it's you know, making sure that these
data centers are built responsibly and sustainably and don't add
to you know, the power burden non sort of existing communities.
Speaker 1 (13:53):
We've got to talk about data centers later in the episode,
but read like what is it? What are people in
the tech industry feed I mean, is there a sense
that the midterms, well the next Democrat president may may
kind of swing the the sort of ark of how
these companies are treated in a in a way which
could be very disturbing to their interests. And what's what
(14:14):
what's the level of kind of excitement that the wheels
are off versus fear that the good times may be
coming to an end.
Speaker 3 (14:20):
I think there's always I mean, they all have these
big policy shops you know who, and they're always you know,
they're out there lobbying, and they're always kind of like,
as a as a rule, freaked out about what is
the next shoe to drop, like what are what are
the new regulations They're always concerned about, you know, Europe.
I mean, this is just a it's a constant battle
that they've been fighting. I think before before chat GPT, right,
(14:41):
it was it was social media before that, and you know,
privacy and all this stuff. So I think I think
that I don't notice like a spike because of this
news this week or anything like that. I think it's
I think it's sort of the same level of like,
let's you know, we've got to just we've got to
be involved, we've got to be engaged in the process
and try to shape it in our favor as much
(15:05):
as we can. But I do think like in general,
this this theme of AI backlash, which is something like
you know, we've been writing about it semophor for a
while and I think we were very early on that,
like the backlash to the data centers, et cetera. Is
a real problem. Like it's very it's very real. I
mean it's part I think it's partly why Like they
(15:25):
want to build data centers in space, I.
Speaker 4 (15:27):
Mean, which I think kind of sounds like a good idea.
Speaker 2 (15:30):
Yeah, put it in space? Why not? I mean.
Speaker 5 (15:35):
It tesla there.
Speaker 4 (15:39):
I mean, there was so much sort of backlash when
Elon Musk somebody these you know, proposals to to put
I think it was something like a million or something.
It was quite it was a lot of satellites in space.
But I just think, like you know, if you ask
a lot of Americans, where would you rather have a
data center somewhere up in space? Yeah, potentially polluting our
atmosphere or directly next to your house. Like I mean,
(16:01):
I just think there's like major noise concerns with these
data centers. There's environmental concerns, and I think a lot
could be addressed by putting them up there.
Speaker 2 (16:10):
You know, the noise concerned. Though.
Speaker 3 (16:11):
They do have these this hum. The data centers have
this hum. That's actually of all the complaints that people have,
that's the one that I sort of relate to the most,
Like if there was this thing humming all the time,
like I that would really bother me. But I think
the new ones are the liquid cooling will be a
lot quieter.
Speaker 1 (16:28):
Actually got a whole segmental data center, so I don't
want to got it. We don't want to. Oh no,
we don't want to serve too much. Just before we
go to the break though, is there? I mean there's
been discussion of working groups between government officials and tech executives.
Is this and what's the difference between that? And you know,
President Trump sitting flanked by all the tech executives kind
of chising with Diret Coke, like what's the what's the
(16:50):
is it? Is it going to be? Is anything meaningful
that comes out of this? Or is it like a
bell weather for something the more meaningful that may come
in the future.
Speaker 2 (16:56):
Working groups? I mean, come on, you guys.
Speaker 3 (16:58):
These guys are at the White House all the time,
like anytime they want, like Trump will pick up the phone.
I remember, like you try to get people to come
out from the from Silicon Valley to our like DC events,
and I'm like, you'll, we're gonna have all these people
from the administration there.
Speaker 2 (17:10):
It'll be great. You can talk to them.
Speaker 3 (17:11):
They're like read, I can go to the White House
tomorrow if I want to like it's like that is
not a.
Speaker 1 (17:16):
Like you so much feel gracious invitation, but unfortunately I'm
otherwise engaged at the White House.
Speaker 2 (17:22):
I mean, these people are running, they're running the White House.
Speaker 1 (17:26):
All right, We're gonna take a short break now. When
we come back, Taylor is going to tell us about
a dark money campaign paying influences to frame Chinese AI
as a threat. Don't miss it. When you're traveling abroad,
(17:52):
Wi Fi is always an issue. Maybe you can't find it,
or maybe you can, but the Wi Fi network that's
a vain looks a little bit sketchy and makes you
think it might not be safe to connect to. I've
certainly had that feeling, and it's why having a local
simcard on your phone can make all the difference. But
you don't have to literally buy one. That's where Sale
(18:13):
comes in, an e SIM service that is as simple
as downloading an app, and that gives you instant Internet
access wherever you're traveling. That is safe and secure Internet
with built in cybersecurity features like web protection and ad blockers,
and no need for local Wi Fi. All you have
to do is download this salely app on your mobile
device and choose the plan that works best for you.
(18:34):
Once you install, the e sim sale will be instantly activated,
so you can start your trip off right. Download salely
in your app store and use our code tech Stuff
at checkout to get an exclusive fifteen percent off your
first purchase. See details in the podcast episode description box.
Welcome back to tech Stuff, Taylor. You've wrote a fascinating
(18:54):
story for Why this Week that I alluded to before
the break, which I understand all again. When you approached
yourself to appear in an influencer marketing campaign about AI,
what happened?
Speaker 4 (19:07):
Yeah, So I got this quite shocking email in my
inbox that I can only assume was AI generated. I'm
assuming this AI super pac was using AI to reach
out to people, because why on earth would you put
a tech reporter and a tech reporter who's known specifically
for reporting critically on dark money influence campaigns, you know,
(19:28):
on this sort of outreach list. But so, yeah, so
I got this sort of vague email that was like, hey,
you know, we want to pay you thousands of dollars.
There could be this potential for ongoing work to like
promote AI. And what stood out to me because I
do get sort of spam. You know, I have a
big sort of YouTube channel and stuff, so I get
sometimes offers from like a cell phone case company or something,
(19:49):
and I usually have to decline most of these offers
for conflict of interest. But it was it was kind
of vague, and it was like, you know, we want
you to talk about AI in China, but it didn't
really say who was sponsoring it. So I got back
and I thought, is this a big AI company? You know,
sort of who's behind this? I replied, And I was like, hey,
I'd love more information about this. By the way, my
auto signature says you know Taylor Lorenz Tech Reporter and
(20:12):
this marketing this influencer marketing company got back to me
and said, you know, it's for this group build American AI.
But don't worry. They don't want you to mention them,
tag them or anything in the post. We basically just
want you to repeat all of these talking points. And
I said, wow, that's fascinating. Who's funding this? And she
(20:33):
got back and said, oh, well we've got you know,
execs from palent here and open AI and so yeah,
and also here's some examples of you know, previous influencers
who have said yes to this campaign, and here's their
content that went live. Easy if only. But but you know,
I did want to sort of confirm all of this
(20:53):
and also speak to some of the influencers involved. I
looked at you know, Notice has done some great rewarding
and axios previously about Build American AI and the super
pac that they're affiliated, which which is leading the future.
And so I spoke to a bunch of other influencers
that were approached as part of this campaign. I was
actually at the White House Correspondence dinner, uh, you know
events that weekend when all this is happening, and I
(21:16):
was going up to all these other political influencers. I
was like, Hey, were you approaching and they were like yeah,
I actually got that email too. One of them was
considering taking it.
Speaker 1 (21:23):
Rita touchie one invited to the Putty.
Speaker 3 (21:26):
Oh man, I need to work on my my influencer clout.
How many do I need more than a couple thousand followers?
Speaker 2 (21:34):
Is that?
Speaker 4 (21:35):
I think you do need a few hundred. I have
over half a million on TikTok reads, so I might
be uh, And I love.
Speaker 5 (21:41):
This as a genre just like people mistakenly approaching Taylor
because she has like such a huge following. It's just
it makes me, you know, that's just always like people
are like, why would they email Taylor?
Speaker 1 (21:55):
What was the moment when you when? When the when?
The when? The when the when the executive sort of
stomach hit the floor and realized that you weren't about
sign up for the five thousand dollars, but instead we're
about to expose the campaign.
Speaker 4 (22:06):
Yeah, I mean, I was shocked at how quickly they
sent me the documents. I mean clearly, like again, I
believe that they were and this is just my suspicion
is that they were using AI for outreach. And I'm
assuming I because again I have a large verified following
on TikTok specifically, and this campaign was mostly TikTok and Instagram,
(22:26):
that it was just sort of sending general outreach. And
then this probably I'm assuming low level staffer is just
sort of blanket spamming out information. She sent me this
influencer brief, all of this documentation. Again as I wish
my job could be this easy all the time. But
when I followed up and I said, well, thanks for
sending me all this info, I will be reporting on
(22:46):
this for Wired. That's when the boss came in and
they started to threaten to assue me and all this stuff,
and which was hilarious.
Speaker 2 (22:53):
I did stress you for what.
Speaker 4 (22:56):
For god knows what? But can I just read you
what my TikTok bio is again, This is they are
approaching me to pay me thousands of dollars for Build
American AI, a dark money group. Because I report so
frequently on dark money influence campaigns, my TikTok bio literally
reads independent journalist who has never accepted a penny of
(23:17):
dark money funding.
Speaker 2 (23:21):
I think accept it. Get you.
Speaker 4 (23:24):
We were joking. One of the editors was like, well,
they were like, you know, it doesn't say that you
never would accept dark money.
Speaker 3 (23:33):
The Taylor like, how amazing would that be if they
sued you? Because they're using AI clearly to send these
things out, Like could you imagine the discovery it just
the AI chat you know whatever traffic I mean, it
would be increasing.
Speaker 4 (23:47):
And also all the influencers that took that secretly were
promoting this messaging campaign on behalf of this notorious dark
you know, dark money group that's working with this super
pac I mean the money behind it is Greg Brockman,
Joe Lonsdale, great.
Speaker 1 (24:01):
Fact when the co founder of open Ai, co.
Speaker 4 (24:04):
Founder of open Ai, co founder of Talentier, the two
co founders of Andresen Horowitz. This group has been some
of the most aggressive anti regulation people. They're pushing this
messaging about China because that's sort of their talking point
to sort of stave off US regulation. But yeah, I
mean it's it's hilarious. I would love if they see me.
Of course, the threats just disappeared and the story went
(24:26):
up and you know, honestly build American AI and leading
the future. I did speak to them as well on
the record for the story and they were kind of like, well,
you got us. They were like they gave a statement.
They were like, well, yeah, we're doing this campaign and okay,
we didn't really mean to reach out to you, but
you know, we have to compete, and we do believe
these things about China, and you know, we don't regret it,
(24:47):
and so at least they're brazen about it.
Speaker 2 (24:50):
Should we like take anything influencers say seriously at all? No?
Speaker 4 (24:54):
And this is this is what drives me crazy and
like why I continue to write these stories. You know,
I did another big investigation for Wired last year on
this Democrat you know, dark money group paying people. I
just these influencers accept these deals constantly with no disclosure,
and this is like about as unethical as you can be.
Speaker 1 (25:14):
I mean, you guys don't know like illegal on the
Paolo or is polar only applied to like mainstream media. No.
Speaker 4 (25:20):
So the thing is, and I've written extensively about this,
not only do is the FTC kind of asleep at
the wheels. So a lot of these influencers don't disclose
brand deals, but the brands. The brands are more legally cautious.
So if you're doing a campaign for Unilever, the influencers
probably doesn't care that much about compliance. Dove's chief compliance
officer does care. The issue is is that there is
(25:41):
no regulation for political content because it doesn't fall under
the FTC. It would theoretically fall under the FEC, which
is basically a sort of non existent entity that has
no interest in regulating the space. So what we've seen
over the past couple of years, really especially since twenty
twenty four, is just this vast proliferation dark money groups
C fours that are that they don't have to disclose
(26:04):
who's funding them, and the influencers don't have to disclose
that they got paid to put push messaging, and so
they're out here pushing political messaging left right and center,
and I think it's it's so corrosive to democracy. We
have a quote from a media scholar in the story saying, like,
you know, this is this is like extremely bad for democracy.
Speaker 1 (26:22):
Taylor, could you give an example of how it's showing
Like how is the how is the kind of content
showing up in the influencer ecosystem, Like what are the influences?
How are they delivering on the talking points?
Speaker 2 (26:32):
Wow?
Speaker 1 (26:32):
Are their influences?
Speaker 4 (26:34):
Yeah? I mean the content that we saw from these
influencers is they're literally repeating almost verbatim, the messaging from
this dark money influence group, really the super Pack, and
they're just putting it out there and their content making
videos as if this is something that they believe and
thought of and came up with, and there's no just again,
(26:56):
there's no disclosure of who funded it, why they funded it.
I mean, some times they'll hashtag gets sort of buried
down just a generic hashtag ad, but that almost makes
it seem like a brand paid them and nobody. There's no. Again,
there's no, and that's that's like maybe one percent of
them do that. Ninety nine percent of them don't disclose
it as an ad. I mean this dark money group
that I reported on last year that has been funding
(27:17):
hundreds and hundreds of influencers to push overt political messaging.
They they didn't disclose any of it. So your followers,
you as a consumer, And we just saw this poll
come out recently that you know, the majority of gen
z get their news directly from influencers. They're just consuming
paid messaging left, right and center. And I just I
also think that I know every influencer thinks their beat
(27:39):
is the most important beat, but I really feel like
these influence campaigns are affecting everything from business to politics, everything,
Like we have a huge amount of money getting in here,
and there's really not critical reporting on the influencer economy.
I think most tech reporters are so focused on the
actual tech leaders as we should be, right like Elon
Mussam Altman and stuff. But it's like, hey, let's take
(28:00):
a look at what's happening on these platforms. It's gotten
really bad, and I just think we need a lot
more oversight and critical reporting on all that stuff.
Speaker 1 (28:10):
Natasha, you've been also reporting on AI and super PACs.
I guess maybe not from the influencer angle, but but
what what have you been She read out influencer stories
to you.
Speaker 5 (28:20):
Yeah, Yeah, I was looking at I was looking at
the AI safety faction. You know, that is the folks
who you can think of them as sort of anthropical ligned.
You know, a lot of them came from the effective
ultruist rationalist community. They're the ones who are concerned about
existential risk around AI, and they just they had like
(28:47):
it's it's sort of been a decade long project for
them to have influence in a lot of these elite circles,
like through through nonprofit, through philanthropies like open philanthropy, which
is now called coefficient giving. There was actually like a
whole plan that they laid out in twenty sixteen that said, hey,
(29:08):
you know, there's not a lot of respect, there's not
a lot of prestige for people talking about extinction risks
and you know, existential risks from AI. So let us
fund you know, think tanks, trainings for people to study
AI alignment. That means, like you know, making AI safer
(29:30):
in a like having technical solutions. So they've been influencing
all of these elite circles journalism, politics, policy, academia, you know,
without that much money. But they've only just recently started
embracing like content creators and influencers. And I thought that
(29:50):
this was just super interesting and happening, you know, very quietly.
But this this group has so much money, and so
they're starting to bring in kind of regular people who
are really new to this language, to this idea, which
is a you know, very like Bay Area native concept.
(30:10):
So I had gone to this event at this like
rationalist hangout in Berkeley where they had this class of
content creators and it was really interesting. You know, some
of them came from climate change, some of them came
from book talk. You know, they had all expressed a
little bit of interest in this. But there's there's different
programs going around. And then I, you know, a couple
(30:33):
of days after my story came out, this new this
new program called Please Please Don't Kill Us, which was
about AI extinction and bringing a bunch of like making
it into a creator house for influencers for AI safety
came out. So I felt very vindicated that I.
Speaker 1 (30:51):
Was on top of things like it creators of people
like living to influencers living together. I mean it's the.
Speaker 5 (30:57):
Same house, like this is where all the events that
are AI safety focus happen. It's this like kind of
they call it a campus. It's a former hotel, kind
of a sprawling area, and yet it's like they'll pay
for you to be there. The only rule you don't
even have to make good content. The only rule is
you have to make content every single day or you'll
get kicked out.
Speaker 1 (31:16):
And are you serious?
Speaker 5 (31:18):
Yeah, I mean that's like, yeah, Winser Rules Publisher.
Speaker 4 (31:24):
We're gonna, yeah, we're gonna have the AI safety content
house on one corner. We're gonna have the AI Pro
AI super Pac content house on the other corner. It's
gonna be a whole, you know, a whole mess.
Speaker 5 (31:37):
Yeah, Taylor is really right. I mean it's wild the
distinction that you're talking about, where, you know, if it's
a political candidate, you have to go with the FEC.
If it is an exact product, you have to go
with the FTC. But even Mark Andresen was talking very openly,
he was referring to this AI safety influencer thing, but
he was like, there's this huge gray area where you
(31:59):
can put any amount of funding you want. You know,
if you are just talking generally about vibes or ideas
or like political concepts, and A sixteen Z has their
own they have I think they've done two classes now
of their new media fellows. There's been about like a
dozen or so disclosed, but there's sixty five of them.
(32:20):
So you know, it's not only the bild American AI
paid folks, I mean folks that are paid in that way.
There's also you know, and they've been they've been signing up,
like some of the most popular tech Twitter accounts.
Speaker 1 (32:37):
For exclusive deals like could you could you be shilling
for the AI safety community one week and then the
pro AI community the next week. Is there are any
tracking sece?
Speaker 5 (32:46):
I think that might be harder to pull off, but
I just find it. You know. Taylor has always views, Yeah,
I don't think that's happening, but Ken, but there's lots
of people who could fund you on one side or
the other. And Taylor has always emphasized this that like
it's really ironic that the tech investors have been so
late to appreciating the way that the economies of their
(33:10):
platforms work, and I just think is so interesting to
me that this is all tech money, right, But they
are like just discovering content creators in twenty twenty six.
I mean, I'll say, like, EA definitely want to AI
Safety was having a moment pre Sambankment Freed going to
jail where they thought maybe they could go mainstream. I
(33:30):
don't think they were like quite in content creators. But
I've been thinking about Taylor this whole time because she's
been making this argument for like a decade, like try
to understand your own platforms, guys, and now they finally are.
Speaker 4 (33:43):
One thing I'll say just about the flip flopping is
we have seen a lot of content creators change their
political positioning based on what's more profitable. And I actually
was just speaking to a manager yesterday of a big
progressive content creator, and he said he's been trying to
push him to adopt a slightly different political ideology online
(34:04):
because it's more profitable. And you know, we've seen political
evolutions like this is what people talk about of like
drifting to the right, especially promoting like right leaning or
sort of reactionary ideology pro corporation. You know, we have
influencers doing spaon con for natural gas companies like the
energy industry is also huge in the influencer marketing world.
(34:26):
Miranda Green, who is a phenomenal environmental journalist, has written
extensively about this. These companies pushing you know, like using
influencers to push big oil. You know, as Natasja mentioned,
the big tech industry has discovered it. So I feel
like all the worst people in the world over the
past couple years have finally discovered like, oh well, this
is just great. You know, we could just pay people
(34:47):
online to put out our messaging and this is how
Americans are getting their news now. We don't have to
deal with journalists at all.
Speaker 3 (34:53):
We have Rachel Jones was at the Anthropic Code conference
yesterday and she said there were it was just she's
writing about this, which I think will be out when
this when this podcast runs. How there's sort of this weird,
I don't know, tension I guess between like there's there's
these influencers walking around this developer conference and then there's
developers going like, wait a minute, I thought this was
(35:15):
for us, Like what are all these influencers doing here?
And it just creates this like really weird environment. I
think they've been going like really Edthropic's been going really
hard on the influencer market.
Speaker 2 (35:24):
Actually that's this last.
Speaker 1 (35:26):
Yeah, we're always ahead of it. We're always ahead of
the We always ahead of the cycle here on tech
stuff read You also had a news letter. You didn't
write the story. One of your colleagues did that. There
are some accusations going around that China is paying for
anti data center propaganda in the US. Is that is
that true? Or is it the rumor?
Speaker 2 (35:48):
Or is it?
Speaker 3 (35:49):
I mean, you know, I it sounds like it's true,
it doesn't. I mean, it's not very far fetched, right,
if you're China, like you're gonna do everything you can
to slow down American you know AI, right, and that's
a that's a tried and true way to do it.
Speaker 2 (36:03):
Right. The Russians figured this out, you know, a long
time ago.
Speaker 4 (36:07):
But I would say, or sorry to interrupt, but do
you think that's paid? Like from what I've seen from
most foreign influence campaigns, is it's not necessarily paid influencer content.
It's more like organic messaging. Well, they'll set up like
UGC accounts or AI accounts or things like that or
sort of like run ads on Facebook traditionally when they
could do that but more easily.
Speaker 3 (36:26):
Well, it might not be paid as in like here's
a here's cash to the influencer, but it costs them money, right,
they're putting they're putting money into the into the the
effort one way or another. So someone's getting paid somewhere
in the stack.
Speaker 4 (36:40):
Well, all the influencers can make money either way. So yeah, Well, in.
Speaker 3 (36:45):
China, when you go if you're a journalists and you
go to like any press event, there's just you sit
down at your table, and there's an envelope of cash.
Speaker 2 (36:51):
Like that's just how it works.
Speaker 3 (36:53):
I mean, China, that's not that much cash.
Speaker 4 (37:02):
Journal I live to try to just stay here and
just stay here and start promoting the you know, you
should just.
Speaker 2 (37:08):
Just become an AI researcher like me.
Speaker 4 (37:11):
This is the launch of Reid's influencer career.
Speaker 1 (37:14):
I know he's not only the tech founder but also
an influencer and tech Journalist's.
Speaker 2 (37:19):
Got many of my routing numbers.
Speaker 1 (37:20):
Exactly after the break, we're going to talk all about
data center backlash, what's behind it, and whether there are
any signs of its stopping.
Speaker 2 (37:29):
Stay with us, H.
Speaker 1 (37:42):
YH, welcome back. So I actually have a personal anecdote
about data center backlash. My mother recently moved to central
Florida and there's a data center being constructed down the
(38:07):
road by Larry Ellison called Project Tango, although actually it
was the plans were were halted by some local community activism.
And I found it pretty interesting to sort of live
in the midst of a kind of national story and
see a little bit firsthand, and also have my own
visceral emotional reaction, being like, I don't want a data
(38:28):
centered near my mum, and I think I think I'm
you know, I'm far from alone in that. Natasha, You've
been working on this store, You've been talking to some
of the sort of anti data center activists like what
is the what's going on here? Like, so, what's the
what are the parameters of this story?
Speaker 5 (38:45):
Right? So, I mean this is in the context of,
you know, some of this super pac reporting that I've
been doing. But there was just today, actually earlier today,
heat Map, which is an environmental a great environmental new site,
put out some new data that they have looked at
like very closely. These numbers haven't been reported before, and
(39:07):
they found that at least like in the first months,
first three months of this year, at least twenty proposed
data center projects were canceled after local pushback, and that
there are at least one hundred new data center fights
going on. And they have this chart where you can
just watch since Q two of last year, the number
(39:27):
of canceled projects just go up and up, and I think,
you know, it's it seems very similar to me, like
the dynamics that we talked about with with read story
about the executive Order. You know, I think that there,
I think that the industry thought that maybe they could
characterize some of the pushback as you know, initially like
(39:52):
the water concerns were unfounded, or you know, uh, this
is reactionary, that there are really rich nimbi's involved, and
so you know, to try to say, like who has
the authentic populist backlash, you know, or or they point
out that a lot of these people are are simultaneously
(40:12):
maybe signing up for chat GPT accounts, or you know,
you know that there's some irony there. But I think
that you know, just in the past week, we've seen
Kevin O'Leary's project in Utah get pushed through by by
the county commissioners in that case, you know, it's it's uh,
(40:32):
it was going to be run on one hundred percent,
you know, natural gas was going to use more than
the entire state. That's that was the proposal. And there
was this like viral video of people in like gathered
together in a gym just yelling like shame, shame, shame,
(40:52):
and you have you know, Kevin O'Leary, like you're talking
about Larry Ellison. And then there were two moratoriums, one
in Montgomery County, Maryland and Durham County, North Carolina. I
spoke to uh one of the residents of Apex Uh Apex,
North Carolina, which was the in the middle of this
(41:15):
mid term election that a lot of the super PACs
had put in a lot of money, was Valerie Fouchet,
the incumbent against this Democrat UH nedle Alam Progressive, and
he was saying, you know that in their local coalition
there are Republicans, there are Democrats, There are a lot
of you know, it's in the research triangles. So he
(41:38):
was like, there's a lot of you know, they thought
that they were going to get like rural farmers and
it's instead like environmental scientists from UNC and stuff, and
he was just like, if not for the super PACs,
like this would be a very bipartisan issue. And I
think one thing that's really been lost in in the
like you know, in the talking heads like Hunt picating
(42:00):
about it on X is the fact that when you
drill down into a lot of the deals, it's just
totally opaque for the residents. I don't know if your
mom found this, but it's like very hard.
Speaker 1 (42:11):
To tale for a long time.
Speaker 5 (42:15):
And yeah, yeah, it's like who is behind this? Can
I get any details? You know? In the Apex Center case,
they found out like three hours before you know that
there was going to be a group there was only
a mandate that they had to inform three different houses.
And so you're just seeing this I think, you know,
the balance of power is just so strongly in the
(42:35):
direction of the data center developers that yeah, it's just
it's just a really fascinating thing to be happening in
May on you know, May twenty twenty six, like with
all of these midterm elections coming up, and.
Speaker 1 (42:50):
What are the concer mean that obvious says that the
hum the noise, there's the pressure on the energy grid.
There's the particulate the kind of you know, pollution essentially,
like how how do we know like what is the
what are the negative effects on a local community from
a data center and what are the positive effects? Obviously
there's jobs and in some cases more than in others,
(43:12):
But what's the kind of what is the cost benefit
and us is if there's a data center popping up
next door.
Speaker 5 (43:18):
Well, I think part of the issue, the backlash is
that the normal procedures in terms of studying the environmental
impact are not being done. These things are being pushed
through very quickly. You know, there's federal legislation around like
pushing making permitting faster. So a they don't feel residents
don't feel like they're getting an adequate picture of what
(43:39):
the potential environmental impact will be. They don't feel like
they're getting you know, a clear picture of what's being built.
You know, they'll say like two stories, but it's actually
seventy feet tall, or in the case of the Kevin
O'Leary Shark Tank Utah one, they lowered the tax burden
(44:01):
from six percent to like zero point five percent. So
I think communities are very concerned about like what are
they getting in return, you know those jobs, data center jobs,
those are obviously temporary. And then yeah, what is the
environmental impact you hear when you hear that companies are
bringing in their own electricity, it's like diesel generators, and
(44:22):
then you know that we know the environmental impact there.
So I think it's just like a total you know,
you can only push people so far before they're going
to push back.
Speaker 3 (44:33):
Well, I mean, I just first of all, like, this
is total. This is like totally what we predicted would happen,
like you know, a year and a half ago or
two years ago, whenever we first wrote about it, and
that Rachel Jones was actually doing most of the reporting
on this, and I thought great. I mean that you
talk about the tech industry response to this, and I
think they were actually really late to understand that this
(44:56):
was going to happen because all this like opaque. You know,
we're going to come into this and we're not going
to say that we're AWS or Meta or whoever.
Speaker 2 (45:03):
You know, we're going to We're going to have like a.
Speaker 3 (45:05):
Shell company that wants to build this thing and do
it kind of in secret, Like that whole playbook was
in place, and it was just so clearly not going
to work in this new era of like massive scaling,
and so they're just bumping into these these completely predictable
problems building these data centers. But I think that what
I was going to say is like I think this
(45:26):
is going to lead to a lot of like just
interesting innovation. Like there's these parts of the rust belt
in the country where there's like this stranded power and
the companies are talking about, like you know in like Syracuse,
New York and places like that, where you can actually
create sort of almost like a like you have in
the Gulf States where people get like paid out for
(45:47):
the the energy generation like you.
Speaker 1 (45:50):
Know, geomal or access capacity from power plants.
Speaker 3 (45:54):
So access capacity like all these old power plants that
are like shut down or running it like not even
close to capacity because all the factories left, and now
they have this power. It's like, well, yeah, you can
come in and build these data centers. But I think
people are going to say, well, we want something out
of that, and what it might be is like a
check in the mail every month, you know. I think
that's the kind of stuff that we're talking about. These
(46:16):
companies are going to have to do and then I
think the.
Speaker 1 (46:18):
Other story in the journal about exactly that, right, which
is a proposed tax on data centers, and some of
the proposals are basically texting the data centers themselves. Other
proposal texting tokens, either the issue of tokens by tech
companies or the use of tokens by users. I mean,
it's kind of interesting question is.
Speaker 2 (46:35):
Where does that go? I mean, like I think you
might look at it.
Speaker 3 (46:38):
It might look like Alaska, you know, where people get
a check in the mail for the oil production, or
New Mexico where there's this like sovereign wealth fund with
from the oil money, things like that, but for communities.
So I think that's one thing, and then there's probably other,
probably many other innovative ideas. But then I think you're
also just going to see innovation on everything from energy
(46:59):
production into you know, the space data center concept, right
which which actually, like right in today as we speak,
makes no financial sense and is probably not feasible, but
like you know, any I mean, if it's possible, they're
going to do it because it's like they just need
to build these things. It's just there's so much demand.
(47:20):
So that would be great. Like I think that would
be a great outcome if, like all if things that
are super unprofitable today become profitable and it causes new
you know, new innovation, new science and stuff like that.
So I sort of look at it as a positive.
I guess that's a positive spin on this whole thing.
Speaker 4 (47:35):
I think, you know, one big issue here, and I
totally agree with Reid that there's going to be a
lot of innovation. But the average person in like Utah
or whatever small town you know is getting the data
center in Texas doesn't understand why we need data centers.
The tech industry has done such an incredibly poor job
of explaining, you know, how these things work, what they do.
And Natasha, I think mentioned earlier that a lot of
(47:57):
these activists are actually using chat GPT, like there is
a use case for data centers, Like we all use
these products, and I'm not just talking about AI products,
which are increasingly integrated into everyday life. The consumers obviously
get a lot of value from because we see the
adoption of you know, not just these chatbots, but lots
of AI tools. We also have the explosion of streaming,
you know, all of these other sort of online services
(48:18):
that rely on data centers, storage, et cetera. A lot
more people are going to be building their own technology,
running their own local programs, et cetera, with the rise
of vibe coding. So I think that, like, we need
a positive case for data centers, like we need the
tech industry needs to make that. I think the name
data center sounds incredibly dystopian. You know, these things just
(48:40):
were you know, they take up a huge amount of space,
Like it is very kind of scary. I wouldn't want
to live next to one. The environmental concerns are you know,
obviously real too. So I think that, like there just
needs to be a lot of education as well, and say,
like what are we using these data centers for? What
are these data centers for? You know, and explaining to
consumers how technology work and why we need more energy production,
(49:02):
why why it's a good thing to have technological development
and sort of expansion of access.
Speaker 3 (49:07):
To these But don't you think, Taylor, like even if
even if people understood that, there'd still be the same problem.
Like we everybody understands we need more housing, right because
of the housing crisis. Nobody wants an apartment complex built
next door to them, right, It's well, that's.
Speaker 4 (49:21):
Just listen, there's gonna be some like Ymbi's. I'm not
saying that, but I think that people will tolerate, uh,
you know, certain buildouts of technology. I mean we've seen
this with like even just like them. I mean, Amazon
warehouses provide a lot more sort of direct jobs. I
think if you made the case, I think that if
these data centers, as you mentioned, sort of like the
(49:43):
tax you know, taxing giving sort of residence credits, like,
I think if it was sort of more financially neutral
for them and more environmentally neutral, uh, and they sort
of understood the value of it, and there wasn't the again,
the concerns around the sound, I don't think it would
be as big of a deal. You know, a lot
of these are also not directly next to people's homes,
(50:05):
like you see people getting outraged about data centers or
in their states or you know states like Maine where
the residents came together to try to put this data
center moratorium when no one's even building data centers in Maine.
So I think it's more of this like emotional reaction.
Right now, I'm not saying that the people that will
be directly living next to the data centers won't sort
of have something to say. But I think most of
the opposition right now is actually not even coming from
(50:25):
people that are going to be directly living next to
these things.
Speaker 3 (50:28):
Yeah, it was interesting in Memphis because they they put
these turbines in without a permit, the XAI turbines, Yes,
and then they wanted to expand and build more turbines
and they wouldn't let them, so they just put them
across state lines in Mississippi and just like and just
built power lines. I mean, it's like fascinating the lengths.
Speaker 4 (50:47):
And those turbines for XAI are totally unpermitted. Like I
mean that Elon Musk is an example of somebody that
I would call an irresponsible actor in this space. And
you know, I agree with you read that, like hopefully
we'll see some times.
Speaker 1 (50:59):
They measured to description Taylor, but.
Speaker 2 (51:03):
Graphic is going to be using them so right, you.
Speaker 1 (51:06):
Know, Yeatasha, I want to get the last word to you.
You mentioned twenty data centers that are paused. Is is
that twenty out of forty or twenty out of twenty million?
Like how much will this activism slow down the construction
of data centers? Across the US and will it affect
the tech industry?
Speaker 5 (51:21):
Will it affect Yeah, it's actually already affecting data centers
that are not even tied to AI companies. And I
think that this is I mean, to me, this is
just like around like December of last year, you started
hearing like David Sacks and other people talk about how
the AI industry needed a new narrative. We really need
a new narrative, you know, like why why aren't people
(51:44):
excited the way they are about way moos? You know,
what's what's going on here? And they I mean, I
think it's very funny also to like be saying we
need a new narrative rather than just like coming up
with a new narrative and then and then seeding that
through influencers, which is how people do it. But they
I think, yeah, the positive use case for data centers
(52:05):
has not been made clear at all. You're even seeing
you know this one former Andrees and Horowitz investor who's
an investor in in anthropic His name is Midha. Yeah,
he yeah, he was just on a podcast talking about
and I heard him talking about this too, and he
(52:27):
was saying, Oh, maybe we'll need labels for data centers
that say, you know, here is one that is being
used to make semiconductors more efficient, here's one that's being
used to like, you know, innovate on bioscience, and then
people can decide whether or not they want to support that.
I mean, that is a degree of transparency that is
(52:48):
just so far from where we are right now. But
I just think, yeah, the the upside to residents has
not been made clear. The upside to the public has
not been made clear. So they're really in for a
very interesting, yeah, very interesting political battle. I think for
(53:09):
so long that rhetoric about you know, this existential risk
technology that's going to take all of our jobs, like
that works really well on the bosses, that works really
well on Wall Street shareholders, not so well with with voters.
And yeah, like here we are watching that that clash
(53:30):
play out.
Speaker 3 (53:31):
I just think it's that I think there's always going
to be a disconnect between the data centers and the
end product because and Anngenay knows this, he's a smart
guy that you know, you can't the same GPU could
be used to cure cancer and or video.
Speaker 1 (53:47):
Battles totally identification. I mean, it's not one to use
case as sort of upisode argument.
Speaker 3 (53:52):
Yeah, and it's it's general purpose, so it but that
also means the tokens are general purpose, and what we're
seeing is and this is like, I mean, we could
go into more detail on this, we don't have time,
but like this whole anthropic XC ideal just shows how
these tokens are just like a unit of They're almost
just like money, like currency at this point. And I
think residents, it's not I think the goal should not
(54:15):
be to get them to think of, oh, this data
center could cure cancer. It's like this data center is
printing money and we should get you just need to
get a cut of it. Like I think it's just
as simple as that.
Speaker 1 (54:26):
That's all we have time for today, but I do
want to end, as always with uh West Week in
Tech and best Week in Tech read I'm sorry I
scooped to a worst week in Tech with with U
stock coom. Can I go first, you guys first?
Speaker 4 (54:38):
Teddy, Yes, Stam Altman's having the worst week in tech.
Speaker 1 (54:43):
It's been true every week since we started this podcast.
Why specifically this week this.
Speaker 2 (54:49):
Week, why I was so excited about ass.
Speaker 4 (54:51):
Dot com I listen, I think we know again why
tech CEOs don't see each other. You know, there's this
Elon Musk Sam Altman lawsuit that's playing out right now
in the courts. And yesterday we got this release of
a trove of text messages between Sam Altman and Mira
Murradi discussing Sam's firing from open AI as CEO a
(55:13):
couple of years ago. I don't know if you remember
this drama. Sam was sort of temporarily fired CEO from
open Ay before coming back, and he just looks so
pitiful in these messages. He's like and and people have
been screenshotting. Now they becoming memes. He texts Mira while
miras in this meeting. She was on the board saying
can you indicate directionally good or bad? She says directionally
(55:34):
very bad. Okay, so I've seen people saying how are
things going? When my editor asked for my draft, I say,
directionally very bad. You know, become a joke.
Speaker 5 (55:44):
I actually thought that Sam was was redeemed somewhat by
those texts in that he said he was willing to
step aside. I mean, certainly he doesn't look like the
alpha like that's not the takeaway from the text, but
the fact that he was. I mean, it was just
like that felt. I think people felt a lot more
(56:05):
empathy for him seeing his position and Mira. You know,
it reflected very differently on Mirah. I think because she
had not made a lot of those things clear and
you had board members tested it was.
Speaker 1 (56:20):
It was.
Speaker 5 (56:21):
It was an interesting week for sure.
Speaker 3 (56:23):
I think it's funny you hear both of you just
criticizing these tech people for not being alpha enough.
Speaker 1 (56:30):
Not suspecially loops mixed. I saw maybe one hundred and
twenty thousand tech workers have been laid off so far
this year. Nothing nothing. I gave them the worst week
in tech this week. Is there any sign? Are they
just being fired and then immediately being rehired by another
(56:52):
tech company or what's going on? You know in the
in the Silicon Valley area, Well, how are people reacting
to this?
Speaker 5 (56:59):
I mean, I know a lot of unemployed people.
Speaker 2 (57:02):
I was gonna say the same thing.
Speaker 3 (57:04):
It's like, I've never had this experience living in the
Bay Area now for over ten years, where it's like
I know all these people who are getting laid off,
and you know, it's it's just it's just a new
it's a new thing. And I think it's mostly people
in management, I think, right, I mean that, do you
(57:24):
think Natasha like, I don't know.
Speaker 4 (57:26):
I don't know if you guys saw the message from
Brian Armstrong that he posted the co founder and CEO
at Coinbase. He posted that he had made the difficult
decision to reduce the size of coinbase by fourteen percent,
and he said it basically in this long, sort of
the Blackman style post, that that just average people are
(57:48):
like shipping production ready code using AI, which to me
is quite shopping shocking. I think a lot of people
were quote tweeting it like uh oh, like non technical
people are just shipping production code like that's concerning. But
I think it shows how quickly a lot of these
roles are being automated away and whether this code is good.
(58:08):
I think it makes me nervous about Coinbase. I would,
you know, but it's clearly you know they're able to
replace a lot of roles and cut out that like
middle manager level, the product manager level.
Speaker 1 (58:18):
Best we can take for me goes to There was
a sort of Willy Wonka's Chocolate factory thing which happened
in San Francisco this week, which I'm sure you will
know about. But two hundred lucky folks were selected out
of eight thousand applicants to attend the launch party of
GPT five point five. The lucky losers, all seven eight hundred,
(58:40):
got a huge increase in their token rate limits. So
congratulations to all eight thousand yuural winners, but the two
hundred you got to go. Do you hear anything about
this party?
Speaker 3 (58:50):
Rinatasha a lifetime supply of tokens. God, I didn't hear
about it. I didn't try to get my golden ticket.
I feel like Charlie Bucket here one one candy bar
or something.
Speaker 5 (59:02):
Yeah, Kylie Robinson from Core Memory went and she was
talking about I mean I saw some pictures with like
Sam Altman and his husband and there was like this
Samultman picture. She was talking about the cult of Sam Altman.
But like I was at the Anthropic, you know, with
Claude event yesterday, Yeah, I would I would say it's
(59:24):
definitely the year of the Cult of Dario over Samultman.
But yeah, I think people who got the Lucky ticket
were very stoked. And also there was free tokens or
at least like, you know, more compute capacity given out,
like Oprah after the SpaceX Anthropic announcement. So yeah, check
(59:45):
under your chairs.
Speaker 3 (59:47):
Who gets your best WEE can take I I thought
SpaceX I mean being it like the ability to offload
the tokens to Anthropic while they sort of figure out
their their whole like hard a strategy. Their answer to
Codex and cloud Cowork is like, like that's that's great,
Like going into the IPO gives them like a revenue stream,
(01:00:08):
and it also just like underscores this fascinating dynamic where
tokens are becoming like this transferable asset in you know,
it's just I've never seen anything like this and in
the economy before.
Speaker 5 (01:00:21):
And Compute brings everybody together.
Speaker 2 (01:00:26):
Until it kills us. All.
Speaker 1 (01:00:29):
Well, I know it's pretty a good place to end.
That's it for the weekend Tech. Thank you so much
for joining us that.
Speaker 4 (01:00:36):
This is the launch of your AI Doomer influencer Era Reid.
Speaker 3 (01:00:39):
I'm telling you I'm an AI researcher, so I have
to be a doomer.
Speaker 1 (01:00:42):
Now that's the rules for tech stuff. I must be
lost in. This episode was produced by e Lizah Dennis
and Men. It's a slow to executive produced by Me
Julian Nutster and Kate Osborne for the Kaleidoscope, and Katria
Novel for I Hop Podcasts. Paul Bowman engineered this episode,
Jack Intley mixed it, and Kyle mud upwards our theme
song special thank you to our three panelists today. Please
(01:01:04):
check out all of their work. We're lucky to call
them friends of Tech Stuff