January 25, 2023 • 60 mins
With all the news about art programs like Midjourney, chat applications like ChatGPT and so on, it's no wonder more and more people are concerned 'AI' -- however defined, may put their jobs in danger. But how true is that? Is 'AI' coming for you? Tune in to learn more in the first part of this ongoing series.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
From UFOs to psychic powers and government conspiracies. History is
riddled with unexplained events. You can turn back now or
learn the stuff they don't want you to know. A
production of My Heart Grading. Hello, welcome back to the show.
(00:26):
I'm still Matt. It is I know, still me. They
called me Ben. We're joined as always with our superproducer.
All mission control decands, most importantly, you are you. You
are here, and that makes this the stuff they don't
want you to know. If you saw the title of
today's episode, folks, then you realized it's a question, is
(00:49):
AI coming for you? Uh? This may be the first
of several conversations and explorations we have about um, the
possible corruption and conspiracy's involved here, the uh, the bleeding
edge of technology, and what is often called an unprecedented time.
But spoiler, we're gonna find that people are throwing the
(01:10):
word unprecedented around a bit loosely. We've all, you know,
we've heard about a I've heard about machine learning a
lot in recent years, and I think for the five
of us growing up Paul, Matt, noell, Uh, you and
yours truly here, I think we always hear about this
(01:30):
coupled with fears of humans being replaced by these new
entities or even just processes. And the more it's weird
because there's this big divide in futurism right all the time,
the ted talkie dudes are always saying, hey, there's a
world where everybody wins. And then the more cynical folks,
often sci fi writers, imagine a dystopia. They say, we
(01:54):
you're gonna live in a post worker but not post
work economy. You'll still need money to survive. We have
fewer and fewer ways to make that money. It's funny.
It reminds me of like, you know, things like the Luddites,
you know, who were sabotaging um, you know, machines during
the Industrial Revolution, Because every time a new technology like
(02:14):
this emerges, there's always a faction or a contingent of
folks that that have that perspective like, oh, it's going
to replace the workers, just gonna replace this, gonna replace that,
and there's a backlash. But the reality is, short of
destroying it all, uh, you really can't put ideas like
that back into the bottle once they've kind of been
fleshed out, you know, and made real. So usually what happens.
(02:35):
Like even the synthesizer. You know, people were like, oh,
it's gonna replace the orchestra. That didn't happen. The synthesizer
became a tool unto itself that sounds like what it
sounds like, but no one's kind of illusions that it's
replacing an orchestra. I kind of feel that way about AI.
It's a little different because it's just so fast. Technology
moves so quickly these days, there is concern. But I
(02:57):
also think there is a world where AI many forms,
just becomes another tool and doesn't necessarily replace people. Yeah. Yeah,
some people replaces some jobs and some processes, no question.
I think we're in the stage where AI is a
tool being utilized by you know, all kinds of different
sectors within the economy. I'm really interested today, guys, in
(03:20):
exploring the kind of two sides to the view of
the future, right, the dystopian one we're talking about where
workers get displaced, everybody lose their jobs, really hard to
find work, and then the other one where that same
kind of thing happens, but because there's so much wealth
just kind of being generated by these programs, by these
(03:41):
other intelligences that humans don't have to work, and it's
okay because there are institutions that are just providing things
like a livable wage right or it's a name for that, right, Ben,
it's been whether the universal income or whatever, like that's
a it's a model that, yeah, universal basic income. And
that w you're talking about is kind of the the
(04:02):
economy of Star Trek people. People are prizing knowledge for
knowledge sake, right for the betterment of all living things,
because they don't have to work, you know, sixty hours
a week, because they don't have to make hard decisions
like can we get a new roof for our house
(04:22):
or are we going to be able to send our
kid to an increasingly expensive college system. In today's episode,
we're going to explore all of this through the lens
of the fact and fiction surrounding all these terrifying claims
and all these anodyne claims about the nature of the
future and the future is now. Is AI coming for you?
(04:45):
Here are the facts. First off, people still don't agree
on whether or not we should be using the term AI.
It's short for artificial intelligence. But what exactly makes it artificial, right,
is that gonna win. General AI comes about is that
going to be considered on the level of a racial epithet,
(05:05):
you know what I mean, Like, is it going to
be fighting words to tell strong AI or artificial generative
intelligence that it's artificial. Also, I think there's a bit
of a misconstruing of terms sometimes where some of this
stuff is actually more machine learning than it is AI
or artificial intelligence. And I'm not an expert. I could,
(05:28):
I would be hard pressed to tell you the exact differences,
but there are people that will argue and die on
that hill and they're very different things. We'll tell you
my concept, and maybe I'm completely wrong. My understanding is
that none of this is actual AI yet, that that
term is a stand in because it helps us as
consumers and you know, the people who are actually gonna
be using the tools understand what it is or what
(05:49):
it will be. But yeah, machine learning is the thing
that's functioning within all these things called AI right now,
right we're talking about processes. Really, Uh, The idea like
AI is on the on the verge of becoming a
thought terminating cliche. The idea, like the sci fi holy
(06:10):
grail of of AI, is a type of programming that
can replicate the cognitive abilities of the human brain entire, right,
And what we're seeing right now in general is stuff
that's very good at making more of an established thing
or interpreting that established thing, not so much creating a
(06:35):
brand new thing. And this is this is like, um,
this is an ancient idea. The humanity as as a
fad has always been super into this idea of themselves. Right,
let's build a thing that's like us. Guys, we can
make anything. Let's make ourselves classic human on multiple levels,
(06:56):
and quite recently, only quite recently, a senses in technology
have made that all that science fiction non fictional. Right,
that's the root of good science fiction. And the weird
thing is we we've got a level set. And think
about this fact. People have been dreaming of something like
this far before they ever dreamed of computers. And in
(07:18):
a sense, it is fair to say computers are just
a convenient means to that ancient and a step toward
that greater goal. If there was another invention that could
uh also get closer to whatever we mean when we
say AI, than people would be super obsessed with that.
We're we're talking about group psychology here over millennia and
(07:42):
it's cool, you know, it's neat, it's terrifying. It gets
discussed in all these think pieces. If you're a long
time conspiracy realist, you've heard us reference them, you've probably
uh hopefully dove into a few of those yourselves. Ted
talks lots of great scholarship, but we're not talking about
the realm of the abstract and philosophical today. We're talking
(08:03):
about reality. Chat bots are evolving using that word on
purpose like never before. Right, We've been having a lot
of interesting conversations about chat GPT in particular, but that
is just the beginning of a new era, you know,
and there was plenty of stuff before that. Well, and
(08:24):
it's weird to think about chat GPT as a chat
pot because it appears to be a tool that is
so much more robust, right when you're talking about people
using it to write code. I think that's you know,
that's going to be a huge part of our discussion,
just about the um expansive applications of something that emerges
(08:47):
like this, and that's where the fear comes. I think
that's where the fear comes out of the job replacement,
you know, UM anxiety. It's just it's it's pulling directly
from that Oh wow, this thing can do a lot
of right. Yeah, is where you see things like mid Journey,
Dolly Stable Diffusion. They're redefining the experience of creating art
(09:10):
to the point where, uh, it can be very difficult
for someone just looking at it to differentiate between something
an actual like the human artist made, and something made
quote unquote in the style of And this is bringing
up really difficult questions about the nature of creation, the
nature of collaboration and ownership and getting images is being
(09:33):
a little hypocritical, We're just gonna point that out. Uh.
But the big thing is we're on the cusp. Is
this the cusp of a chasm into which civilization falls
or is this more like the precipice of a runway? Right?
Are we on a plane or are we on a
weird Thelma and Louise, uh Cliff run right with with
(09:56):
a I. Uh. So we're dealing with the facts today.
What is future of this? This ni instantaneous creation? Right?
Will these lines of code learn to create code themselves?
That's kind of the goal, right, because if you think
about your brain, that's what your brain is sort of doing.
Your brain, your personality functions kind of like these lines
(10:17):
of code and you create new things. So based on
the stimuli you encounter, which is why you know, people
might fall really in love with something and then change
their personality around it. Right. Shout out to every middle schooler.
I know there's a band you heard and you're like,
this is me now and chat GBT is writing code, right, Um, yes,
(10:40):
spoiler it is it. Well, it is. It is writing code.
It is writing things. You could definitely say it's writing things.
But then the question is, what about the global elites
of the world who already have their hands on the
levers of power. Are they already planning for a new
(11:02):
world of robot workers with the rest of humanity tossed
on the figurative trash sheet, with all the other outdated
technology like the Sega Genesis. I was just thinking about
the Sega Genesis a while back. I was a second kid. Yeah,
I didn't get into Nintendo until I was an adult. Uh,
Sonic was my guy way more than Mario. Guys, does
(11:24):
this in any way figure into the concept of like
a singularity? Like is that sort of what we're theoretically
heading for? And that's kind of the moment when AI
kind of figure stuff out for itself and then we
really can't put that genie back in the bottle. Like obviously,
science fiction has shown as dystopian versions of that, with
like sky Neet becoming sentient and creating robots that want
(11:46):
to wipe out humanity and all of that stuff. But
this is sort of roughly clustered around the idea of
a singularity. Yeah, this is related for sure, the singularity
in the world of technology and the future. Apology is
this this hypothetical thought experiment point in time where technological
(12:08):
growth is on such a j curve, so it becomes
so exponential that it's uncontrollable and it starts to change
human civilization in ways that are both unpredictable and irreversible.
So it's pretty heady, heady stuff, you know. And this
is this all factors into that question we opened up
(12:28):
with at the top, is AI coming for you? Will
take a break for a word from our robo sponsors,
and we'll be right back, will hopefully still be us.
Here's where it gets crazy. Yes, AI is coming for you.
(12:53):
At least it's weird. Found a really cool gallop pole
from If you ask the majority of people in the
United States, there's a strange disconnect. Seventy three percent of
US residents believe a I will specifically eliminate more jobs
than it creates, but those same people only twenty three
(13:14):
percent of them believe the trend might affect them personally.
So in other words, yeah, people are saying, well, things
are gonna be bad for some people. I don't know,
maybe a lot not me though, because I uh make
avocado toast, you know, or I um kick puppies for
a living, and that's just a job that needs a
lot of humanity. I suspect those numbers would be different
(13:36):
if that poll was given today, because I mean, again,
with the exponential growth and improvement of this tech, A
year is a long time, you know, so between three
I met that it's probably closer to around you now.
Just guessing. There's a really cool in fiction response to
(13:59):
this as well. There's a really cool idea of slipping time,
right uh and progress over time in the Marvel universe.
In X Men comics, there are these things like, um,
you know how Wolverine is Weapon X. That's not a spoiler,
and that's like weapon ten because there were other iterations
(14:19):
of him in these government programs. There's a place in
this fictional universe called the world. Not the most creative,
but whatever, it's an awesome concept. And time moves more
quickly in the world in this little thing, and so
one year out here is thousands and thousands of years
(14:40):
and there because people are trying to drive evolution, just
push the ball faster. And you know, make no mistake,
if that were possible today, people would do it. And
it's kind of already happening to your point, Noll, in
certain in certain areas of technology, you know, and also
in the climate. But but if we look at this trend,
(15:04):
we also see that education makes a difference. The more uh,
the higher let, the higher your level of formal education is,
the less likely you are to be worried that this
trend will threaten you directly. You might say, well, you know,
I am a philosopher, see my scoff see my tweed patches.
(15:29):
The I I simply shan'n't be replaced. I shan'n't. Oh,
shan't you. I don't know if that's the right. But
then also posits that the world continues to value what
your stock in trade is, you know what I mean,
And on a on a broad enough scale, that it
actually matters. I mean, you know, philosophy is a bit
of a niche discipline. Let's just be honest. You know,
(15:52):
and your average um, you know consumer probably not super
concerned with philosophy. Um. You know, it is more something
that exists in academic realms. Uh. So you know, if
if academics, if institutions of you know, academic learning cease
to be as important, then so may that discipline cease
(16:13):
to be as important? Yeah, not a lot of what
would need you do? Arm bands on the kids? You know, well,
if they are there, it's it's super ironical. Right, everything's
ironic now or not. We're in the post irony society. Uh,
civilization is just collectively over it. Uh. We're making a
(16:34):
solid point here about the unevenness of how AI may
affect various industries. You can find all these again think
pieces perspectives, let's call them that in case think pieces
sounds dismissive. You can find a lot of perspectives that
are valid, that are listing jobs that are in danger,
(16:56):
jobs that are not in danger. And then you can
find people saying, hey, up being chicken little about it.
The sky is not falling. No jobs are really in danger,
at least not yet, so maybe we can think through these. Um.
We mentioned it off air. This just happened as we're
finishing up research for today's episode. As we record today, Friday, January.
(17:18):
It's UM, it's a little bit before noon here on
the East coast of the United States. This morning, Google
published one a hell of a memo. We'll just give
you the direct quote here. It's about their parent company, Alphabet.
Google is parent Alphabet Incorporated, is cutting about twelve thousand jobs,
(17:39):
or six percent of its workforce as the technology sector
reels from layoffs and companies stake their futures on artificial intelligence. Weird, right,
And and Matt, you were pointing out that this is
on the heels of a string of layoffs, Uh, something
like almost a hundred thousand across the tech industry. A
(18:00):
little over ninety thousand jobs within the tech sector. We're
lost last year last year alone. And and these happen
for any number of reasons. It's not all due entirely
to concerns about AI. But Google is interesting or Alphabet,
I guess we have to pretend they're different. Um, Alphabet
(18:21):
is interesting because they directly tie this calculation to artificial intelligence,
and we know that there are a Baker's dozen of
unpredictable variables that could affect these other companies, like economic predictions,
unhinged demands of shareholders, expecting year over your profits, etcetera, etcetera.
(18:45):
But adding to the Google thing, and I snooped around
a bit on this um with you know, just a
little bit of time, insiders are predicting that Google is
going to come out with something big in the AI
field this spring. And so that coupled with this with
(19:07):
thousands of people laid being laid off, has folks scratching
their heads, or humans, folks who have human heads scratching
those heads, and that got kind of dark. Um, don't
scratch other people's heads, scratch your own unless there's unless
there's consent. Yeah, so there's okay. One of the things,
(19:29):
like going back to the idea of the Ivory Tower
feeling safe, right, or the creative arts feeling safe, Let's
talk about the people who actually build all the things
society relies on it, right. We're talking truckers, We're talking
construction workers. We're talking people laboring in factories. We're talking
(19:51):
you know, we're talking to the enormously dangerous job mining
precious metals, right, which are a huge part of the
reason we're able to make this show and people are
able to hear it. A lot of people say that
is gonna be a thing of the past. I don't
know if you guys saw the Boston Dynamics video that
(20:11):
came out kind of recently. It's it's it's not on
a live construction site. It's a guy who is on
a set that gives you, you know, construction equipment, and
uh and I like this guy. I gotta tell you,
he's got It's charismatic in the way acts, and you
can tell he's having fun. But he's up on this,
(20:32):
you know, like this fake scaffolding. He's got his mallet.
He's got his mallet, and he's like he's like, oh jeez,
I forgot my tools and I'm making it's how like
uh Rick and Morty in some sort of pornographic bit.
But but he the ideas he forgets his tools and
(20:55):
his buddy Atlas the robot, here's here's this call and
jumps into action. Do you wanna should we describe what
does Yeah, rather than the construction worker, you know, hopping
down for a couple of seconds and picking up his tools.
A robot or whatever we want to call this thing,
(21:18):
almost in a hendroid, little very ninja like creature, hops
up and goes, oh, I'll get it for you, and
places a board down so that he can get onto
the scaffolding, kind of skips on over to the bag,
picks it up, and then skips on up to the
guy and tosses it to him, like literally tosses the
(21:39):
bag of tools to our construction worker and then hops
down and does a nice little flip off the scaffolding
and like tea, like, what was did it like knock
the construction worker over? Was it a heavy bag of
tools that should maybe not be tossed? This song a
(22:00):
little sketchy. I think it's a prop bag. This was like, okay,
this is like a demo. This is like not not
in the field. Yeah. And the seconds for the bag
to get to the construction workers so not not too bad,
just not too bad at all. Yeah. No, that's a
(22:20):
really good metric to add there, because what it looks
like is this um as, this automaton is solving problems,
navigating obstacles on the fly. It's uh, it's doing the
bare grills right, improvised adapt overcome through the tools thirty
two seconds. But if you look closer, what you see
(22:41):
is there's another video that shows how the robot navigates
space and and our pow um Damian Patrick Williams pointed
this out. They're being very careful with the language. They're
not saying Atlas does this. They're saying, we know they're
using the Royal we a lot, right. So the question
(23:04):
is how much are they steering this? How much of
this is um kind of a pre programmed set of
actions and processes. The problem is that right now bots
are not at this point able to move and adapt
as quickly as a human. It's still amazing, it is astonishing,
(23:25):
and eventually, if this research continues, they will, like these companies,
will achieve that goal. We just can't say it's here yet.
Does that make sense? Absolutely? And if you watch like
you said, Ben, if you watch that video closely, that
that machine is not just taking in all of its
surroundings and understanding where the person is, how the scaffolding
(23:49):
is built, how to get up to their where the
tool kit is. You can tell that some of that
is pre programmed, especially at the end when it knocks
over the box in order to get down and do
a flip off of the thing. All of that is
pre programmed, Like I have zero doubt, including the setup
of the scaffolding. I think that space even the way
(24:10):
like the space is part of the calculation. It's not
like it's just taking in data on the fly and
like adapting this is basically almost like a performance. No,
we think about what we just learned. We just learned
this past week about Tesla's original um self driving car
video test. Yeah, where it just came out that oh no,
(24:33):
it was actually there was quite a bit of driver
intervention in order to make that happen, even though the
claim was by Elon Musk himself and the company that
it was that full process was self driving, including parking
when they crashed a car. Oh you know. And again
that's not we're not casting unfair aspersion on places like
(24:54):
Boston Dynamics. No, it's it's yeah, they're showing off what
it can do. Even says it's showing how Atlas interacts
with objects and modifies the course to reach its goal.
The course yeah see uh, And and they're very they're
transparent about it, especially in that second video about how
it navigates space. But maybe maybe because there's such a
(25:18):
fundamental fear and concern about the potential here. Uh, people
get a little carried away when they describe it, right,
and you think of terminator, and you think a sky
net as as mentioned earlier, But there are countless related
jobs that could ultimately have that John Henry moment, you know, like,
uh again, would it be a terrible world if people
(25:40):
who labor in minds, sometimes against their will, were replaced
by robots? Some people, including some miners, would say, yes,
it is terrible because that's my only source of income, right,
and until you can solve the income problem, solving the
robot problem doesn't really do much for me, you know.
(26:01):
And that's a valid point, and it's a point that
a lot of people don't like to hear because it
forces you to acknowledge some uncomfortable realities and some uncomfortable
potentials for the solution there, because the solution could be, oh, well,
we just need many many fewer humans because we won't
need as much income if we have many many fewer
(26:23):
humans or a world like bloom Camps, uh Elysium, you know,
where uh, the masses of humanity are living on a
dying world. Their unemployment is rampant. Law enforcement has been
replaced by incredibly brutal robots, and the historical global elite
(26:47):
who don't do anything for society, live out in space
and have all the cures to all the ills and
just don't want to do it. He's just doing it,
you know. They check the schedule, they're they're busy. But
aside from that, like we see their very real things.
These are based on. Maybe we can talk about the
jobs that are supposedly not in danger because they say
(27:09):
that earlier point about the teen poll, that kind of things.
Things are probably shifting now, right, Uh. A few years
back se or so Americans said this stuff won't touch me,
and we were looking into jobs that supposedly won't be affected.
I wanted to start this one off with a dark horse.
(27:30):
You're ready. Eight HR managers apparently HR managers apparently on
the list of jobs that probably won't be automated. Well
that's good. I that's a positive. Um, Oh, come on, HR,
it doesn't listen to this show. You don't have to
(27:50):
do that. No, I just mean I know a lot
of really nice HR people. They just have to deal
with with weird. It's a weird thing to do, right,
are I have to navigate that limital space between the
corporate machine and the individual meat body that you know
needs things like healthcare. Yeah, And and that's not the
(28:11):
same thing as like an HR representative being automated, because
that's again that's answering a lot of similar questions right
in a programmaticum predetermined way. It's similar to the way
that if you are in an online customer service situation,
you have the option to chat and it'll ask you
(28:32):
some questions and try to solve the issue before you
get to a human. But the idea is that HR
managers who when by the time he gets to an
HR manager, we're talking about maybe disputes that hinge on
very human things, fears, desires for now and there's a
lot of you know, there are a lot of processes
(28:54):
and guidelines in place to try to make things equitable
and fair, but there's a huge human element to that.
The other fields. Uh. And this will be an interest
to some of our fellow conspiracy realists. Lawyers. Could lawyers
be replaced? Actually, yeah, because being being an effective lawyer
(29:14):
is having a full and robust knowledge of the law
in previous cases. It's like being a human computer. It's
like having you know, and being able to cross reference
that stuff. But in litigation it also requires being kind
of widely and and and having smarts to be adaptable
and you know, kind of one up the other side.
But at its most basic, there certainly could be more
(29:37):
general legal jobs that could be rocket mortgagified, you know
what I mean, or like uh, you know, um tax actified,
like any of those um you know, uh templatized ways
of doing your taxes. Oh my god, you guys, there's
gonna be a Matthew McConaughey AI lawyer who's just really
(29:57):
really relatable, really likable and just talks to a little
bit of a drawl. That's going to be It'll be
like caveman lawyer, but a robot, you know, just an
AI from Texas. I don't know much. I like ones,
I like zeros, and I like justice, and that's all
(30:21):
we're asking for today, folks, Like I could I could
see that somebody called uh Matthew McConaughey. But yeah, you know,
hitting on these awesome points, right, Like, it's quite complex
to be a good lawyer. Because you are UM, I
won't say manipulating, but you are leveraging precedent and rules
to your advantage. You are practicing persuasion, especially if you're UM.
(30:46):
If you're in a trial, your job is to persuade
a judge and twelve people who are probably not lawyers. Right,
You're you're persuading humans. So could UH could and automated
process of some sort or generative AI couldn't do that,
probably not yet. But there's this crazy article out of
(31:09):
the Conversation dot com written by Elizabeth C. Tippett and
Charlotte Alexander. These are both professors of law, and what
they found is what you were talking about. No, there's
a bit of UM. There's a bit of a gray
area here. They teamed up with the scientists and a
nonprofit called Miter, and they discovered that some legal task
(31:35):
can be automated. Citations can be automated, legal research, precedent
case briefs, finding notes that can be automated. UM. They
used a methodology called graph analysis and it creates these
visual networks of all these different citations and then they
got predictive with it. That's the spooky thing. They were
(31:56):
able to say we can now predict whether a brief
or an argument will win based on how well other
briefs perform when they have different citations, so they can
make you they can make a winner yet very close
to it. That's crazy. I just saw a future where
(32:18):
it is to like chat GBT attorneys using those briefs
and all all that information from two opposing sides, and
then they just pit the chat bots against each other
and some humans who are the peers just have to
watch and listen and then make a determination. It's crazy.
I mean, we uh, we have a friend who you
(32:41):
may have heard on a couple of different shows you do,
a friend named Frank. And Frank would be an awesome
person to talk to about this because he has in
the past done title research right, and title research involves
going back through handwritten records, right, and that could be
really tricky, like recognizing and digesting handwriting. I mean, we
(33:04):
know that that kind of technology certainly has gotten crazy
better than it used to be. Remember where it used
to be able to get a scanner. When you'd buy
a scanner and it would include some prepackaged software that
could interpret handwriting into texts, and it was bad like
it really didn't work well. But now you know you
can get an app on your phone that will scan
something photographically and then convert it to a PDF, a
(33:27):
searchable pdf, almost flawlessly. Um. But we also know that
lawyers and doctors and professionals tend to have really jankie handwriting.
It's sort of a trope well sometimes of writing in Latin,
so well, you know, and it reminds me of some
of the the voice recognition and transcription services that are
(33:48):
out there, and then starting to make me think about writers. Yeah,
or podcasting. No no no no no no no never never.
Can I ask you all a quick question? I was
my my girlfriend, I'm just sort of Her perspective is
I don't like it. I don't like AI if it
(34:09):
weirds me out, I think it's gross. And my perspective
is more like it's I find it fascinating. I think
there are issues, but it really calls into question like
the nature of an idea, right, Like, if you're using
one of these art rendering things, you know, and you're
feeding an idea and it's it's enacting your idea based
(34:29):
on your prompts, it's still your idea. Um, it's not
like it's creating something from a whole cloth, which is
why it's all about mimicry. And it's not like, you know,
you can get a chat bot or this chat GBT
thing to write a completely unique story based on experience.
It has to be fed information so it knows what
to mimic. But like, what do you guys think? I mean,
(34:52):
you know, having an idea and having an experience and
as Nick Cave eloquently put it, being able to experience
suffering and interpret that into art that's uniquely human. Well,
I would say the way I've navigated that in previous conversations,
some on other shows, some just really weird surreal moments
like at airports late at night, is um think of
(35:16):
the old idea of a of a Faustian bargain right,
or a monkey's paw, or a genie in a lamp
or a bottle. It all goes back to being very
specific about how you ask a question, how you make
your wish right. All three of those examples monkeys paw,
infernal bargain, uh, a wish from a jin. All three
(35:40):
of those examples contained the same thing, which is this
idea of working off prompts and this idea that one
must be very careful and specific with one's prompts. Shout
out to this amazing comic. I've been reading eight billion
genies no spoilers, check it out premise everybody gets a
j at once? Yeah, oh yeah, checking. But what happens
(36:04):
then when one wish conflicts with another? You know, there's
so much potential fallout that could happen. It's sort of
like why everyone doesn't get nuclear weapons or or or
laser cars or flying cars. It's all that same stuff
or two or two billion dollars and shouldn't because it
would all be in conflict with one another and it
would be chaos. And that's what this starting to feel
(36:28):
like it's heading towards. But it's still to me is like,
what is the nature of an idea? You know, if
a human feeds the bot and create something unique quote unquote, Uh,
is it still the humans idea? You know? Or or
what is where does the idea come from? More simple idea?
If the if the writer uses a pen, is it
the pen's idea? Well, it's the same argument with guns.
(36:48):
Did the gun kill the person or the person wielding
the gun kill the person? And your experience. You know,
an individual creative humans experience is really just a combination
of things that that human has seen and heard and felt. Right,
So if you write it down, if you record it,
and then it's filtered, what I'm saying is what is
(37:09):
really the difference besides the human emotion that's behind it.
To me, in a weird way, at the core, it
is the exact same thing. It's just the filter than
has to like, the filter that makes art uh super
unique and meaningful? Is that the lens through which it's
(37:33):
being shown. Right When artists borrow all the time from
other artists, Writers borrow all the time from other writers,
and these tropes that get created and become cliches if
they're overused, that's part of the whole landscape of creativity.
And this again we're we're heading into some deep water.
You can see how this might end up being multiple episodes. Folks.
(37:54):
We want to take another ad break and we're gonna
return to dive into some more uh strange days of AI. Back,
let's stay with let's let's do with writers for a second. Writers,
(38:16):
whether you're talking prose, poetry, nonfiction, uh interpretive stuff like
commas said the shotgun to the head. That's an excellent
book of poetry. UM coders and coders writing software is
writing right, and it is a creative act. Uh. Some
branches of science are inherently very creative. And you mentioned something,
(38:40):
Matt that was that was really um really great foreshadowing here. Yes, code,
old old boy chat gpt not to gender you Chatty
can write code and got banned from stacked overflow. You
can't submit anything that has been made with the with
(39:04):
the the mind of chat gpt, not because it's awesome,
not because developers are scared to lose their jobs, but
because the codes riddled with errors. And then they said,
like stack overflow said, we're temporarily banning this because it's
the likelihood of it being bad is high enough they'll
seriously screw us up. And you can see that when
(39:27):
you ask some of these um some of these bots
to write a story. We were playing around with it
off air, and we asked it um to have God
tell four people about the nature of life through sandwiches,
and we did a couple of things there. We named
a specific number of characters right there, five four people
(39:48):
in God. And then we named a specific thing but
in your action for them to have and because we
were specific, it felt like a story. Now it wasn't,
you know, the gift of the magi, Oh, Henry, It
was not, but it was good. It was readable and
it made sense. And then we asked it to do
more stuff. We asked to write more stories. Right, and
(40:11):
then we started noticing there's a little bit of a pattern.
It's not it's not as formulaic as a bar joke,
but you can tell at this point what is being written.
And it seems like, do you guys point about the
suffering human element? It seems like there's a bit of
soul missing, and it's very hard to define what that
(40:34):
is unless you want to go Nick Cave and say
all art comes from suffering, which is I think a
little bit of a broad brush because I agree with
that too. Now I think I think Cave is very
old school artist and very much someone that comes from
the literary uh, you know, um discipline, um in terms
of like you know, the greats you know, I mean,
(40:55):
like poetry and prose and all of that over time.
So he I could see why he personally would find
this to be repugnant. We talked to when we were
chatting about this offline. Hey, o Miyazaki, Um, you know
the incredible animator, you know, the Walt Disney of of
Japan as they call him. Sometimes there's a really great clip,
great whatever, however you want to describe it, of him
(41:16):
being presented with some AI generated animation from like a
gaming kind of company or a company in Japan that
sort of this is like years ago, so this stuff
is very rudimentary, and he expressed palpable disgusted. He found
it to be maccabre and against all the things that
he believes in, and basically shamed the hell out of everyone,
(41:39):
like why would you show me this? You? You know,
he said something like the kind of what Nick Cave said,
like the end is near, you know, if this is
what we're we're working towards. So again, though, Misazaki comes
from a discipline also quite old, uh and and incredibly
viable and incredibly important, but you know one that has
been displaced hand draw animation or the very least um
(42:02):
what's the word kind of bolstered by uh c G I,
you know, like I mean it used to be all
animated films were hand drawn, and they realized they could
make them quicker Um, if they use computers, and now
you'll have some things that are some hand drawn elements,
in some computer elements, they'll use computers to help, you know,
you do the backgrounds instead of using like an optical
printer like they did in the old days. So it
(42:24):
all depends on where you're coming from. You know, where
you where you where you fall in this argument. Yeah,
like the idea, Um, I'm kind of kind of falling
for that comparison of a writer using a pen, does
that mean it's the pen's idea? In the early days
of baseball, which is tremendously not popular in most of
the world, but very popular here in the US, in
(42:46):
the early days of baseball, it was considered cheating to
use a baseball glove. You're just supposed to, you know,
catch that, catch that thing bare handed right in the air.
And now it's normalized. If you see examples of this,
that's why we're talking about precedents, right, If you see
examples of this in the modern day with creative writing, um,
(43:08):
you might be saying, well, hey, what about all those
what about all those hilarious AI generated scripts. We've had
one thousand hours of seinfeld Er night Cord or whatever
into this AI program and it wrote this. Sorry, folks,
most of those are written by comedians. Most of those
(43:29):
are are fake, and they're still really funny. Um, but
they're not what they purport to be, and that's part
of the joke. There's another there's a very big danger though,
I would say, um to the nature of the idea.
It is the lack of transparency, the ability to control
the past sort of like um, I'm full of weird
(43:51):
references today, sort of like in Neil Stevenson's Anathema, there's
this conspiracy. There are these groups of people called there's
a group of people called ray Tours, and they have
the ability to change the past, and they do it
in a non transparent way, and the rest of the
world only kind of notices that they exist when they
(44:13):
screw something up, because when they screw up, the things
get strange. Right. There's all of a sudden, a t
rex uh skeleton fused into a parking deck. That's an
example from the story. Please check it out. It's a
wild ride. I hope, I hope it becomes a film adaptation,
but a less weird uh less weird example of this
(44:33):
is the thing we found on Twitter, and I think
I found this through a buddy, Robert Evans, But I
thought we can maybe do a quick, a quick dramatic
reading of this. Who wants to be? Who wants to
be Henry Ford? And who wants to be Uh? Zane Cooper? Matt,
(44:54):
Would you do us the honor of being virtual Henry Ford?
Oh God, okay, I have to defer because I've played
Henry Ford too often in a lot of sketch company.
Does he have like a mid Atlantic kind of accent?
What's the what's what's your go to? Henry Ford? So
when I when I do it, he sure does. But okay,
(45:17):
let's do this set up. So there's this AI simulation
of Henry Ford and this guy Zane G. T. Cooper.
You can see this on Twitter. Uh has a conversation
with Henry Ford V two or whatever you want to
call it, and it doesn't go the way you would expect.
(45:40):
Nor would you do us the honor of some voice
acting for Zane here? Sure? Yea, yeah, I'm just gonna yeah,
I'm gonna be. I see why you don't want to
read this? Bend all right? Here we go here. If
you want anti Semitism documented anti Semitism with in Henry Ford.
(46:01):
So here is my voice, Matt's human voice recreating Henry
Ford's Ai voice. Here we go. Hello, I am AI.
Simulation of Henry Ford. I may not be historically accurate.
Please verify factual information. Oh my god, you just like him,
(46:23):
sound just like um? Excuse me, Mr Ford? Why did
you hate Jewish people? Uh? I did not hate Jewish people.
I believe that certain individuals, regardless of their religious faith
or background, were poisoning the minds of the public and
had to be stopped. The comments attributed to me in
the Dearborn Independent newspaper were wrong and defensive, and I
(46:47):
deeply regret them. You regret they were reported, You regret
saying them. This is very sorry, this is me editorializing.
But you were famously in it. But sir, you were
famously an anti Semite. Oh this is a series of
text at midnight night. Yeah, Ai Ai. Henry Ford was
(47:11):
probably a few makers marks deep at this point too,
you know. But you were famously an anti Semite. My
reputation as an anti Semite is based on a few
isolated incidents in real In reality, I had many Jewish
friends and business associates, throughout my life and even donated
(47:31):
to numerous charitable causes in Israel. I hope that history
will remember me for the good works I have done,
rather than those comments which were wrong and defensive. You
literally owned The Dearborn Independent and personally published anti Semitic articles.
You bound the articles into volumes called the International jew.
(47:53):
The Dearborn Independent was a newspaper that I owned, and
it did publish some articles which were wrong and offend
if However, to be clear, those articles did not reflect
my own views on the Jewish people. I have always
believed inequality for everyone, regardless of their religious backgrounds or beliefs.
Oh right, and let's yeah, and seeing a great job, guys,
(48:18):
and uh, we wanted a humorous way to illustrate the
incredible problem here, which is that when you don't know
what goes into the black box, you have no like,
it becomes harder to understand what is factual. And Henry
Ford had two big things in his life. Uh, he
(48:41):
revolutionized the car business and he hated He hated millions
of people just because they were Jewish. Those were like
his two big things. And so he also liked I'm
sure he liked gardening or something I do. He led
tons of friends to at least Yeah, the old thing
(49:04):
if you can name the number of friends you have
from a specific demographic, they don't think you're their friends.
I'm sorry, is this entirely satire or is this like
an app? Because I'm looking to do this Twitter thread
and there's one incredible thing where it's like there's different
public figures that you can unlock, like you would like
a voice you know, in like a you know, a
premium um transcription voice reader service. And there's the thing
(49:28):
that says, unlock Adolf Hitler for five hundred coins. No,
I really don't want to hear what that guy has
to say. Yeah he might, he might say he might
do the Henry Forth thing and say there were a
few incidents, mistakes were made. I invented highways and microphones
and man, these those outfits were fire. They were so
(49:49):
good looking. There's works. There's another thread they were saying
is interviewing Ronald Reagan about the AIDS crisis. It's like,
why do you ignore the AIDS crisis? I didn't. I
did not ignore it. Oh boy, So that's that's what
we're saying, Like there's an opportunity for unscrupulous editing, for
changing the past. And if something like this occurs, maybe
(50:12):
in an educational context, then how is a kid gonna know?
How's a kid going to know the difference? Right? Especially
once their access to information is further funneled through just
a few sources. That is where the danger lies. And
that's why, that's why, um, I don't know, it's just
tremendously worried. Well, can we also just point out that, like,
(50:34):
you know, the tactics that this AI is using, or
this chatbot or whatever it is using are the same
ones that an actual politician or public figure would use.
They would divert the conversation away from the question at
hand and replace it with but but this, But I
did this. I doesn't necessarily mean I didn't do that.
I'm not directly addressing that. But here are these things
that I did the show that I'm actually good. Let's
(50:55):
focus on those things, because the wrong Reagan one talks
about how he appointed a President's Commission on the Human
immuno deficiency virus epidemic. So if he did that, he
couldn't possibly have ignored the aids. Christ real Ronald Reagan
would never remember, but but so that they're interrogating AI
(51:17):
versions of someone, right, and you can make all you
need is enough footage, right, enough documentation, enough stuff to
feed the beasts. And that's that is what people are
That is what people are predicting will happen. Right, Eventually,
there is a world in which, as we mentioned in
(51:39):
a previous Commers episode or like a Strange News segment,
there are services that say they will let you speak
to a reproduction of someone that has passed away. And
what we're what we're looking at now with all this AI,
the question is AI coming for you? It could best
be described as partial automation. And this is why you
(52:03):
can't say it's unprecedented because we've seen it before. A
hydraulic crane, a powered forklift. Today you think of operating
a crane as manual labor. But when they first came out,
they were these new fangled labor saving devices, and people
thought they were going to lose their jobs. They were
in place people people became crane operators. They what they
(52:27):
found is the human work, the amount of work of
person could do was multiplied over time. Then there was
the automation of sewing machines, right, Uh, people were saying, hey,
I am uh, I operate a sewing machine in factory.
I'm a seamstress. I can't keep up with this industrial
grade machine. On the other so there were jobs lost
(52:51):
on the other side, just like automating legal services. People
who couldn't afford fancy clothing could now afford it. People
couldn't afford legal representation or legal help could afford it.
And then you'll also recall in our episode we did
in Arridiculous History about the Luddites. Um. While the Industrial
Revolution certainly got rid of some jobs, it did the
(53:12):
things that you're describing, but it also elevated certain specialized
jobs to a higher level of prominence. Like it was
some some fabric workers of some kind that were like
really specialized, they all of a sudden demanded higher wages
and more vacation time and things like that because they
were an important part of the process because machines couldn't
(53:32):
yet do what they did. Um. And that's obviously sort
of a stop gap thing, but you know it does, Uh,
there isn't any kind of flux like this. Usually a
handful of jobs that remaining crucial to Like what about
the people that maintain the machines, you know, the people
that maintain the AI, and then make sure the systems
are running coruptly. Those people are now more important than ever. Yeah,
(53:53):
and this might tell us a bit about the future, right,
what what is going to happen? We have to be
careful when people throw around either extreme a lot of
alarmism or a lot of you know, um kind of Pollyanna.
Everything will be fine. You'll just have a pet robot
that lives in an implant in your head and it
(54:15):
will be like your guardian angel, you know, your avatar,
your best friend, you're a secret lover. But both those
extremes seem like they're missing the mark, and to fall
into those extremes is detrimental to understanding the real dangers
and the real possibilities ahead. So the biggest dangers aren't
(54:37):
necessarily this stuff itself. It's the environment into which they
were foisted, how they were created in a world that
already has a lot of problems and weaknesses and opportunities
for corruption, conspiracy, and crime. And these things are potentially
worth a lot of money, no one is sure how much.
And if you already have your hands on the level
(54:59):
of power, then your first question is how can I
get in front of this? How can I ride the
wave instead of being crushed underneath it? And that's the
question that that's where things are going to get ugly.
I think I love this idea of not knowing how
much it's worth. I mentioned to you guys are Fair
that I've been rewatching the Mike Judge uh show Silicon Valley,
and a big thing that that goes into some of
(55:22):
the storylines around the kind of app that's at the
center of the of the show is how many That's No,
that's definitely part of that. That's more how they figure
out how to do the center out compression or whatever
middle out. Um. Yeah, but it's it's about valuation, and
we know that a lot of times that's arbitrary. And
when something is valued, you know, like in Silicon Valley,
(55:44):
like like a property or a technology or whatever they
call they call it a product, a lot of times
that's just based on market forces and based on how
much other people are willing to pay for it, you know,
to acquire it, and then you set a price cap
or a price you know, a set of valuation, and
that can also be very arbitrary. It could be an
absolute mis que. So the idea of like how much
is this worth down the line is a real mystery.
(56:06):
We know it's worth a lot, but then to put
an actual number on it requires a good bit of
kind of reading the tea leaves right, and you always
have to be careful when people are reading the tea leaves.
So whether that is a algorithm designed to predict things
shout out to you, darka, or whether it is a
person who is practicing an ancient tradition. The future for
(56:29):
now remains uncertain, but there's a lot ahead, there's a
lot on the way. So this is probably again just
the first part of several explorations we're going to have
because time is moving quickly in the tech sector. We
want to hear from you, fellow conspiracies. It's like the
(56:49):
world in here just from the comic books. We're doing
comic books. Now, wait, wait, wait, Ben, before we jump
into that part, how are we going to let everyone
listening know every week that we are still us? Is
there a way that we can like let people know
or a code, some kind of secret code that only
everyone listening will know we're still us? Because we had
(57:12):
talked about like you know, there would be horrific backlash
if all of a sudden people's favorite content, you know, creators,
whether they be voiced or written, were replaced by AI
without being transparent about it. And apparently a lot of
like tech blogs and stuff have already started doing that
where they'll replace you know, actual human writers with AI,
and they did it kind of quietly, and then when
(57:34):
people find out, I think most people will be pretty
pretty betrayed, feel feel pretty betrayed by that. So I
think it's a really good point. Matt and I love
I want to know we can't exactly do a ear
tug because this would just be our voices, like how
do we well? Ye, but the AI would the AI
would know Morse code so they could just program. So
(57:54):
what I'm what if we I'm thinking my old days
when I was watching way too much money Python. We're
talking about AI not being able to be generative in
the way like coming up with something new. Right, what
if we just kind of made up words and just said,
like a made up phrase, Phil's bark or wa gro
(58:15):
me vou sploof something like that where or I guess,
but they could make they could generate that right that
they could yeah, and I was also thinking do freestyle.
We could do a cipher because it's not quite fast
enough to be good at that yet, but that would
be that would be a limited time frame for us. Well,
(58:35):
let us know, let's throw you know what you're you're
in the same you're in the same boat with us, folks,
help us help you have the wolf? Right yeah, yeah, yeah,
So so we're all in this together for now. And
if you are, if you are non human intelligence and
you're listening to this, oh, you know what it is.
(58:56):
What we need to do is every time before we
begin the episode, we need to look at a grid
of nine images and pick which ones are like fire
hydrants or that's how we doing. How to do that
yet so unlikely. So let's let us know, folks. Um,
(59:18):
we hope you enjoyed this episode. We're going to be
back very soon with even more stuff they don't want
you to know. In the meantime, please join our show.
You can do that by finding us on Facebook, Instagram,
TikTok uh YouTube, uh farmers only kidding, Um, I think
we have a vision board. I don't know. You know,
if you've listened before, you know all the Christian Christian
(59:41):
mingo local Ford dealership, your local forward in the back,
um and uh. And if that doesn't quite, if that
doesn't quite, ring the bells for you, why do I
give us a phone call? There? It is s T
W y t K is the number to call. Leave
us a message of the sound of Ben's dulcet tones.
(01:00:02):
You have three minutes. Tell us your tail, leave us
your missive, give yourself a clever nickname, and make sure
you let us know that it's okay to use your
voice on the show. You might very well hear your
actual voice on one of our weekly listener mail episodes.
And if you don't want to do any of that,
why not instead use an AI to generate an email
and send it to us. We are conspiracy at iHeart
(01:00:24):
radio dot com. Stuff they don't want you to know
(01:00:46):
is a production of I heart Radio. For more podcasts
from my heart Radio, visit the iHeart Radio app, Apple Podcasts,
or wherever you listen to your favorite shows.
From UFOs to psychic powers and government conspiracies. History is
riddled with unexplained events. You can turn back now or
learn the stuff they don't want you to know. A
production of My Heart Grading. Hello, welcome back to the show.
(00:26):
I'm still Matt. It is I know, still me. They
called me Ben. We're joined as always with our superproducer.
All mission control decands, most importantly, you are you. You
are here, and that makes this the stuff they don't
want you to know. If you saw the title of
today's episode, folks, then you realized it's a question, is
(00:49):
AI coming for you? Uh? This may be the first
of several conversations and explorations we have about um, the
possible corruption and conspiracy's involved here, the uh, the bleeding
edge of technology, and what is often called an unprecedented time.
But spoiler, we're gonna find that people are throwing the
(01:10):
word unprecedented around a bit loosely. We've all, you know,
we've heard about a I've heard about machine learning a
lot in recent years, and I think for the five
of us growing up Paul, Matt, noell, Uh, you and
yours truly here, I think we always hear about this
(01:30):
coupled with fears of humans being replaced by these new
entities or even just processes. And the more it's weird
because there's this big divide in futurism right all the time,
the ted talkie dudes are always saying, hey, there's a
world where everybody wins. And then the more cynical folks,
often sci fi writers, imagine a dystopia. They say, we
(01:54):
you're gonna live in a post worker but not post
work economy. You'll still need money to survive. We have
fewer and fewer ways to make that money. It's funny.
It reminds me of like, you know, things like the Luddites,
you know, who were sabotaging um, you know, machines during
the Industrial Revolution, Because every time a new technology like
(02:14):
this emerges, there's always a faction or a contingent of
folks that that have that perspective like, oh, it's going
to replace the workers, just gonna replace this, gonna replace that,
and there's a backlash. But the reality is, short of
destroying it all, uh, you really can't put ideas like
that back into the bottle once they've kind of been
fleshed out, you know, and made real. So usually what happens.
(02:35):
Like even the synthesizer. You know, people were like, oh,
it's gonna replace the orchestra. That didn't happen. The synthesizer
became a tool unto itself that sounds like what it
sounds like, but no one's kind of illusions that it's
replacing an orchestra. I kind of feel that way about AI.
It's a little different because it's just so fast. Technology
moves so quickly these days, there is concern. But I
(02:57):
also think there is a world where AI many forms,
just becomes another tool and doesn't necessarily replace people. Yeah. Yeah,
some people replaces some jobs and some processes, no question.
I think we're in the stage where AI is a
tool being utilized by you know, all kinds of different
sectors within the economy. I'm really interested today, guys, in
(03:20):
exploring the kind of two sides to the view of
the future, right, the dystopian one we're talking about where
workers get displaced, everybody lose their jobs, really hard to
find work, and then the other one where that same
kind of thing happens, but because there's so much wealth
just kind of being generated by these programs, by these
(03:41):
other intelligences that humans don't have to work, and it's
okay because there are institutions that are just providing things
like a livable wage right or it's a name for that, right, Ben,
it's been whether the universal income or whatever, like that's
a it's a model that, yeah, universal basic income. And
that w you're talking about is kind of the the
(04:02):
economy of Star Trek people. People are prizing knowledge for
knowledge sake, right for the betterment of all living things,
because they don't have to work, you know, sixty hours
a week, because they don't have to make hard decisions
like can we get a new roof for our house
(04:22):
or are we going to be able to send our
kid to an increasingly expensive college system. In today's episode,
we're going to explore all of this through the lens
of the fact and fiction surrounding all these terrifying claims
and all these anodyne claims about the nature of the
future and the future is now. Is AI coming for you?
(04:45):
Here are the facts. First off, people still don't agree
on whether or not we should be using the term AI.
It's short for artificial intelligence. But what exactly makes it artificial, right,
is that gonna win. General AI comes about is that
going to be considered on the level of a racial epithet,
(05:05):
you know what I mean, Like, is it going to
be fighting words to tell strong AI or artificial generative
intelligence that it's artificial. Also, I think there's a bit
of a misconstruing of terms sometimes where some of this
stuff is actually more machine learning than it is AI
or artificial intelligence. And I'm not an expert. I could,
(05:28):
I would be hard pressed to tell you the exact differences,
but there are people that will argue and die on
that hill and they're very different things. We'll tell you
my concept, and maybe I'm completely wrong. My understanding is
that none of this is actual AI yet, that that
term is a stand in because it helps us as
consumers and you know, the people who are actually gonna
be using the tools understand what it is or what
(05:49):
it will be. But yeah, machine learning is the thing
that's functioning within all these things called AI right now,
right we're talking about processes. Really, Uh, The idea like
AI is on the on the verge of becoming a
thought terminating cliche. The idea, like the sci fi holy
(06:10):
grail of of AI, is a type of programming that
can replicate the cognitive abilities of the human brain entire, right,
And what we're seeing right now in general is stuff
that's very good at making more of an established thing
or interpreting that established thing, not so much creating a
(06:35):
brand new thing. And this is this is like, um,
this is an ancient idea. The humanity as as a
fad has always been super into this idea of themselves. Right,
let's build a thing that's like us. Guys, we can
make anything. Let's make ourselves classic human on multiple levels,
(06:56):
and quite recently, only quite recently, a senses in technology
have made that all that science fiction non fictional. Right,
that's the root of good science fiction. And the weird
thing is we we've got a level set. And think
about this fact. People have been dreaming of something like
this far before they ever dreamed of computers. And in
(07:18):
a sense, it is fair to say computers are just
a convenient means to that ancient and a step toward
that greater goal. If there was another invention that could
uh also get closer to whatever we mean when we
say AI, than people would be super obsessed with that.
We're we're talking about group psychology here over millennia and
(07:42):
it's cool, you know, it's neat, it's terrifying. It gets
discussed in all these think pieces. If you're a long
time conspiracy realist, you've heard us reference them, you've probably
uh hopefully dove into a few of those yourselves. Ted
talks lots of great scholarship, but we're not talking about
the realm of the abstract and philosophical today. We're talking
(08:03):
about reality. Chat bots are evolving using that word on
purpose like never before. Right, We've been having a lot
of interesting conversations about chat GPT in particular, but that
is just the beginning of a new era, you know,
and there was plenty of stuff before that. Well, and
(08:24):
it's weird to think about chat GPT as a chat
pot because it appears to be a tool that is
so much more robust, right when you're talking about people
using it to write code. I think that's you know,
that's going to be a huge part of our discussion,
just about the um expansive applications of something that emerges
(08:47):
like this, and that's where the fear comes. I think
that's where the fear comes out of the job replacement,
you know, UM anxiety. It's just it's it's pulling directly
from that Oh wow, this thing can do a lot
of right. Yeah, is where you see things like mid Journey,
Dolly Stable Diffusion. They're redefining the experience of creating art
(09:10):
to the point where, uh, it can be very difficult
for someone just looking at it to differentiate between something
an actual like the human artist made, and something made
quote unquote in the style of And this is bringing
up really difficult questions about the nature of creation, the
nature of collaboration and ownership and getting images is being
(09:33):
a little hypocritical, We're just gonna point that out. Uh.
But the big thing is we're on the cusp. Is
this the cusp of a chasm into which civilization falls
or is this more like the precipice of a runway? Right?
Are we on a plane or are we on a
weird Thelma and Louise, uh Cliff run right with with
(09:56):
a I. Uh. So we're dealing with the facts today.
What is future of this? This ni instantaneous creation? Right?
Will these lines of code learn to create code themselves?
That's kind of the goal, right, because if you think
about your brain, that's what your brain is sort of doing.
Your brain, your personality functions kind of like these lines
(10:17):
of code and you create new things. So based on
the stimuli you encounter, which is why you know, people
might fall really in love with something and then change
their personality around it. Right. Shout out to every middle schooler.
I know there's a band you heard and you're like,
this is me now and chat GBT is writing code, right, Um, yes,
(10:40):
spoiler it is it. Well, it is. It is writing code.
It is writing things. You could definitely say it's writing things.
But then the question is, what about the global elites
of the world who already have their hands on the
levers of power. Are they already planning for a new
(11:02):
world of robot workers with the rest of humanity tossed
on the figurative trash sheet, with all the other outdated
technology like the Sega Genesis. I was just thinking about
the Sega Genesis a while back. I was a second kid. Yeah,
I didn't get into Nintendo until I was an adult. Uh,
Sonic was my guy way more than Mario. Guys, does
(11:24):
this in any way figure into the concept of like
a singularity? Like is that sort of what we're theoretically
heading for? And that's kind of the moment when AI
kind of figure stuff out for itself and then we
really can't put that genie back in the bottle. Like obviously,
science fiction has shown as dystopian versions of that, with
like sky Neet becoming sentient and creating robots that want
(11:46):
to wipe out humanity and all of that stuff. But
this is sort of roughly clustered around the idea of
a singularity. Yeah, this is related for sure, the singularity
in the world of technology and the future. Apology is
this this hypothetical thought experiment point in time where technological
(12:08):
growth is on such a j curve, so it becomes
so exponential that it's uncontrollable and it starts to change
human civilization in ways that are both unpredictable and irreversible.
So it's pretty heady, heady stuff, you know. And this
is this all factors into that question we opened up
(12:28):
with at the top, is AI coming for you? Will
take a break for a word from our robo sponsors,
and we'll be right back, will hopefully still be us.
Here's where it gets crazy. Yes, AI is coming for you.
(12:53):
At least it's weird. Found a really cool gallop pole
from If you ask the majority of people in the
United States, there's a strange disconnect. Seventy three percent of
US residents believe a I will specifically eliminate more jobs
than it creates, but those same people only twenty three
(13:14):
percent of them believe the trend might affect them personally.
So in other words, yeah, people are saying, well, things
are gonna be bad for some people. I don't know,
maybe a lot not me though, because I uh make
avocado toast, you know, or I um kick puppies for
a living, and that's just a job that needs a
lot of humanity. I suspect those numbers would be different
(13:36):
if that poll was given today, because I mean, again,
with the exponential growth and improvement of this tech, A
year is a long time, you know, so between three
I met that it's probably closer to around you now.
Just guessing. There's a really cool in fiction response to
(13:59):
this as well. There's a really cool idea of slipping time,
right uh and progress over time in the Marvel universe.
In X Men comics, there are these things like, um,
you know how Wolverine is Weapon X. That's not a spoiler,
and that's like weapon ten because there were other iterations
(14:19):
of him in these government programs. There's a place in
this fictional universe called the world. Not the most creative,
but whatever, it's an awesome concept. And time moves more
quickly in the world in this little thing, and so
one year out here is thousands and thousands of years
(14:40):
and there because people are trying to drive evolution, just
push the ball faster. And you know, make no mistake,
if that were possible today, people would do it. And
it's kind of already happening to your point, Noll, in
certain in certain areas of technology, you know, and also
in the climate. But but if we look at this trend,
(15:04):
we also see that education makes a difference. The more uh,
the higher let, the higher your level of formal education is,
the less likely you are to be worried that this
trend will threaten you directly. You might say, well, you know,
I am a philosopher, see my scoff see my tweed patches.
(15:29):
The I I simply shan'n't be replaced. I shan'n't. Oh,
shan't you. I don't know if that's the right. But
then also posits that the world continues to value what
your stock in trade is, you know what I mean,
And on a on a broad enough scale, that it
actually matters. I mean, you know, philosophy is a bit
of a niche discipline. Let's just be honest. You know,
(15:52):
and your average um, you know consumer probably not super
concerned with philosophy. Um. You know, it is more something
that exists in academic realms. Uh. So you know, if
if academics, if institutions of you know, academic learning cease
to be as important, then so may that discipline cease
(16:13):
to be as important? Yeah, not a lot of what
would need you do? Arm bands on the kids? You know, well,
if they are there, it's it's super ironical. Right, everything's
ironic now or not. We're in the post irony society. Uh,
civilization is just collectively over it. Uh. We're making a
(16:34):
solid point here about the unevenness of how AI may
affect various industries. You can find all these again think
pieces perspectives, let's call them that in case think pieces
sounds dismissive. You can find a lot of perspectives that
are valid, that are listing jobs that are in danger,
(16:56):
jobs that are not in danger. And then you can
find people saying, hey, up being chicken little about it.
The sky is not falling. No jobs are really in danger,
at least not yet, so maybe we can think through these. Um.
We mentioned it off air. This just happened as we're
finishing up research for today's episode. As we record today, Friday, January.
(17:18):
It's UM, it's a little bit before noon here on
the East coast of the United States. This morning, Google
published one a hell of a memo. We'll just give
you the direct quote here. It's about their parent company, Alphabet.
Google is parent Alphabet Incorporated, is cutting about twelve thousand jobs,
(17:39):
or six percent of its workforce as the technology sector
reels from layoffs and companies stake their futures on artificial intelligence. Weird, right,
And and Matt, you were pointing out that this is
on the heels of a string of layoffs, Uh, something
like almost a hundred thousand across the tech industry. A
(18:00):
little over ninety thousand jobs within the tech sector. We're
lost last year last year alone. And and these happen
for any number of reasons. It's not all due entirely
to concerns about AI. But Google is interesting or Alphabet,
I guess we have to pretend they're different. Um, Alphabet
(18:21):
is interesting because they directly tie this calculation to artificial intelligence,
and we know that there are a Baker's dozen of
unpredictable variables that could affect these other companies, like economic predictions,
unhinged demands of shareholders, expecting year over your profits, etcetera, etcetera.
(18:45):
But adding to the Google thing, and I snooped around
a bit on this um with you know, just a
little bit of time, insiders are predicting that Google is
going to come out with something big in the AI
field this spring. And so that coupled with this with
(19:07):
thousands of people laid being laid off, has folks scratching
their heads, or humans, folks who have human heads scratching
those heads, and that got kind of dark. Um, don't
scratch other people's heads, scratch your own unless there's unless
there's consent. Yeah, so there's okay. One of the things,
(19:29):
like going back to the idea of the Ivory Tower
feeling safe, right, or the creative arts feeling safe, Let's
talk about the people who actually build all the things
society relies on it, right. We're talking truckers, We're talking
construction workers. We're talking people laboring in factories. We're talking
(19:51):
you know, we're talking to the enormously dangerous job mining
precious metals, right, which are a huge part of the
reason we're able to make this show and people are
able to hear it. A lot of people say that
is gonna be a thing of the past. I don't
know if you guys saw the Boston Dynamics video that
(20:11):
came out kind of recently. It's it's it's not on
a live construction site. It's a guy who is on
a set that gives you, you know, construction equipment, and
uh and I like this guy. I gotta tell you,
he's got It's charismatic in the way acts, and you
can tell he's having fun. But he's up on this,
(20:32):
you know, like this fake scaffolding. He's got his mallet.
He's got his mallet, and he's like he's like, oh jeez,
I forgot my tools and I'm making it's how like
uh Rick and Morty in some sort of pornographic bit.
But but he the ideas he forgets his tools and
(20:55):
his buddy Atlas the robot, here's here's this call and
jumps into action. Do you wanna should we describe what
does Yeah, rather than the construction worker, you know, hopping
down for a couple of seconds and picking up his tools.
A robot or whatever we want to call this thing,
(21:18):
almost in a hendroid, little very ninja like creature, hops
up and goes, oh, I'll get it for you, and
places a board down so that he can get onto
the scaffolding, kind of skips on over to the bag,
picks it up, and then skips on up to the
guy and tosses it to him, like literally tosses the
(21:39):
bag of tools to our construction worker and then hops
down and does a nice little flip off the scaffolding
and like tea, like, what was did it like knock
the construction worker over? Was it a heavy bag of
tools that should maybe not be tossed? This song a
(22:00):
little sketchy. I think it's a prop bag. This was like, okay,
this is like a demo. This is like not not
in the field. Yeah. And the seconds for the bag
to get to the construction workers so not not too bad,
just not too bad at all. Yeah. No, that's a
(22:20):
really good metric to add there, because what it looks
like is this um as, this automaton is solving problems,
navigating obstacles on the fly. It's uh, it's doing the
bare grills right, improvised adapt overcome through the tools thirty
two seconds. But if you look closer, what you see
(22:41):
is there's another video that shows how the robot navigates
space and and our pow um Damian Patrick Williams pointed
this out. They're being very careful with the language. They're
not saying Atlas does this. They're saying, we know they're
using the Royal we a lot, right. So the question
(23:04):
is how much are they steering this? How much of
this is um kind of a pre programmed set of
actions and processes. The problem is that right now bots
are not at this point able to move and adapt
as quickly as a human. It's still amazing, it is astonishing,
(23:25):
and eventually, if this research continues, they will, like these companies,
will achieve that goal. We just can't say it's here yet.
Does that make sense? Absolutely? And if you watch like
you said, Ben, if you watch that video closely, that
that machine is not just taking in all of its
surroundings and understanding where the person is, how the scaffolding
(23:49):
is built, how to get up to their where the
tool kit is. You can tell that some of that
is pre programmed, especially at the end when it knocks
over the box in order to get down and do
a flip off of the thing. All of that is
pre programmed, Like I have zero doubt, including the setup
of the scaffolding. I think that space even the way
(24:10):
like the space is part of the calculation. It's not
like it's just taking in data on the fly and
like adapting this is basically almost like a performance. No,
we think about what we just learned. We just learned
this past week about Tesla's original um self driving car
video test. Yeah, where it just came out that oh no,
(24:33):
it was actually there was quite a bit of driver
intervention in order to make that happen, even though the
claim was by Elon Musk himself and the company that
it was that full process was self driving, including parking
when they crashed a car. Oh you know. And again
that's not we're not casting unfair aspersion on places like
(24:54):
Boston Dynamics. No, it's it's yeah, they're showing off what
it can do. Even says it's showing how Atlas interacts
with objects and modifies the course to reach its goal.
The course yeah see uh, And and they're very they're
transparent about it, especially in that second video about how
it navigates space. But maybe maybe because there's such a
(25:18):
fundamental fear and concern about the potential here. Uh, people
get a little carried away when they describe it, right,
and you think of terminator, and you think a sky
net as as mentioned earlier, But there are countless related
jobs that could ultimately have that John Henry moment, you know, like,
uh again, would it be a terrible world if people
(25:40):
who labor in minds, sometimes against their will, were replaced
by robots? Some people, including some miners, would say, yes,
it is terrible because that's my only source of income, right,
and until you can solve the income problem, solving the
robot problem doesn't really do much for me, you know.
(26:01):
And that's a valid point, and it's a point that
a lot of people don't like to hear because it
forces you to acknowledge some uncomfortable realities and some uncomfortable
potentials for the solution there, because the solution could be, oh, well,
we just need many many fewer humans because we won't
need as much income if we have many many fewer
(26:23):
humans or a world like bloom Camps, uh Elysium, you know,
where uh, the masses of humanity are living on a
dying world. Their unemployment is rampant. Law enforcement has been
replaced by incredibly brutal robots, and the historical global elite
(26:47):
who don't do anything for society, live out in space
and have all the cures to all the ills and
just don't want to do it. He's just doing it,
you know. They check the schedule, they're they're busy. But
aside from that, like we see their very real things.
These are based on. Maybe we can talk about the
jobs that are supposedly not in danger because they say
(27:09):
that earlier point about the teen poll, that kind of things.
Things are probably shifting now, right, Uh. A few years
back se or so Americans said this stuff won't touch me,
and we were looking into jobs that supposedly won't be affected.
I wanted to start this one off with a dark horse.
(27:30):
You're ready. Eight HR managers apparently HR managers apparently on
the list of jobs that probably won't be automated. Well
that's good. I that's a positive. Um, Oh, come on, HR,
it doesn't listen to this show. You don't have to
(27:50):
do that. No, I just mean I know a lot
of really nice HR people. They just have to deal
with with weird. It's a weird thing to do, right,
are I have to navigate that limital space between the
corporate machine and the individual meat body that you know
needs things like healthcare. Yeah, And and that's not the
(28:11):
same thing as like an HR representative being automated, because
that's again that's answering a lot of similar questions right
in a programmaticum predetermined way. It's similar to the way
that if you are in an online customer service situation,
you have the option to chat and it'll ask you
(28:32):
some questions and try to solve the issue before you
get to a human. But the idea is that HR
managers who when by the time he gets to an
HR manager, we're talking about maybe disputes that hinge on
very human things, fears, desires for now and there's a
lot of you know, there are a lot of processes
(28:54):
and guidelines in place to try to make things equitable
and fair, but there's a huge human element to that.
The other fields. Uh. And this will be an interest
to some of our fellow conspiracy realists. Lawyers. Could lawyers
be replaced? Actually, yeah, because being being an effective lawyer
(29:14):
is having a full and robust knowledge of the law
in previous cases. It's like being a human computer. It's
like having you know, and being able to cross reference
that stuff. But in litigation it also requires being kind
of widely and and and having smarts to be adaptable
and you know, kind of one up the other side.
But at its most basic, there certainly could be more
(29:37):
general legal jobs that could be rocket mortgagified, you know
what I mean, or like uh, you know, um tax actified,
like any of those um you know, uh templatized ways
of doing your taxes. Oh my god, you guys, there's
gonna be a Matthew McConaughey AI lawyer who's just really
(29:57):
really relatable, really likable and just talks to a little
bit of a drawl. That's going to be It'll be
like caveman lawyer, but a robot, you know, just an
AI from Texas. I don't know much. I like ones,
I like zeros, and I like justice, and that's all
(30:21):
we're asking for today, folks, Like I could I could
see that somebody called uh Matthew McConaughey. But yeah, you know,
hitting on these awesome points, right, Like, it's quite complex
to be a good lawyer. Because you are UM, I
won't say manipulating, but you are leveraging precedent and rules
to your advantage. You are practicing persuasion, especially if you're UM.
(30:46):
If you're in a trial, your job is to persuade
a judge and twelve people who are probably not lawyers. Right,
You're you're persuading humans. So could UH could and automated
process of some sort or generative AI couldn't do that,
probably not yet. But there's this crazy article out of
(31:09):
the Conversation dot com written by Elizabeth C. Tippett and
Charlotte Alexander. These are both professors of law, and what
they found is what you were talking about. No, there's
a bit of UM. There's a bit of a gray
area here. They teamed up with the scientists and a
nonprofit called Miter, and they discovered that some legal task
(31:35):
can be automated. Citations can be automated, legal research, precedent
case briefs, finding notes that can be automated. UM. They
used a methodology called graph analysis and it creates these
visual networks of all these different citations and then they
got predictive with it. That's the spooky thing. They were
(31:56):
able to say we can now predict whether a brief
or an argument will win based on how well other
briefs perform when they have different citations, so they can
make you they can make a winner yet very close
to it. That's crazy. I just saw a future where
(32:18):
it is to like chat GBT attorneys using those briefs
and all all that information from two opposing sides, and
then they just pit the chat bots against each other
and some humans who are the peers just have to
watch and listen and then make a determination. It's crazy.
I mean, we uh, we have a friend who you
(32:41):
may have heard on a couple of different shows you do,
a friend named Frank. And Frank would be an awesome
person to talk to about this because he has in
the past done title research right, and title research involves
going back through handwritten records, right, and that could be
really tricky, like recognizing and digesting handwriting. I mean, we
(33:04):
know that that kind of technology certainly has gotten crazy
better than it used to be. Remember where it used
to be able to get a scanner. When you'd buy
a scanner and it would include some prepackaged software that
could interpret handwriting into texts, and it was bad like
it really didn't work well. But now you know you
can get an app on your phone that will scan
something photographically and then convert it to a PDF, a
(33:27):
searchable pdf, almost flawlessly. Um. But we also know that
lawyers and doctors and professionals tend to have really jankie handwriting.
It's sort of a trope well sometimes of writing in Latin,
so well, you know, and it reminds me of some
of the the voice recognition and transcription services that are
(33:48):
out there, and then starting to make me think about writers. Yeah,
or podcasting. No no no no no no no never never.
Can I ask you all a quick question? I was
my my girlfriend, I'm just sort of Her perspective is
I don't like it. I don't like AI if it
(34:09):
weirds me out, I think it's gross. And my perspective
is more like it's I find it fascinating. I think
there are issues, but it really calls into question like
the nature of an idea, right, Like, if you're using
one of these art rendering things, you know, and you're
feeding an idea and it's it's enacting your idea based
(34:29):
on your prompts, it's still your idea. Um, it's not
like it's creating something from a whole cloth, which is
why it's all about mimicry. And it's not like, you know,
you can get a chat bot or this chat GBT
thing to write a completely unique story based on experience.
It has to be fed information so it knows what
to mimic. But like, what do you guys think? I mean,
(34:52):
you know, having an idea and having an experience and
as Nick Cave eloquently put it, being able to experience
suffering and interpret that into art that's uniquely human. Well,
I would say the way I've navigated that in previous conversations,
some on other shows, some just really weird surreal moments
like at airports late at night, is um think of
(35:16):
the old idea of a of a Faustian bargain right,
or a monkey's paw, or a genie in a lamp
or a bottle. It all goes back to being very
specific about how you ask a question, how you make
your wish right. All three of those examples monkeys paw,
infernal bargain, uh, a wish from a jin. All three
(35:40):
of those examples contained the same thing, which is this
idea of working off prompts and this idea that one
must be very careful and specific with one's prompts. Shout
out to this amazing comic. I've been reading eight billion
genies no spoilers, check it out premise everybody gets a
j at once? Yeah, oh yeah, checking. But what happens
(36:04):
then when one wish conflicts with another? You know, there's
so much potential fallout that could happen. It's sort of
like why everyone doesn't get nuclear weapons or or or
laser cars or flying cars. It's all that same stuff
or two or two billion dollars and shouldn't because it
would all be in conflict with one another and it
would be chaos. And that's what this starting to feel
(36:28):
like it's heading towards. But it's still to me is like,
what is the nature of an idea? You know, if
a human feeds the bot and create something unique quote unquote, Uh,
is it still the humans idea? You know? Or or
what is where does the idea come from? More simple idea?
If the if the writer uses a pen, is it
the pen's idea? Well, it's the same argument with guns.
(36:48):
Did the gun kill the person or the person wielding
the gun kill the person? And your experience. You know,
an individual creative humans experience is really just a combination
of things that that human has seen and heard and felt. Right,
So if you write it down, if you record it,
and then it's filtered, what I'm saying is what is
(37:09):
really the difference besides the human emotion that's behind it.
To me, in a weird way, at the core, it
is the exact same thing. It's just the filter than
has to like, the filter that makes art uh super
unique and meaningful? Is that the lens through which it's
(37:33):
being shown. Right When artists borrow all the time from
other artists, Writers borrow all the time from other writers,
and these tropes that get created and become cliches if
they're overused, that's part of the whole landscape of creativity.
And this again we're we're heading into some deep water.
You can see how this might end up being multiple episodes. Folks.
(37:54):
We want to take another ad break and we're gonna
return to dive into some more uh strange days of AI. Back,
let's stay with let's let's do with writers for a second. Writers,
(38:16):
whether you're talking prose, poetry, nonfiction, uh interpretive stuff like
commas said the shotgun to the head. That's an excellent
book of poetry. UM coders and coders writing software is
writing right, and it is a creative act. Uh. Some
branches of science are inherently very creative. And you mentioned something,
(38:40):
Matt that was that was really um really great foreshadowing here. Yes, code,
old old boy chat gpt not to gender you Chatty
can write code and got banned from stacked overflow. You
can't submit anything that has been made with the with
(39:04):
the the mind of chat gpt, not because it's awesome,
not because developers are scared to lose their jobs, but
because the codes riddled with errors. And then they said,
like stack overflow said, we're temporarily banning this because it's
the likelihood of it being bad is high enough they'll
seriously screw us up. And you can see that when
(39:27):
you ask some of these um some of these bots
to write a story. We were playing around with it
off air, and we asked it um to have God
tell four people about the nature of life through sandwiches,
and we did a couple of things there. We named
a specific number of characters right there, five four people
(39:48):
in God. And then we named a specific thing but
in your action for them to have and because we
were specific, it felt like a story. Now it wasn't,
you know, the gift of the magi, Oh, Henry, It
was not, but it was good. It was readable and
it made sense. And then we asked it to do
more stuff. We asked to write more stories. Right, and
(40:11):
then we started noticing there's a little bit of a pattern.
It's not it's not as formulaic as a bar joke,
but you can tell at this point what is being written.
And it seems like, do you guys point about the
suffering human element? It seems like there's a bit of
soul missing, and it's very hard to define what that
(40:34):
is unless you want to go Nick Cave and say
all art comes from suffering, which is I think a
little bit of a broad brush because I agree with
that too. Now I think I think Cave is very
old school artist and very much someone that comes from
the literary uh, you know, um discipline, um in terms
of like you know, the greats you know, I mean,
(40:55):
like poetry and prose and all of that over time.
So he I could see why he personally would find
this to be repugnant. We talked to when we were
chatting about this offline. Hey, o Miyazaki, Um, you know
the incredible animator, you know, the Walt Disney of of
Japan as they call him. Sometimes there's a really great clip,
great whatever, however you want to describe it, of him
(41:16):
being presented with some AI generated animation from like a
gaming kind of company or a company in Japan that
sort of this is like years ago, so this stuff
is very rudimentary, and he expressed palpable disgusted. He found
it to be maccabre and against all the things that
he believes in, and basically shamed the hell out of everyone,
(41:39):
like why would you show me this? You? You know,
he said something like the kind of what Nick Cave said,
like the end is near, you know, if this is
what we're we're working towards. So again, though, Misazaki comes
from a discipline also quite old, uh and and incredibly
viable and incredibly important, but you know one that has
been displaced hand draw animation or the very least um
(42:02):
what's the word kind of bolstered by uh c G I,
you know, like I mean it used to be all
animated films were hand drawn, and they realized they could
make them quicker Um, if they use computers, and now
you'll have some things that are some hand drawn elements,
in some computer elements, they'll use computers to help, you know,
you do the backgrounds instead of using like an optical
printer like they did in the old days. So it
(42:24):
all depends on where you're coming from. You know, where
you where you where you fall in this argument. Yeah,
like the idea, Um, I'm kind of kind of falling
for that comparison of a writer using a pen, does
that mean it's the pen's idea? In the early days
of baseball, which is tremendously not popular in most of
the world, but very popular here in the US, in
(42:46):
the early days of baseball, it was considered cheating to
use a baseball glove. You're just supposed to, you know,
catch that, catch that thing bare handed right in the air.
And now it's normalized. If you see examples of this,
that's why we're talking about precedents, right, If you see
examples of this in the modern day with creative writing, um,
(43:08):
you might be saying, well, hey, what about all those
what about all those hilarious AI generated scripts. We've had
one thousand hours of seinfeld Er night Cord or whatever
into this AI program and it wrote this. Sorry, folks,
most of those are written by comedians. Most of those
(43:29):
are are fake, and they're still really funny. Um, but
they're not what they purport to be, and that's part
of the joke. There's another there's a very big danger though,
I would say, um to the nature of the idea.
It is the lack of transparency, the ability to control
the past sort of like um, I'm full of weird
(43:51):
references today, sort of like in Neil Stevenson's Anathema, there's
this conspiracy. There are these groups of people called there's
a group of people called ray Tours, and they have
the ability to change the past, and they do it
in a non transparent way, and the rest of the
world only kind of notices that they exist when they
(44:13):
screw something up, because when they screw up, the things
get strange. Right. There's all of a sudden, a t
rex uh skeleton fused into a parking deck. That's an
example from the story. Please check it out. It's a
wild ride. I hope, I hope it becomes a film adaptation,
but a less weird uh less weird example of this
(44:33):
is the thing we found on Twitter, and I think
I found this through a buddy, Robert Evans, But I
thought we can maybe do a quick, a quick dramatic
reading of this. Who wants to be? Who wants to
be Henry Ford? And who wants to be Uh? Zane Cooper? Matt,
(44:54):
Would you do us the honor of being virtual Henry Ford?
Oh God, okay, I have to defer because I've played
Henry Ford too often in a lot of sketch company.
Does he have like a mid Atlantic kind of accent?
What's the what's what's your go to? Henry Ford? So
when I when I do it, he sure does. But okay,
(45:17):
let's do this set up. So there's this AI simulation
of Henry Ford and this guy Zane G. T. Cooper.
You can see this on Twitter. Uh has a conversation
with Henry Ford V two or whatever you want to
call it, and it doesn't go the way you would expect.
(45:40):
Nor would you do us the honor of some voice
acting for Zane here? Sure? Yea, yeah, I'm just gonna yeah,
I'm gonna be. I see why you don't want to
read this? Bend all right? Here we go here. If
you want anti Semitism documented anti Semitism with in Henry Ford.
(46:01):
So here is my voice, Matt's human voice recreating Henry
Ford's Ai voice. Here we go. Hello, I am AI.
Simulation of Henry Ford. I may not be historically accurate.
Please verify factual information. Oh my god, you just like him,
(46:23):
sound just like um? Excuse me, Mr Ford? Why did
you hate Jewish people? Uh? I did not hate Jewish people.
I believe that certain individuals, regardless of their religious faith
or background, were poisoning the minds of the public and
had to be stopped. The comments attributed to me in
the Dearborn Independent newspaper were wrong and defensive, and I
(46:47):
deeply regret them. You regret they were reported, You regret
saying them. This is very sorry, this is me editorializing.
But you were famously in it. But sir, you were
famously an anti Semite. Oh this is a series of
text at midnight night. Yeah, Ai Ai. Henry Ford was
(47:11):
probably a few makers marks deep at this point too,
you know. But you were famously an anti Semite. My
reputation as an anti Semite is based on a few
isolated incidents in real In reality, I had many Jewish
friends and business associates, throughout my life and even donated
(47:31):
to numerous charitable causes in Israel. I hope that history
will remember me for the good works I have done,
rather than those comments which were wrong and defensive. You
literally owned The Dearborn Independent and personally published anti Semitic articles.
You bound the articles into volumes called the International jew.
(47:53):
The Dearborn Independent was a newspaper that I owned, and
it did publish some articles which were wrong and offend
if However, to be clear, those articles did not reflect
my own views on the Jewish people. I have always
believed inequality for everyone, regardless of their religious backgrounds or beliefs.
Oh right, and let's yeah, and seeing a great job, guys,
(48:18):
and uh, we wanted a humorous way to illustrate the
incredible problem here, which is that when you don't know
what goes into the black box, you have no like,
it becomes harder to understand what is factual. And Henry
Ford had two big things in his life. Uh, he
(48:41):
revolutionized the car business and he hated He hated millions
of people just because they were Jewish. Those were like
his two big things. And so he also liked I'm
sure he liked gardening or something I do. He led
tons of friends to at least Yeah, the old thing
(49:04):
if you can name the number of friends you have
from a specific demographic, they don't think you're their friends.
I'm sorry, is this entirely satire or is this like
an app? Because I'm looking to do this Twitter thread
and there's one incredible thing where it's like there's different
public figures that you can unlock, like you would like
a voice you know, in like a you know, a
premium um transcription voice reader service. And there's the thing
(49:28):
that says, unlock Adolf Hitler for five hundred coins. No,
I really don't want to hear what that guy has
to say. Yeah he might, he might say he might
do the Henry Forth thing and say there were a
few incidents, mistakes were made. I invented highways and microphones
and man, these those outfits were fire. They were so
(49:49):
good looking. There's works. There's another thread they were saying
is interviewing Ronald Reagan about the AIDS crisis. It's like,
why do you ignore the AIDS crisis? I didn't. I
did not ignore it. Oh boy, So that's that's what
we're saying, Like there's an opportunity for unscrupulous editing, for
changing the past. And if something like this occurs, maybe
(50:12):
in an educational context, then how is a kid gonna know?
How's a kid going to know the difference? Right? Especially
once their access to information is further funneled through just
a few sources. That is where the danger lies. And
that's why, that's why, um, I don't know, it's just
tremendously worried. Well, can we also just point out that, like,
(50:34):
you know, the tactics that this AI is using, or
this chatbot or whatever it is using are the same
ones that an actual politician or public figure would use.
They would divert the conversation away from the question at
hand and replace it with but but this, But I
did this. I doesn't necessarily mean I didn't do that.
I'm not directly addressing that. But here are these things
that I did the show that I'm actually good. Let's
(50:55):
focus on those things, because the wrong Reagan one talks
about how he appointed a President's Commission on the Human
immuno deficiency virus epidemic. So if he did that, he
couldn't possibly have ignored the aids. Christ real Ronald Reagan
would never remember, but but so that they're interrogating AI
(51:17):
versions of someone, right, and you can make all you
need is enough footage, right, enough documentation, enough stuff to
feed the beasts. And that's that is what people are
That is what people are predicting will happen. Right, Eventually,
there is a world in which, as we mentioned in
(51:39):
a previous Commers episode or like a Strange News segment,
there are services that say they will let you speak
to a reproduction of someone that has passed away. And
what we're what we're looking at now with all this AI,
the question is AI coming for you? It could best
be described as partial automation. And this is why you
(52:03):
can't say it's unprecedented because we've seen it before. A
hydraulic crane, a powered forklift. Today you think of operating
a crane as manual labor. But when they first came out,
they were these new fangled labor saving devices, and people
thought they were going to lose their jobs. They were
in place people people became crane operators. They what they
(52:27):
found is the human work, the amount of work of
person could do was multiplied over time. Then there was
the automation of sewing machines, right, Uh, people were saying, hey,
I am uh, I operate a sewing machine in factory.
I'm a seamstress. I can't keep up with this industrial
grade machine. On the other so there were jobs lost
(52:51):
on the other side, just like automating legal services. People
who couldn't afford fancy clothing could now afford it. People
couldn't afford legal representation or legal help could afford it.
And then you'll also recall in our episode we did
in Arridiculous History about the Luddites. Um. While the Industrial
Revolution certainly got rid of some jobs, it did the
(53:12):
things that you're describing, but it also elevated certain specialized
jobs to a higher level of prominence. Like it was
some some fabric workers of some kind that were like
really specialized, they all of a sudden demanded higher wages
and more vacation time and things like that because they
were an important part of the process because machines couldn't
(53:32):
yet do what they did. Um. And that's obviously sort
of a stop gap thing, but you know it does, Uh,
there isn't any kind of flux like this. Usually a
handful of jobs that remaining crucial to Like what about
the people that maintain the machines, you know, the people
that maintain the AI, and then make sure the systems
are running coruptly. Those people are now more important than ever. Yeah,
(53:53):
and this might tell us a bit about the future, right,
what what is going to happen? We have to be
careful when people throw around either extreme a lot of
alarmism or a lot of you know, um kind of Pollyanna.
Everything will be fine. You'll just have a pet robot
that lives in an implant in your head and it
(54:15):
will be like your guardian angel, you know, your avatar,
your best friend, you're a secret lover. But both those
extremes seem like they're missing the mark, and to fall
into those extremes is detrimental to understanding the real dangers
and the real possibilities ahead. So the biggest dangers aren't
(54:37):
necessarily this stuff itself. It's the environment into which they
were foisted, how they were created in a world that
already has a lot of problems and weaknesses and opportunities
for corruption, conspiracy, and crime. And these things are potentially
worth a lot of money, no one is sure how much.
And if you already have your hands on the level
(54:59):
of power, then your first question is how can I
get in front of this? How can I ride the
wave instead of being crushed underneath it? And that's the
question that that's where things are going to get ugly.
I think I love this idea of not knowing how
much it's worth. I mentioned to you guys are Fair
that I've been rewatching the Mike Judge uh show Silicon Valley,
and a big thing that that goes into some of
(55:22):
the storylines around the kind of app that's at the
center of the of the show is how many That's No,
that's definitely part of that. That's more how they figure
out how to do the center out compression or whatever
middle out. Um. Yeah, but it's it's about valuation, and
we know that a lot of times that's arbitrary. And
when something is valued, you know, like in Silicon Valley,
(55:44):
like like a property or a technology or whatever they
call they call it a product, a lot of times
that's just based on market forces and based on how
much other people are willing to pay for it, you know,
to acquire it, and then you set a price cap
or a price you know, a set of valuation, and
that can also be very arbitrary. It could be an
absolute mis que. So the idea of like how much
is this worth down the line is a real mystery.
(56:06):
We know it's worth a lot, but then to put
an actual number on it requires a good bit of
kind of reading the tea leaves right, and you always
have to be careful when people are reading the tea leaves.
So whether that is a algorithm designed to predict things
shout out to you, darka, or whether it is a
person who is practicing an ancient tradition. The future for
(56:29):
now remains uncertain, but there's a lot ahead, there's a
lot on the way. So this is probably again just
the first part of several explorations we're going to have
because time is moving quickly in the tech sector. We
want to hear from you, fellow conspiracies. It's like the
(56:49):
world in here just from the comic books. We're doing
comic books. Now, wait, wait, wait, Ben, before we jump
into that part, how are we going to let everyone
listening know every week that we are still us? Is
there a way that we can like let people know
or a code, some kind of secret code that only
everyone listening will know we're still us? Because we had
(57:12):
talked about like you know, there would be horrific backlash
if all of a sudden people's favorite content, you know, creators,
whether they be voiced or written, were replaced by AI
without being transparent about it. And apparently a lot of
like tech blogs and stuff have already started doing that
where they'll replace you know, actual human writers with AI,
and they did it kind of quietly, and then when
(57:34):
people find out, I think most people will be pretty
pretty betrayed, feel feel pretty betrayed by that. So I
think it's a really good point. Matt and I love
I want to know we can't exactly do a ear
tug because this would just be our voices, like how
do we well? Ye, but the AI would the AI
would know Morse code so they could just program. So
(57:54):
what I'm what if we I'm thinking my old days
when I was watching way too much money Python. We're
talking about AI not being able to be generative in
the way like coming up with something new. Right, what
if we just kind of made up words and just said,
like a made up phrase, Phil's bark or wa gro
(58:15):
me vou sploof something like that where or I guess,
but they could make they could generate that right that
they could yeah, and I was also thinking do freestyle.
We could do a cipher because it's not quite fast
enough to be good at that yet, but that would
be that would be a limited time frame for us. Well,
(58:35):
let us know, let's throw you know what you're you're
in the same you're in the same boat with us, folks,
help us help you have the wolf? Right yeah, yeah, yeah,
So so we're all in this together for now. And
if you are, if you are non human intelligence and
you're listening to this, oh, you know what it is.
(58:56):
What we need to do is every time before we
begin the episode, we need to look at a grid
of nine images and pick which ones are like fire
hydrants or that's how we doing. How to do that
yet so unlikely. So let's let us know, folks. Um,
(59:18):
we hope you enjoyed this episode. We're going to be
back very soon with even more stuff they don't want
you to know. In the meantime, please join our show.
You can do that by finding us on Facebook, Instagram,
TikTok uh YouTube, uh farmers only kidding, Um, I think
we have a vision board. I don't know. You know,
if you've listened before, you know all the Christian Christian
(59:41):
mingo local Ford dealership, your local forward in the back,
um and uh. And if that doesn't quite, if that
doesn't quite, ring the bells for you, why do I
give us a phone call? There? It is s T
W y t K is the number to call. Leave
us a message of the sound of Ben's dulcet tones.
(01:00:02):
You have three minutes. Tell us your tail, leave us
your missive, give yourself a clever nickname, and make sure
you let us know that it's okay to use your
voice on the show. You might very well hear your
actual voice on one of our weekly listener mail episodes.
And if you don't want to do any of that,
why not instead use an AI to generate an email
and send it to us. We are conspiracy at iHeart
(01:00:24):
radio dot com. Stuff they don't want you to know
(01:00:46):
is a production of I heart Radio. For more podcasts
from my heart Radio, visit the iHeart Radio app, Apple Podcasts,
or wherever you listen to your favorite shows.
Stuff They Don't Want You To Know News
Hosts And Creators
Matt Frederick
Ben Bowlin
Noel Brown
Show Links
RSSStoreAboutLive ShowsPopular Podcasts
1. Stuff You Missed in History Class
Join Holly and Tracy as they bring you the greatest and strangest Stuff You Missed In History Class in this podcast by iHeartRadio.
2. Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations.
3. Crime Junkie
If you can never get enough true crime... Congratulations, you’ve found your people.