August 19, 2026 • 30 mins
The AI conversation has become a battle between utopian promises and apocalyptic warnings. But what happens when we stop asking what AI will do to the world and start asking what it’s actually good for? Oz sits down with Josh Tyrangiel, journalist and author of AI for Good: How Real People Are Using Artificial Intelligence to Fix Things That Matter, to find out where AI is already making a difference.
Oz and Josh explore why successful implementation depends more on stubborn humans than flashy technology. They also unpack the growing backlash against AI, the uncertain impact on jobs, and why Josh thinks we still have a narrow window to decide what we actually want AI to become.
See omnystudio.com/listener for privacy information.
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:14):
Welcome to tech stuff. I'm as Voloscian if my LinkedIn
feed is to be believed. There are two kinds of
people in the world. Those who think AI is the
future of everything and will save us all, and those
who see it as a harbinger of the apocalypse that
will destroy society as we know it. But how are
real people actually using AI? What is it good for
(00:34):
when we think of it as a tool and not
some kind of demigod. Josh Sharangel is a journalist at
The Atlantic, a documentary producer, and the author of a
new book where he tries to answer that very question.
It's called AI for Good. How real people are using
artificial intelligence to fix things that matter? Josh, Welcome to
text Stuff.
Speaker 2 (00:53):
Thanks so much for having me.
Speaker 1 (00:54):
What made you want to write this book?
Speaker 2 (00:55):
I started doing ar reporting as a columnist for the
Wahian Post, right and so I was sort of immersed
really right after the debut of ch GPT three point five,
which is the moment most people think of as the
kind of a rival of modern AI. And I've found
myself entertained but not enlightened by the conversation, by which
(01:16):
I mean, you know, I spent a lot of time
in the valley, and people were telling me basically this
sort of like almost routine level of repetition about how
AI is going to make everything better and we're going
to cure cancer and we're going to mitigate climate change,
but we really have to go go, go to make
those things happen, right, and so on the one hand,
(01:37):
it's like, oh, it's very interesting. And then there was
this other side, as you mentioned in the intro, of
people who are like, you have no idea what is
about to come for us? Come with me if you
want to live like all the apocalyptic stuff. And as
a columnist when you're just trying to like get through
a couple of cycles, you're like, well, this is great.
You have these two very clear narrative sides. The stakes
(01:59):
are very high. You're talking to very smart people, by
the way, like these are uncommonly smart but also uncommonly
egocentric people who have.
Speaker 1 (02:06):
Been uncommonly rewarded for their activities, which is not the
best thing for your sense of a reality check.
Speaker 2 (02:12):
I guess, yeah. No, they're very richly rewarded for what
they say. They think quite highly of themselves, which is
again not to undercut the accomplishments or the intelligence. But
at some point you're like, Okay, I've heard enough of this.
I don't actually care. Like, you know, they start telling
you about their fundraising and it's like, do you know
how little I care about like your latest round? Like
(02:34):
I don't care at all. That's great for you, but
you got the wrong guy. And so at some point
I was just like, what is this good for? And
so I was having this conversation with a guy named
Danny Hillis, who for people who don't know, Danny is
the creator of cloud computing. And if you were to
make a sort of Mount Rushmore of computer science, he'd
definitely be on there, and he would be kind of
(02:56):
in the Teddy Roosevelt spot of just like broad shoulder
and bearded and fun. And you know, I said, like, Danny,
what the fuck is this good for? And he just
kind of laughed at me, and then he said, look,
you have to imagine the tech without the tech companies.
Speaker 1 (03:12):
I mentioned the tech without the tech companies?
Speaker 2 (03:14):
Was there? Yeah, And I'll be honest with you, I'm
a little embarrassed that it took that long for the
thought to occur to me. But it does say something
about the lockdown that these big companies have on the tech.
And what Danny really meant by that was, if you're
only listening to them, they have already predetermined the sales point,
the revenue they need, and the practization that is required.
(03:37):
But this is a wildly important technology. Go find people
who are using it and experimenting it in ways that
may not be immediately profitable, but will help us do
things that we care about in the world. And so
that really opened my eyes. And from there, you know,
I just started exploring largely around things that I cared about. Right,
So can AI be useful in government? I care a
(03:59):
lot about the future of the republic and the competence
of government services healthcare, both in providing healthcare but also
making the experience of healthcare better, education, personal connection. So
that was really the genesis for the book. And I
owe Danny, you know, a ton of credit for kind
of shaking me by the shoulders and just being like,
you got to get away from the big companies.
Speaker 1 (04:20):
Yeah, this was a debate you know, we had on
tech stuff recently, which is like, even if you accept
that that may be a financial bubble around AI right now,
and that financial bubble may crash. It seems like the
resids you will nonetheless be intensely valuable, unlike say Web
three or other kind of tech type bubbles.
Speaker 2 (04:41):
Yeah. I think that's a very important point of discernment,
which is that it is not going to go away.
Like I'm a true believer that this is the most
powerful software we've invented, which is not to say it's
instantly miraculous. It's just very powerful. I can understand why
people are like, oh, it's a bubble. Well, no, the
finances are well, technologies real.
Speaker 1 (05:01):
How have you found your own views evolving on this
through writing the book? Did you have a significant change
in going from thinking has had no value to being
very persuadive? Is value? And what was the kind of
well the turning points along the way of your Josh
journey on this topic.
Speaker 2 (05:17):
I mean, it's a great question. I started to see
patterns in the challenge of implementing AI that really opened
my eyes to kind of the pace of implementation, the
pace of change. And really the most interesting one of
those challenges is human beings. Like we kick up a
lot of friction, right and for good reason. So like
(05:37):
we you know, when you talk about AI, to change
government like man. Government is a political system, thousands and
thousands of inputs and small rules and regulations. It has
human beings who just want to work an eight hour day.
It has other human beings who have political imperatives. And
so what you see is like the tech doesn't just
arrive and rule itself out and all of a sudden
(05:58):
everything solved. Tech still needs human beings to implement it
and refine it and mess with it, and they're really
like in every aspect that I was dealing with, what
you would see is like, you know, the tech almost
arrives in this kind of heroic caravan of like we're here,
we're going to fix it, and then everybody's like, mm
hmmm are you because our problems are not the problems
(06:21):
you think we have, right, And so what I got
out of that is, first of all, like, okay, human beings,
we're not going anywhere anytime soon. And two, we are
actually really really good at diagnosing how we want technology
to work and hammering away at it until it works
the way we want. It just takes time.
Speaker 1 (06:41):
This is one of my big questions about the drone moment,
because obviously there's been so much investment in anderil and
healsing in Europe and then in Iran and Ukraine, Like
people are actually cobbling together drones and making them for
combat situation. So to your point about like the devil
is in the deployed and the environment of making stuff
(07:02):
when you needed to save your life maybe much more
fertile for innovation than the environment when you have tensive
billions of dollars of venture capital.
Speaker 2 (07:10):
Well, I think that's right. I would say that there's
one distinction with drones that's pretty important, which is what
does success look like? There's only two phenomenon. Did you
hit the enemy did you avoid the ally? That's what
drone success is. It's pretty binary. And so I actually
think that in a way, as shocking as impressive as
(07:30):
some of that technology is, it's actually quite simple from
a decision tree point view. It's very different when you're saying, Okay,
we're in a hospital and we're trying to create an
AI program that can predict accurately whether a patient is septic.
You're dealing with massive amounts of variation, which is not
to say there isn't variation in drone technology. You've got
(07:52):
to know the terrain, you got to know the weather,
you got to know lots of different stuff. But the
unpredictability that can be kicked up by human beings internal
organs is vastly larger, and you don't necessarily get this
binary outcome. Each one of these problems, which is what
I learned, has a very distinct, unique kind of application,
(08:15):
and you have to pay attention to all that. It's
not as simple as hey, here's GPT, go solve it,
or hey, Claude code will agentify your world. That's kind
of meaningless. The devil is in the details that a
patent shape. You're thinking, I mean, it's such a wild company.
I mean, and I think you know. I'll go back
(08:37):
to the beginning for listeners who may not know this,
but basically it was created by these two guys, Alex
Karp and Peter Tiel. And Peter Tiel most people have
heard of largely because he is very successful Silicon Valley entrepreneur,
early investor in Facebook. Also, his politics are libertarian to
very very hard right. He's an ally of Donald Trump.
(08:57):
Alex carp'si opposite. He's a half black caf jew New Yorker.
Socialists gave money to Hillary Clinton and Kamala Harris, and
the two of them met in law school and they
decided very early on that artificial intelligence and software could
order the world if properly deployed. And so they got
money originally from inq Tel, which is the investment arm
(09:19):
of the CIA, and you know Pollunteer. They named it
after the mysterious Stones and the Lord of the Rings,
and the company has this mystique. What's hilarious about it
is it does the most boring thing in all of technology.
Their chief architect was like, I don't know what to
tell you. We are the mole people of Silicon Valley.
We deal with plumbing. And that's basically what they do
(09:40):
is that they take data pipelines, which are very unglamorous
things to build and to maintain. They hook into large
data sets that are active. They improve the performance of
those data sets, and then they pipe them into a
useful interface so that you can play a system like
a video game, and so that system could be for example,
you know, think Lows and Home Depot used Palenteer to
(10:02):
manage your inventory. Cleveland, Clinton, which I've spent a lot
of time reporting in, uses Pollenteer to manage their hospital events,
so they know like when new patients coming in and
when they're going out, which is critical to getting even
the minuscule margins that a hospital system has. And so
on the one hand, like obviously they're now a very
politicized company, and on the other they build infrastructure so
(10:25):
that you can actually live in a sort of more
enlightened digital world. And I kept running across them in
various places because in a lot of ways, you know,
people would run into problems and pollunteer would be able
to bail them out, not because what they build is
so beautiful or so elegant, but because it's very useful.
It's very bare bones. They're often much cheaper than their
(10:48):
competitors for reasons that we can get into. So, yeah,
I have a much more nuanced view of that company
than I think a lot of people do. You know.
I've certainly people on Twitter have certainly come after me.
Speaker 1 (10:58):
We took a lot of criticisms, kind of holding up
as an example of a company that was making helpful
AI deployment.
Speaker 2 (11:05):
Yeah, and listen, I'm not suggesting that you should agree
with their decision making to help the Trump administration with ICE,
but they're very clear it's not like they're hiding from it.
Alex Krp is very open about the fact that he
believes if you work at Palenteer, you should want to
make the United States government have the best possible software.
(11:26):
And what I saw time and time again is like,
they make really good software and so you know what,
people can can misinterpret that in any way they want,
But like, that's what I found.
Speaker 1 (11:35):
Talk about the kind of human stories that you grounded
your book in. There was a school superintendent in Indiana
called Peggy Buffington who had kind of an interesting story.
Speaker 2 (11:45):
Yeah, I mean, look, we'll talk shop just for a second.
But if you are going to write a book about AI,
the very first consideration you have is like, well, how
is this not going to be obsolete? Before I start,
and one of the things I sort of thought about
is like, well, you know, I've done enough reporting to
know that this stuff doesn't roll itself out. The human
beings are really going to be critical, and also that
(12:06):
there is a type of human being who is successful
at using and implementing AI. They're just it is, and
they're usually stubborn. They're usually really wetted to a problem,
like they really care about it. Almost irrationally. I kept
running into people who are much more dedicated to solving
the problem in their own domain than they are to AI,
(12:26):
and they've been so frustrated by the inability to make
progress that they're like, well, what the hell is this?
Is this AI thing worth a shot? Many of them
have no technical backgrounds, but they are just wedded to
either like improving health care quality, improving the way students learn,
and they start dabbling. And so Peggy's a really good example.
Is Peggy Actually you know, she's in a town called Hobert,
(12:49):
small town, it's like thirty thousand people. She's the school superintendent,
and she is a force of nature. She's very small,
but you can tell like she's the celebrity of this city.
And so when AI came along, she got a call
from Salcon. Salcon was aware that this was a person
who with a sort of force to her, and Con
(13:12):
Academy had done this collaboration with Open AI in secret
to build an AI bot, basically a tutoring bot. And
so they call Peggy and Peggy's like, yeah, I'm in.
We'll test it here, will beat the crap out of it,
and we're going to involve students, We're gonna involve teachers.
We're gonna give you the unvarnished truth about this bot
(13:32):
so we can understand it. And so the reason she's
so important is that what she really understands is school
she's in Indiana. Indiana is a politicized educational environment in
which they want high quality education, good scores, and they
want it cheap. And so Peggy's like, well, let's try it.
And what I saw from her was this incredible dedication
(13:54):
to get the stuff in the right kids and the
right teacher's hands, and then lots of pfational development with
teachers teaching them how to use it. And she's just relentless.
And in the end, I mean, the kind of interesting
thing about the entire open ai Con Academy experiment is
like didn't really work much for students. You know, they
(14:14):
went to the trouble of engineering this incredible tutoring bod
and the younger students were just like kind of ignored it,
weren't really all that enthusiastic. The older students, oftentimes, you
were very angry about it because they thought it was
an attempt to get between the relationship between them and
their teachers. What it was able to do. And this
is what Peggy was specifically so good at was it
(14:37):
really helped teachers reinvent the way they interact with their students.
So they made a bunch of teaching tools in con Migo,
which is the name of the bod. And it turns out,
like a lot of teachers post COVID, we're in this
environment where they were doing the you know the thing
that you and I grew up with, which is they'd
stand in front of the class. They had a parcel
(14:58):
of information that over the or so forty minutes they
were going to teach the class. Eventually the students would
be graded on their ability to synthesize and repeat the information.
And the model is just dead. What the teachers told
me is like post COVID, like what are we even doing?
I know they're looking at their phones, I know they
have multiple tabs open, and I'm up there kind of
bombing basically, And so what they were able to do,
(15:21):
and I saw this in a couple instances, but there's
one teachers just great at it. She basically was like,
all right, I gotta change, and so she would use
the bot to input the lesson planned the curricular points
she needed to hit. And then she would say, help
me take this lecture and make it a lab. Help
me take this test and make it a lab. And
(15:42):
what I saw was that the kids basically were just active.
They were always moving, they were talking to each other,
they were talking to her. She could scaffold the learning
so that, you know, there's this phrase in teaching called
teaching to the middle, where you're basically trying to account
not for the top kids and not for the bottom kids,
but for this imagined median, so that as many kids
(16:04):
as possible. Following along which she discovered, this teacher and
many others was like, actually, you can teach to multiple
groups in differentiation if the BOT is telling them and
helping them understand when they need to call me to
come over and help. And so I saw a lot
of promise out of that. But it just never would
have happened without Peggy Buffington, who kept the Indiana Department
(16:27):
of Education at bay. Didn't tell them this was happening
because she knew they would squash it. Who was relentless
with teachers, who then would take the successful teachers and
make them talk to the unsuccessful teachers.
Speaker 1 (16:39):
This to your point is about making it technology useful
within a system rather than about technology.
Speaker 2 (16:44):
Yeah is it? Like, yes, the software is cool, but
change is hard in any realm, and if you want
the software to actually make the changes that we want
to see in the world, it's not going to happen
without multiple advocates who are stubborn and grumpy and willing
to get head in your face if you're standing in
a way. And so Pegy just stood out to me.
It's like, oh yeah, like this thing would be dead
(17:08):
as a doornail without her. And again, every place I went,
I would found some version of Peggy, and without it,
you just having an unsuccessful producting unsuccessful implementation.
Speaker 1 (17:39):
Now a quick break, switching topics to one of our
favorite sponsors, Vital Proteins. As a fan of science and technology,
I like knowing what I'm putting into my body and
why we've talked on tech stuff about AI, enhanced blueberries,
superhuman athletes, and health trackers. And while I might not
be joining the enhanced games anytime soon, I still want
(17:59):
to make sure I'm doing what's best for my body.
That's why I love Vital Proteins. I always know exactly
what I'm getting in their products like protein and collagen,
and what they're leaving out like sugar and artificial sweeteners.
I especially love the convenience of their new Collagen and
Protein shake. With Vital Proteins Collagen and Protein Shake, I
(18:20):
can grab and go without too much stress. And they
have both vanilla and chocolate, so I even get a
bit of variety. Not only that, with thirty grams of
protein including ten grams of collagen peptides and every shake,
I'm confident that I'm supporting my hair, skin, nails, bones,
and joints for years to come. But don't take my
(18:40):
word for it. Try the new Vital Proteins Collagen and
Protein Shake for yourself. Thank you again to our sponsor,
Vital Proteins. What do you make of these examples like
Peggy's example, which you're ultimately fairly heroic characters. How do
those kind of singular exsessles they ladder up to like
(19:01):
a new paradigm for how we may interact with AI
or how AI may diffuse through society? Like what do
these what do these examples coalesce into for you?
Speaker 2 (19:11):
Well, there's it's a good question. I mean, I think
there's a there's a couple of ways to approach it, right,
we only have a couple of years to figure out
what we want from AI before it gets kind of
too late, and companies and governments have dictated what we're
going to get from AI because.
Speaker 1 (19:25):
To your point, we're not able to interact with the
text ever from the tech companies anymore. It's kind of
locked in. It's captured.
Speaker 2 (19:30):
Yeah, And so a lot of the response to that
has been for people to say, no, I don't want
it to do this, and that's an important response, but
I think an equally important response is, but I would
like it to do that. And so part of what
they're doing is showing up and demonstrating that, like, actually,
it can improve certain aspects of our lives. It's not
(19:52):
gonna be easy, but with the right person and the
right funding and the right understanding, what we're trying to do,
like this is very useful stuff in lat of different realms.
And so I think there's a couple lessons from it. One,
as I said, is like why don't we show people
what we want? Why don't we demonstrate there's a market
for like, actually, good positive implementations that help our civilization advance?
(20:14):
And on the other you know, look, I do know
that there are lots of companies and lots of cities
and lots of organizations that are like, how are we
going to do this? And if their example is kind
of a roadmap, which I think it is, that's helpful too.
I've been a business journalist for a long time and
when I talk to CEOs, they don't really know much
about their software for the most part, Right.
Speaker 1 (20:37):
And the mandate is use AI is ensure this.
Speaker 2 (20:39):
Is use AI, right, And then most recently, the mandate
is stop using AI. It's too expensive. Right. And part
of this is like, actually, guys, it's not about the expense,
and it's not about this broad brush. It's about asking yourself, Hey,
is there a problem that we're really trying to solve here?
And if so, who's the right person to lead that?
(20:59):
So when I was at the Cleveland Clinic, they had
recently hired a new CTO chief technical officer who'd come
from outside healthcare. And what he realized is that there
are lots of problems in a hospital system today I
would be good for But he wasn't going to start there.
He was actually going to start by finding people in
the organization who he thought had the goods to be
(21:23):
essentially product managers, and so he would find people who
he saw were like super prustrated by a problem, and
then he would start to reverse engineer, like, oh, this
might be a good partner if I wanted to do X,
Y or Z, right, And I said, I was sitting
across from row Hit at lunch and he was telling
(21:44):
me all this just in the most like down fraud
and eyre kind of way, and I was like, Hit,
it's an eighty thousand person organization, Cleveland Clinic. How many
people have you found? He's like eight. Ten. It's like,
I don't know about the but eight or ten. Now
I think he was under selling a little bit. But
(22:06):
the point is taken, which is that this stuff is
actually really hard. You need this domain expertise, but when
you have it and the tech is calibrated to work,
you get pretty incredible results.
Speaker 1 (22:18):
The New Times reviewed your book and sort of pointed
to an irony about surfacing these optimistic case studies at
a time when quote companies are laying off tens of
thousands in the name of automation. How do you respond
to that.
Speaker 2 (22:31):
Well, I'm glad you asked. So. I actually wrote a
cover story for The Atlantic in February about AI and
the future of employment. Here's the deal. At this moment,
we have seen no impact. Like in the data, we
have yet to see a significant impact of any kind
suggesting that AI is compromising the future of employment.
Speaker 1 (22:54):
We have seen mass layoffs at tech companies with AI
as a proviso.
Speaker 2 (22:59):
Yeah, and by the way, like from a narrative perspective,
i'd love for that to be the case. It's much easier, right,
But I haven't seen it bear out. And a lot
of times when people were laying off engineers, what they
discovered is they actually needed those engineers. So how that
plays out is largely a function of speed. There will
probably be some job loss, like I have no doubt
about that. The question is does it happen in three years,
(23:22):
five years, or thirty years. If it happens in thirty years,
we'll be fine. The labor markets have a natural rate
of adjustment. They always have. General purpose technologies usually come in,
they eliminate some jobs, they create new categories, and the
economists are very sanguine about this. In my reporting, the
thing that I found most disturbing and most worthy of
(23:43):
alarm is that there's really no conversation about the what ifs,
and that companies are feeling like they have a financial issue,
as we discussed earlier, which is they've made all this
investment in AI and are not yet seeing massive productivity gains. Therefore,
if they're under pressure, they may have to cut jobs. Therefore,
what's going to happen to all those people? And politically,
(24:04):
there's almost no answer other than Gina Romondo, who's made
a couple of suggestions about ways to do job retraining,
like it's a completely ignored issue by the political class
that said, like to the Times critique, like, well, what
about the jobs. I'm not saying it's not significant, But
the premise of the book is not let me adjudicate
whether AI is going to be good or bad. Writ large,
(24:27):
I don't know. I don't know how anyone can know.
I think that what I'm trying to restore is some
agency to all of us who may feel a little
bit helpless about it. And it goes back to this
point that LIKE doesn't have to be what the tech
companies want it to be. If we can demonstrate what
we want from the technology, and we like what we
(24:48):
demand from it. That is a form of agency, So
is not using it for the things that are terrible? Right,
And like, you know this very well, and I'm guessing
your listeners do too, But like, the the best way
to cast your vote with the technology company is in
your user behavior. If you use Google a certain way,
Google knows and then moves in that direction. If we
(25:10):
use AI in ways that replace human labor, in ways
that contribute to the sort of slopification of information, like,
we're going to get more of it. If we use
AI in other ways, we'll get more of that. Like
we're not without power here. I don't think that power
is going to last forever. This stuff is going to
settle quick. And so part of my belief is, like
(25:33):
we do have an obligation to show and report on
what we want to see from them that that will
help other people see that too. So like, listen, I've
written book reviews too, Like I totally know you're kind
of like you got to make your points, but I
don't think that's what my premise was.
Speaker 1 (25:48):
Coming back to this idea that we don't have to
entirely align on how we perceive the tech with how
we perceive the tech companies. What is the road of
the tech companies in all of this, and how much
blame do you think they can for the anger you've described,
And what do you think the output of that anger
will it? Will it be the election of politicians who
essentially regulate this into as much of a moratorium as
(26:11):
can be, And what could be the harms of that?
I mean, why do you see this interaction between how
we live our daily lives, these trillion dollar IPOs and
the political winds blowing?
Speaker 2 (26:24):
Yeah? I mean, on the one hand, these companies deserve
credit for creating pretty incredible technology, and on the other
they deserve a ton of blame for our overall relationships
with technology. Right, So, I'm thinking specifically Google and Meta
and the ones that have been around all centuries selling
as social media or search in ways that have gradually
(26:45):
deteriorated our ability to trust information, our connection with other
human beings like we've seen it. I think it's not
unfair to say that there are similarities to tobacco companies.
We get it. If you're not anger, you're skeptical, you
haven't been paid. The thing that stands out to me
is that there's no ambiguity about what people want when
(27:06):
it comes to a regulatory regime from AI. So this
is a slightly older poll now, I think it's only
it's maybe two three months old. But Fox News polling,
which is different than Fox News, eighty one percent of
people said it was a very urgent or urgent priority
to regulate AI, that they wanted that from the government.
Problem is we have people in our elected offices who
(27:26):
buy and large don't know anything about technology. And there's
this gap that has existed forever, which is the people
who make technology are generally uninterested in public policy. The
people who make public policy are generally uninterested in technology. Right,
tech always moves faster than the law always does. And
in American particular, we're sort of famous for regulating after
(27:50):
the catastrophe. You know, Europe may regulate to prevent opportunity.
We regulate after the catastrophe. We only have one institutional
way to fix it, which is how we vote. And
so another reason that I'm so interested in talking about
this now is that we have a data election coming up.
We ought to demand that our elected representatives have a
(28:12):
point of view about what kind of regulation they want
of these companies, whether it's some sort of international collaboration,
like you know, International Atomic Energy Commission, which has done
pretty well at regulating nuclear proliferation, whether it's state by state,
whether it's data centers. But we're not having like a
mature conversation about what we want and what we don't
(28:35):
want from the tech. Usually we reserve that for our lawmakers.
I would hope that at town halls and in debates
that that is a pretty big issue. My hunches it
will be. And I'm generally not an optimist about this stuff,
but I think the data center issue, while a little
bit misplaced, it's the only thing people can point to
(28:55):
physically and say I hate this except for flow cameras,
except for flockcamp. It's right, it's kind of time. I
don't you know. I know that this will be a
massive issue in twenty twenty eight, no question, but it's
already massive. So like, could we get to a place
because in the end, I think the company CEOs, many
of whom I've spoken with, they may posture, they may bluff,
(29:18):
they may say they want this, they want that. Pretty
much all of them will tell me privately, like, yeah,
it be a real relief to have some regulations, we
stop having to talk about this, and we want the
right kind of regulation for us. Of course they do,
but I haven't met any of them who are like,
don't regulate it at all. They don't want to spend
time wasted here. But it's about this like gap we have,
(29:39):
and it's you know, my hope is we can make
progress on that in the midterms.
Speaker 1 (29:43):
The book is AI for Good, how real people are
using artificial intelligence to fix things that matter. Josh, thank
you my pleasure for tech Stuff. I'm os Voloshin. This
(30:08):
episode was produced by Eliza Dennis and Tyler Hill. It
was executive produced by me and Julian Nutter for Kaleidoscope
and Katria Novel for iHeart Podcasts. Jack Insley makes this
episode and Kyle Murdoch rotar THEMESI a special thank you
to all our listeners. Please rate, review, and reach out
to us at tech Stuff podcast at gmail dot com.
We love hearing from you.
Welcome to tech stuff. I'm as Voloscian if my LinkedIn
feed is to be believed. There are two kinds of
people in the world. Those who think AI is the
future of everything and will save us all, and those
who see it as a harbinger of the apocalypse that
will destroy society as we know it. But how are
real people actually using AI? What is it good for
(00:34):
when we think of it as a tool and not
some kind of demigod. Josh Sharangel is a journalist at
The Atlantic, a documentary producer, and the author of a
new book where he tries to answer that very question.
It's called AI for Good. How real people are using
artificial intelligence to fix things that matter? Josh, Welcome to
text Stuff.
Speaker 2 (00:53):
Thanks so much for having me.
Speaker 1 (00:54):
What made you want to write this book?
Speaker 2 (00:55):
I started doing ar reporting as a columnist for the
Wahian Post, right and so I was sort of immersed
really right after the debut of ch GPT three point five,
which is the moment most people think of as the
kind of a rival of modern AI. And I've found
myself entertained but not enlightened by the conversation, by which
(01:16):
I mean, you know, I spent a lot of time
in the valley, and people were telling me basically this
sort of like almost routine level of repetition about how
AI is going to make everything better and we're going
to cure cancer and we're going to mitigate climate change,
but we really have to go go, go to make
those things happen, right, and so on the one hand,
(01:37):
it's like, oh, it's very interesting. And then there was
this other side, as you mentioned in the intro, of
people who are like, you have no idea what is
about to come for us? Come with me if you
want to live like all the apocalyptic stuff. And as
a columnist when you're just trying to like get through
a couple of cycles, you're like, well, this is great.
You have these two very clear narrative sides. The stakes
(01:59):
are very high. You're talking to very smart people, by
the way, like these are uncommonly smart but also uncommonly
egocentric people who have.
Speaker 1 (02:06):
Been uncommonly rewarded for their activities, which is not the
best thing for your sense of a reality check.
Speaker 2 (02:12):
I guess, yeah. No, they're very richly rewarded for what
they say. They think quite highly of themselves, which is
again not to undercut the accomplishments or the intelligence. But
at some point you're like, Okay, I've heard enough of this.
I don't actually care. Like, you know, they start telling
you about their fundraising and it's like, do you know
how little I care about like your latest round? Like
(02:34):
I don't care at all. That's great for you, but
you got the wrong guy. And so at some point
I was just like, what is this good for? And
so I was having this conversation with a guy named
Danny Hillis, who for people who don't know, Danny is
the creator of cloud computing. And if you were to
make a sort of Mount Rushmore of computer science, he'd
definitely be on there, and he would be kind of
(02:56):
in the Teddy Roosevelt spot of just like broad shoulder
and bearded and fun. And you know, I said, like, Danny,
what the fuck is this good for? And he just
kind of laughed at me, and then he said, look,
you have to imagine the tech without the tech companies.
Speaker 1 (03:12):
I mentioned the tech without the tech companies?
Speaker 2 (03:14):
Was there? Yeah, And I'll be honest with you, I'm
a little embarrassed that it took that long for the
thought to occur to me. But it does say something
about the lockdown that these big companies have on the tech.
And what Danny really meant by that was, if you're
only listening to them, they have already predetermined the sales point,
the revenue they need, and the practization that is required.
(03:37):
But this is a wildly important technology. Go find people
who are using it and experimenting it in ways that
may not be immediately profitable, but will help us do
things that we care about in the world. And so
that really opened my eyes. And from there, you know,
I just started exploring largely around things that I cared about. Right,
So can AI be useful in government? I care a
(03:59):
lot about the future of the republic and the competence
of government services healthcare, both in providing healthcare but also
making the experience of healthcare better, education, personal connection. So
that was really the genesis for the book. And I
owe Danny, you know, a ton of credit for kind
of shaking me by the shoulders and just being like,
you got to get away from the big companies.
Speaker 1 (04:20):
Yeah, this was a debate you know, we had on
tech stuff recently, which is like, even if you accept
that that may be a financial bubble around AI right now,
and that financial bubble may crash. It seems like the
resids you will nonetheless be intensely valuable, unlike say Web
three or other kind of tech type bubbles.
Speaker 2 (04:41):
Yeah. I think that's a very important point of discernment,
which is that it is not going to go away.
Like I'm a true believer that this is the most
powerful software we've invented, which is not to say it's
instantly miraculous. It's just very powerful. I can understand why
people are like, oh, it's a bubble. Well, no, the
finances are well, technologies real.
Speaker 1 (05:01):
How have you found your own views evolving on this
through writing the book? Did you have a significant change
in going from thinking has had no value to being
very persuadive? Is value? And what was the kind of
well the turning points along the way of your Josh
journey on this topic.
Speaker 2 (05:17):
I mean, it's a great question. I started to see
patterns in the challenge of implementing AI that really opened
my eyes to kind of the pace of implementation, the
pace of change. And really the most interesting one of
those challenges is human beings. Like we kick up a
lot of friction, right and for good reason. So like
(05:37):
we you know, when you talk about AI, to change
government like man. Government is a political system, thousands and
thousands of inputs and small rules and regulations. It has
human beings who just want to work an eight hour day.
It has other human beings who have political imperatives. And
so what you see is like the tech doesn't just
arrive and rule itself out and all of a sudden
(05:58):
everything solved. Tech still needs human beings to implement it
and refine it and mess with it, and they're really
like in every aspect that I was dealing with, what
you would see is like, you know, the tech almost
arrives in this kind of heroic caravan of like we're here,
we're going to fix it, and then everybody's like, mm
hmmm are you because our problems are not the problems
(06:21):
you think we have, right, And so what I got
out of that is, first of all, like, okay, human beings,
we're not going anywhere anytime soon. And two, we are
actually really really good at diagnosing how we want technology
to work and hammering away at it until it works
the way we want. It just takes time.
Speaker 1 (06:41):
This is one of my big questions about the drone moment,
because obviously there's been so much investment in anderil and
healsing in Europe and then in Iran and Ukraine, Like
people are actually cobbling together drones and making them for
combat situation. So to your point about like the devil
is in the deployed and the environment of making stuff
(07:02):
when you needed to save your life maybe much more
fertile for innovation than the environment when you have tensive
billions of dollars of venture capital.
Speaker 2 (07:10):
Well, I think that's right. I would say that there's
one distinction with drones that's pretty important, which is what
does success look like? There's only two phenomenon. Did you
hit the enemy did you avoid the ally? That's what
drone success is. It's pretty binary. And so I actually
think that in a way, as shocking as impressive as
(07:30):
some of that technology is, it's actually quite simple from
a decision tree point view. It's very different when you're saying, Okay,
we're in a hospital and we're trying to create an
AI program that can predict accurately whether a patient is septic.
You're dealing with massive amounts of variation, which is not
to say there isn't variation in drone technology. You've got
(07:52):
to know the terrain, you got to know the weather,
you got to know lots of different stuff. But the
unpredictability that can be kicked up by human beings internal
organs is vastly larger, and you don't necessarily get this
binary outcome. Each one of these problems, which is what
I learned, has a very distinct, unique kind of application,
(08:15):
and you have to pay attention to all that. It's
not as simple as hey, here's GPT, go solve it,
or hey, Claude code will agentify your world. That's kind
of meaningless. The devil is in the details that a
patent shape. You're thinking, I mean, it's such a wild company.
I mean, and I think you know. I'll go back
(08:37):
to the beginning for listeners who may not know this,
but basically it was created by these two guys, Alex
Karp and Peter Tiel. And Peter Tiel most people have
heard of largely because he is very successful Silicon Valley entrepreneur,
early investor in Facebook. Also, his politics are libertarian to
very very hard right. He's an ally of Donald Trump.
(08:57):
Alex carp'si opposite. He's a half black caf jew New Yorker.
Socialists gave money to Hillary Clinton and Kamala Harris, and
the two of them met in law school and they
decided very early on that artificial intelligence and software could
order the world if properly deployed. And so they got
money originally from inq Tel, which is the investment arm
(09:19):
of the CIA, and you know Pollunteer. They named it
after the mysterious Stones and the Lord of the Rings,
and the company has this mystique. What's hilarious about it
is it does the most boring thing in all of technology.
Their chief architect was like, I don't know what to
tell you. We are the mole people of Silicon Valley.
We deal with plumbing. And that's basically what they do
(09:40):
is that they take data pipelines, which are very unglamorous
things to build and to maintain. They hook into large
data sets that are active. They improve the performance of
those data sets, and then they pipe them into a
useful interface so that you can play a system like
a video game, and so that system could be for example,
you know, think Lows and Home Depot used Palenteer to
(10:02):
manage your inventory. Cleveland, Clinton, which I've spent a lot
of time reporting in, uses Pollenteer to manage their hospital events,
so they know like when new patients coming in and
when they're going out, which is critical to getting even
the minuscule margins that a hospital system has. And so
on the one hand, like obviously they're now a very
politicized company, and on the other they build infrastructure so
(10:25):
that you can actually live in a sort of more
enlightened digital world. And I kept running across them in
various places because in a lot of ways, you know,
people would run into problems and pollunteer would be able
to bail them out, not because what they build is
so beautiful or so elegant, but because it's very useful.
It's very bare bones. They're often much cheaper than their
(10:48):
competitors for reasons that we can get into. So, yeah,
I have a much more nuanced view of that company
than I think a lot of people do. You know.
I've certainly people on Twitter have certainly come after me.
Speaker 1 (10:58):
We took a lot of criticisms, kind of holding up
as an example of a company that was making helpful
AI deployment.
Speaker 2 (11:05):
Yeah, and listen, I'm not suggesting that you should agree
with their decision making to help the Trump administration with ICE,
but they're very clear it's not like they're hiding from it.
Alex Krp is very open about the fact that he
believes if you work at Palenteer, you should want to
make the United States government have the best possible software.
(11:26):
And what I saw time and time again is like,
they make really good software and so you know what,
people can can misinterpret that in any way they want,
But like, that's what I found.
Speaker 1 (11:35):
Talk about the kind of human stories that you grounded
your book in. There was a school superintendent in Indiana
called Peggy Buffington who had kind of an interesting story.
Speaker 2 (11:45):
Yeah, I mean, look, we'll talk shop just for a second.
But if you are going to write a book about AI,
the very first consideration you have is like, well, how
is this not going to be obsolete? Before I start,
and one of the things I sort of thought about
is like, well, you know, I've done enough reporting to
know that this stuff doesn't roll itself out. The human
beings are really going to be critical, and also that
(12:06):
there is a type of human being who is successful
at using and implementing AI. They're just it is, and
they're usually stubborn. They're usually really wetted to a problem,
like they really care about it. Almost irrationally. I kept
running into people who are much more dedicated to solving
the problem in their own domain than they are to AI,
(12:26):
and they've been so frustrated by the inability to make
progress that they're like, well, what the hell is this?
Is this AI thing worth a shot? Many of them
have no technical backgrounds, but they are just wedded to
either like improving health care quality, improving the way students learn,
and they start dabbling. And so Peggy's a really good example.
Is Peggy Actually you know, she's in a town called Hobert,
(12:49):
small town, it's like thirty thousand people. She's the school superintendent,
and she is a force of nature. She's very small,
but you can tell like she's the celebrity of this city.
And so when AI came along, she got a call
from Salcon. Salcon was aware that this was a person
who with a sort of force to her, and Con
(13:12):
Academy had done this collaboration with Open AI in secret
to build an AI bot, basically a tutoring bot. And
so they call Peggy and Peggy's like, yeah, I'm in.
We'll test it here, will beat the crap out of it,
and we're going to involve students, We're gonna involve teachers.
We're gonna give you the unvarnished truth about this bot
(13:32):
so we can understand it. And so the reason she's
so important is that what she really understands is school
she's in Indiana. Indiana is a politicized educational environment in
which they want high quality education, good scores, and they
want it cheap. And so Peggy's like, well, let's try it.
And what I saw from her was this incredible dedication
(13:54):
to get the stuff in the right kids and the
right teacher's hands, and then lots of pfational development with
teachers teaching them how to use it. And she's just relentless.
And in the end, I mean, the kind of interesting
thing about the entire open ai Con Academy experiment is
like didn't really work much for students. You know, they
(14:14):
went to the trouble of engineering this incredible tutoring bod
and the younger students were just like kind of ignored it,
weren't really all that enthusiastic. The older students, oftentimes, you
were very angry about it because they thought it was
an attempt to get between the relationship between them and
their teachers. What it was able to do. And this
is what Peggy was specifically so good at was it
(14:37):
really helped teachers reinvent the way they interact with their students.
So they made a bunch of teaching tools in con Migo,
which is the name of the bod. And it turns out,
like a lot of teachers post COVID, we're in this
environment where they were doing the you know the thing
that you and I grew up with, which is they'd
stand in front of the class. They had a parcel
(14:58):
of information that over the or so forty minutes they
were going to teach the class. Eventually the students would
be graded on their ability to synthesize and repeat the information.
And the model is just dead. What the teachers told
me is like post COVID, like what are we even doing?
I know they're looking at their phones, I know they
have multiple tabs open, and I'm up there kind of
bombing basically, And so what they were able to do,
(15:21):
and I saw this in a couple instances, but there's
one teachers just great at it. She basically was like,
all right, I gotta change, and so she would use
the bot to input the lesson planned the curricular points
she needed to hit. And then she would say, help
me take this lecture and make it a lab. Help
me take this test and make it a lab. And
(15:42):
what I saw was that the kids basically were just active.
They were always moving, they were talking to each other,
they were talking to her. She could scaffold the learning
so that, you know, there's this phrase in teaching called
teaching to the middle, where you're basically trying to account
not for the top kids and not for the bottom kids,
but for this imagined median, so that as many kids
(16:04):
as possible. Following along which she discovered, this teacher and
many others was like, actually, you can teach to multiple
groups in differentiation if the BOT is telling them and
helping them understand when they need to call me to
come over and help. And so I saw a lot
of promise out of that. But it just never would
have happened without Peggy Buffington, who kept the Indiana Department
(16:27):
of Education at bay. Didn't tell them this was happening
because she knew they would squash it. Who was relentless
with teachers, who then would take the successful teachers and
make them talk to the unsuccessful teachers.
Speaker 1 (16:39):
This to your point is about making it technology useful
within a system rather than about technology.
Speaker 2 (16:44):
Yeah is it? Like, yes, the software is cool, but
change is hard in any realm, and if you want
the software to actually make the changes that we want
to see in the world, it's not going to happen
without multiple advocates who are stubborn and grumpy and willing
to get head in your face if you're standing in
a way. And so Pegy just stood out to me.
It's like, oh yeah, like this thing would be dead
(17:08):
as a doornail without her. And again, every place I went,
I would found some version of Peggy, and without it,
you just having an unsuccessful producting unsuccessful implementation.
Speaker 1 (17:39):
Now a quick break, switching topics to one of our
favorite sponsors, Vital Proteins. As a fan of science and technology,
I like knowing what I'm putting into my body and
why we've talked on tech stuff about AI, enhanced blueberries,
superhuman athletes, and health trackers. And while I might not
be joining the enhanced games anytime soon, I still want
(17:59):
to make sure I'm doing what's best for my body.
That's why I love Vital Proteins. I always know exactly
what I'm getting in their products like protein and collagen,
and what they're leaving out like sugar and artificial sweeteners.
I especially love the convenience of their new Collagen and
Protein shake. With Vital Proteins Collagen and Protein Shake, I
(18:20):
can grab and go without too much stress. And they
have both vanilla and chocolate, so I even get a
bit of variety. Not only that, with thirty grams of
protein including ten grams of collagen peptides and every shake,
I'm confident that I'm supporting my hair, skin, nails, bones,
and joints for years to come. But don't take my
(18:40):
word for it. Try the new Vital Proteins Collagen and
Protein Shake for yourself. Thank you again to our sponsor,
Vital Proteins. What do you make of these examples like
Peggy's example, which you're ultimately fairly heroic characters. How do
those kind of singular exsessles they ladder up to like
(19:01):
a new paradigm for how we may interact with AI
or how AI may diffuse through society? Like what do
these what do these examples coalesce into for you?
Speaker 2 (19:11):
Well, there's it's a good question. I mean, I think
there's a there's a couple of ways to approach it, right,
we only have a couple of years to figure out
what we want from AI before it gets kind of
too late, and companies and governments have dictated what we're
going to get from AI because.
Speaker 1 (19:25):
To your point, we're not able to interact with the
text ever from the tech companies anymore. It's kind of
locked in. It's captured.
Speaker 2 (19:30):
Yeah, And so a lot of the response to that
has been for people to say, no, I don't want
it to do this, and that's an important response, but
I think an equally important response is, but I would
like it to do that. And so part of what
they're doing is showing up and demonstrating that, like, actually,
it can improve certain aspects of our lives. It's not
(19:52):
gonna be easy, but with the right person and the
right funding and the right understanding, what we're trying to do,
like this is very useful stuff in lat of different realms.
And so I think there's a couple lessons from it. One,
as I said, is like why don't we show people
what we want? Why don't we demonstrate there's a market
for like, actually, good positive implementations that help our civilization advance?
(20:14):
And on the other you know, look, I do know
that there are lots of companies and lots of cities
and lots of organizations that are like, how are we
going to do this? And if their example is kind
of a roadmap, which I think it is, that's helpful too.
I've been a business journalist for a long time and
when I talk to CEOs, they don't really know much
about their software for the most part, Right.
Speaker 1 (20:37):
And the mandate is use AI is ensure this.
Speaker 2 (20:39):
Is use AI, right, And then most recently, the mandate
is stop using AI. It's too expensive. Right. And part
of this is like, actually, guys, it's not about the expense,
and it's not about this broad brush. It's about asking yourself, Hey,
is there a problem that we're really trying to solve here?
And if so, who's the right person to lead that?
(20:59):
So when I was at the Cleveland Clinic, they had
recently hired a new CTO chief technical officer who'd come
from outside healthcare. And what he realized is that there
are lots of problems in a hospital system today I
would be good for But he wasn't going to start there.
He was actually going to start by finding people in
the organization who he thought had the goods to be
(21:23):
essentially product managers, and so he would find people who
he saw were like super prustrated by a problem, and
then he would start to reverse engineer, like, oh, this
might be a good partner if I wanted to do X,
Y or Z, right, And I said, I was sitting
across from row Hit at lunch and he was telling
(21:44):
me all this just in the most like down fraud
and eyre kind of way, and I was like, Hit,
it's an eighty thousand person organization, Cleveland Clinic. How many
people have you found? He's like eight. Ten. It's like,
I don't know about the but eight or ten. Now
I think he was under selling a little bit. But
(22:06):
the point is taken, which is that this stuff is
actually really hard. You need this domain expertise, but when
you have it and the tech is calibrated to work,
you get pretty incredible results.
Speaker 1 (22:18):
The New Times reviewed your book and sort of pointed
to an irony about surfacing these optimistic case studies at
a time when quote companies are laying off tens of
thousands in the name of automation. How do you respond
to that.
Speaker 2 (22:31):
Well, I'm glad you asked. So. I actually wrote a
cover story for The Atlantic in February about AI and
the future of employment. Here's the deal. At this moment,
we have seen no impact. Like in the data, we
have yet to see a significant impact of any kind
suggesting that AI is compromising the future of employment.
Speaker 1 (22:54):
We have seen mass layoffs at tech companies with AI
as a proviso.
Speaker 2 (22:59):
Yeah, and by the way, like from a narrative perspective,
i'd love for that to be the case. It's much easier, right,
But I haven't seen it bear out. And a lot
of times when people were laying off engineers, what they
discovered is they actually needed those engineers. So how that
plays out is largely a function of speed. There will
probably be some job loss, like I have no doubt
about that. The question is does it happen in three years,
(23:22):
five years, or thirty years. If it happens in thirty years,
we'll be fine. The labor markets have a natural rate
of adjustment. They always have. General purpose technologies usually come in,
they eliminate some jobs, they create new categories, and the
economists are very sanguine about this. In my reporting, the
thing that I found most disturbing and most worthy of
(23:43):
alarm is that there's really no conversation about the what ifs,
and that companies are feeling like they have a financial issue,
as we discussed earlier, which is they've made all this
investment in AI and are not yet seeing massive productivity gains. Therefore,
if they're under pressure, they may have to cut jobs. Therefore,
what's going to happen to all those people? And politically,
(24:04):
there's almost no answer other than Gina Romondo, who's made
a couple of suggestions about ways to do job retraining,
like it's a completely ignored issue by the political class
that said, like to the Times critique, like, well, what
about the jobs. I'm not saying it's not significant, But
the premise of the book is not let me adjudicate
whether AI is going to be good or bad. Writ large,
(24:27):
I don't know. I don't know how anyone can know.
I think that what I'm trying to restore is some
agency to all of us who may feel a little
bit helpless about it. And it goes back to this
point that LIKE doesn't have to be what the tech
companies want it to be. If we can demonstrate what
we want from the technology, and we like what we
(24:48):
demand from it. That is a form of agency, So
is not using it for the things that are terrible? Right,
And like, you know this very well, and I'm guessing
your listeners do too, But like, the the best way
to cast your vote with the technology company is in
your user behavior. If you use Google a certain way,
Google knows and then moves in that direction. If we
(25:10):
use AI in ways that replace human labor, in ways
that contribute to the sort of slopification of information, like,
we're going to get more of it. If we use
AI in other ways, we'll get more of that. Like
we're not without power here. I don't think that power
is going to last forever. This stuff is going to
settle quick. And so part of my belief is, like
(25:33):
we do have an obligation to show and report on
what we want to see from them that that will
help other people see that too. So like, listen, I've
written book reviews too, Like I totally know you're kind
of like you got to make your points, but I
don't think that's what my premise was.
Speaker 1 (25:48):
Coming back to this idea that we don't have to
entirely align on how we perceive the tech with how
we perceive the tech companies. What is the road of
the tech companies in all of this, and how much
blame do you think they can for the anger you've described,
And what do you think the output of that anger
will it? Will it be the election of politicians who
essentially regulate this into as much of a moratorium as
(26:11):
can be, And what could be the harms of that?
I mean, why do you see this interaction between how
we live our daily lives, these trillion dollar IPOs and
the political winds blowing?
Speaker 2 (26:24):
Yeah? I mean, on the one hand, these companies deserve
credit for creating pretty incredible technology, and on the other
they deserve a ton of blame for our overall relationships
with technology. Right, So, I'm thinking specifically Google and Meta
and the ones that have been around all centuries selling
as social media or search in ways that have gradually
(26:45):
deteriorated our ability to trust information, our connection with other
human beings like we've seen it. I think it's not
unfair to say that there are similarities to tobacco companies.
We get it. If you're not anger, you're skeptical, you
haven't been paid. The thing that stands out to me
is that there's no ambiguity about what people want when
(27:06):
it comes to a regulatory regime from AI. So this
is a slightly older poll now, I think it's only
it's maybe two three months old. But Fox News polling,
which is different than Fox News, eighty one percent of
people said it was a very urgent or urgent priority
to regulate AI, that they wanted that from the government.
Problem is we have people in our elected offices who
(27:26):
buy and large don't know anything about technology. And there's
this gap that has existed forever, which is the people
who make technology are generally uninterested in public policy. The
people who make public policy are generally uninterested in technology. Right,
tech always moves faster than the law always does. And
in American particular, we're sort of famous for regulating after
(27:50):
the catastrophe. You know, Europe may regulate to prevent opportunity.
We regulate after the catastrophe. We only have one institutional
way to fix it, which is how we vote. And
so another reason that I'm so interested in talking about
this now is that we have a data election coming up.
We ought to demand that our elected representatives have a
(28:12):
point of view about what kind of regulation they want
of these companies, whether it's some sort of international collaboration,
like you know, International Atomic Energy Commission, which has done
pretty well at regulating nuclear proliferation, whether it's state by state,
whether it's data centers. But we're not having like a
mature conversation about what we want and what we don't
(28:35):
want from the tech. Usually we reserve that for our lawmakers.
I would hope that at town halls and in debates
that that is a pretty big issue. My hunches it
will be. And I'm generally not an optimist about this stuff,
but I think the data center issue, while a little
bit misplaced, it's the only thing people can point to
(28:55):
physically and say I hate this except for flow cameras,
except for flockcamp. It's right, it's kind of time. I
don't you know. I know that this will be a
massive issue in twenty twenty eight, no question, but it's
already massive. So like, could we get to a place
because in the end, I think the company CEOs, many
of whom I've spoken with, they may posture, they may bluff,
(29:18):
they may say they want this, they want that. Pretty
much all of them will tell me privately, like, yeah,
it be a real relief to have some regulations, we
stop having to talk about this, and we want the
right kind of regulation for us. Of course they do,
but I haven't met any of them who are like,
don't regulate it at all. They don't want to spend
time wasted here. But it's about this like gap we have,
(29:39):
and it's you know, my hope is we can make
progress on that in the midterms.
Speaker 1 (29:43):
The book is AI for Good, how real people are
using artificial intelligence to fix things that matter. Josh, thank
you my pleasure for tech Stuff. I'm os Voloshin. This
(30:08):
episode was produced by Eliza Dennis and Tyler Hill. It
was executive produced by me and Julian Nutter for Kaleidoscope
and Katria Novel for iHeart Podcasts. Jack Insley makes this
episode and Kyle Murdoch rotar THEMESI a special thank you
to all our listeners. Please rate, review, and reach out
to us at tech Stuff podcast at gmail dot com.
We love hearing from you.
TechStuff News
Hosts And Creators
Oz Woloshyn
Karah Preiss
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
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.