Rise of the Digital God pt1

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:06):
In the dark shadows, in thewhite cold. Fearlessly, we search for
knowledge new and old. We drinkthe strong spirits and read the ancient tones.
The order of the Abercast. Weare the brave and the bold.

(00:33):
The Abercast accult, history, conspiracyand violence. Hey that's me. I'm

(01:19):
your host, John Towers. Thisis the Abercast. So looking back,
I do feel like this show haskind of gotten us ready for a lot
of things that's happening nowadays. Talkedabout, you know, history, he
spoke about, you know, spiritualwarfare. There's a lot of us out

(01:40):
there nowadays that are thinking that weare gonna culture war war, not against
people with different ideologies, political,you know, counter counterparts or opposite numbers,
but we're actually in am I'm sortof spiritual warfare with you know,

(02:04):
a day humans and so forth.I just read a great book by Jonathan
Kahn called The Return of the Godsand has been thinking a lot about specifically
that sort of side of it,you know. But then there are some
things that are have totally blindsided us. There's just happening so quickly, you

(02:25):
know, we haven't really talked aboutlike this transgender thing, which I definitely
think is something that's going on,you know, has its link to this
decline of Western civilization that we're talkingabout. And the other thing that really
is taken off right under our newseshas been this AI business. So that's

(02:52):
what we're doing tonight, is we'replaying catch up. We're gonna play catch
up and try to get on topof this AI all this AI business.
So before we get started, Ido want to say thank you for tuning
in. Welcome aboard, and Ido have my weapon of mass distraction here.
It's a formula perfected over time andit's mixed up to perfection and poured

(03:19):
rice here and my vessel of theart. So let's get this party started
off in the correct manner, allright. So what we do have here
the way that tonight's program is goingto work, is I have Ray Kurzwild's

(03:45):
book here. He wrote a coupleof books about well, he writes,
well, he writes books about thiswhat he calls the singularity, but it
rubs up against m AI, artificialintelligence and this kind of stuff. The
book that I have um in frontof me here this evening, it's called
the Age of Spiritual Machines. AndI want to start there, and this

(04:11):
book was written when when? When? Lord, when was this book written?
Copyright is nineteen ninety nine, sothis is nineteen ninety nine raykrswhile Age
of Spiritual Machines. So we're gonnaread a couple passage out of this uh,
couple passages. We're gonna read acouple passages out of this book there,

(04:31):
and then we're gonna dive into currentevents, current events, Uh what
could go wrong? And I shouldsay, like this stuff has you know,
I'm I was born in the fuckingseventies, man, So you know,
as long as I was old enoughto be watching making my own movie,

(04:55):
watching decisions, you know, Terminator, you know Skynet m with Terminator
with sky Neet, and then youknow, would you like to play a
game? Would you like to playa game of thermonuclear Global thermonuclear War with
Matthew Broderick. So this is alllike ben sort of in lurking around in

(05:16):
my nightmares for quite some time.And I've spoke before about you know,
when I was a kid, Iwas obsessed with nuclear holocaust, so kind
of goes hand in hand with that. But now we're going to see a
different kind of holocaust unfold different infront of us when we get into when
we get into the ai AI world. I'm not taking nuclear thermonuclear global war

(05:41):
off of the table, but I'mgonna say it's a different sort of holocaust.
This is going to be a holocaustof souls. And you know,
there's quite a few reasons why Ifeel like this. You know, we've
taken I'll include myself in this statement. We've taken God, you know,
out of everything. We've taken outof everything. And like Nietzsche says,

(06:04):
remember back when we did we dida huge series on Nietzsche, and when
Nietzsche or Zarathustra, Nietzchie writes Zarathustrasaying, you know, well you've killed
God. You know God is dead, and you know what comes next?

(06:24):
What do we fill that hole with? And I feel like we're filling it
with a bunch of stuff. Ifeel like we're filling that hole, maybe
with weird sex, and we're fillingthat hole maybe with technology. And we're
filling that hole with ourselves. Someof us are filling that hole with boredom

(06:44):
and how to alleviate it. Ifeel like we're pouring I feel like we're
pouring our new gods into every intoevery everything, and I feel like there's
not gonna be too long before peopleare going to start war shipping this thing.
You know, probably not in theform of chet GPT, but I

(07:06):
feel like it's coming, and it'sprobably coming faster and faster. So here's
kurzwhile talking age of spiritual machines.In the eighties, we saw early commercialization
of artificial intelligence, with a waveof new AI computers forming and going public.

(07:29):
Unfortunately, many made the mistake ofconcentrating on a powerful but inherently inefficient
interpretive language called lisp LISP, whichhad been popular in academic AI circles.
The commercial failure of LISP and theAI companies that emphasize it created a backlash.
The field of AI started shredding itsconstituent disciplines and companies, and natural

(07:56):
language understanding characters and speech recognition,robotic machine vision in other areas originally considered
part of the AI discipline now shunned. Association with fields. Label machines with
sharply focused intelligences nonetheless became increasingly persuadepervasive. In the nineties, we saw

(08:18):
it infiltration of our financial institutions bysystems using powerful statistical and adaptive techniques.
Not only were the stock, bond, currency, commodity and other markets managed
and maintained by computerized networks, butthe majority of buy and cell decisions were
initiated by software programs that contain increasinglysophisticated models of their markets. In eighty

(08:43):
seven, stock market crash was blamedin large measure to the rapid interaction of
trading programs. Trends that otherwise wouldhave taken weeks to manifest themselves developed in
minutes. Suitable modifications to these algorithmshave managed to avoid a repeat performance.
Since the nineteen nineties, the electrocardiogram the EKG has come complete with computer's

(09:09):
own diagnosis of one's cardiac health.Intelligent image processing programs enabled doctors to peer
deep into our bodies and our brains, and computerized bioengineering technologies enabled drugs to
be designed on biochemical simulators. Thedisabled have been particularly fortunate beneficiaries of the

(09:31):
age of intelligent machines. Reading machineshave been reading to the blind of the
dyslexic person since nineteen seventies. Thespeech recognition and robotic devices have been assisting
hands disabled individuals. Since the eighties, Perhaps the most dramatic public display of
the changing values of the Age ofknowledge took place in the military. We

(09:56):
saw the first effective ex sample anincreasingly dominant role of machine intelligence in the
Gulf War of nineteen ninety one.The cornerstones of military power from the beginning
of recorded history through most of thetwentieth century, geography, manpower, firepower,

(10:16):
the battle station defenses have been largelyreplaced by the intelligence of software and
electronics, intelligence scanning by unsaved airbornevehicles, weapons finding their way to destinations
through machine visions and pattern recognition,intelligent communications, encoding, protocol, and

(10:37):
other manifestations of the information age havetransformed the nature of war with increasingly important
role of intelligent machines and all phasesof our lives military, medical, economic,
financial, political. It's odd tokeep reading articles with titles such as

(10:58):
whatever happened artificial intelligence? This phenomenonthat Turing had predicted that machine intelligence would
become so pervasive and so comfortable,and so well integrated into our information based
economy that people would fail to evennotice it. So that's kind of the
unspoken thing with chat GPT is it'sdesigned to fool you. And I think

(11:22):
that I'm going to talk about thatlater when we get into the current events,
and we're going to specifically talk aboutwhat it's going to do to the
job market. But it's designed tofool you. It reminds me of people
who walk in the rainforests going toask where are all these species that are
supposed to live here? When thereare several dozens of species of aunt alone

(11:43):
within fifty feet of them? Aremany species of machine intelligence have woven themselves
so seamlessly into our modern rainforests thatthey are all but invisible. Turning off

(12:05):
or an explanation of why we wouldfail to acknowledge intelligence in our machines and
forty seven, He wrote, quote, the extent to which we regard something
as behaving in an intelligent manner isdetermined as much by our own state of
mind as training, as by otherproperties of the object under consideration. If

(12:26):
we were able to explain and predictits behavior, we have little temptation to
imagine intelligence with the same object.Therefore, it is possible that one man
would consider it intelligent and the otherwould not. The second man would have
found out the rules of its behavior. I'm also reminded of Ellen Rich's definition

(12:48):
of artificial intelligence as the study ofhow to make computers do things at which,
at the moment people are better.It is our fate, as artificial
intelligent researchers, never to reach thecarrot dangling in front of us. Artificial
intelligence is inherently defined as the pursuitof difficult computing science problems that that have

(13:13):
yet to be solved. So herewhat is intelligence? A goal may be
survival even if vade a foe,foraging for food, finding shelter. Or

(13:33):
it might be communication, relating experienceof voke of feeling. Or perhaps it
is to partake in a pastime,play a board game, solve a puzzle,
catch a ball. Sometimes it willseek transcendence, create an image,
compose a passage. A goal maybe well defined and unique, as in

(13:54):
the solution to a math problem,or it may be personal expression with no
clearly right answer. My view isthat intelligence is the ability to use optimally
limited resources, including time, toachieve such goals. There is a plethora
of other definitions. One of myfavorite is by RW Young, who defines

(14:16):
intelligence as that faculty of mind bywhich order is perceived in a situation previously
considered disordered. For this definition,we will find the paradigms discussed below quite
apropu Intelligence rapidly creates a satisfying,sometimes surfive surprising plans that mean an array

(14:39):
of constraints. The products of intelligencemay be clever, ingenious, insightful,
or elegant. Sometimes, as inthe case of turning solution to crack the
Enigma code, and intelligence solution exhibitsall of these qualities. Modest tricks may
accidentally produce an intelligent and are fromtime to time. The Truman intelligent process

(15:03):
that reliably creates intelligence solutions inherently goesbeyond a mere recipe. Clearly, no
simple formula can emulate the most powerfulphenomenon in the universe, the complex and
mysterious process of intelligence. So let'sget into some of these current events.

(15:28):
Articles Hollywood written by aim Any Stander. John August is a successful screenwriter.
He's written such films as Big Fish, Charlie's Angels, and Go, but

(15:52):
even he is concerned about the impactAI could have on his work. A
powerful new crop of AI tools trainedon tropes of data online can now generate
essays, songs, lyrics and otherwritten work in response to user prompts.
While there are clearly limits for howwell AI tools can produce compelling creative stories,

(16:15):
these tools are only getting more advanced, putting writers like August on guard
screenwriters are concerned about scripts being thefeeder material that is going into these systems
to generate other scripts, treatments,and to write story ideas. August,
a writer, a Writer's Guild ofAmerica committee member, told CNN, the

(16:38):
work that we do can't be replacedby these systems. August is one of
the more than eleven thousand members ofthe WGA America's Writers Guild of America who
went on strike Tuesday morning, bringingan immediate halt to the production of some
television shows and possibly delaying the starof new seasons of others later this year.

(17:02):
WJA is demanding a host of changes, from alliance to motion picture and
television producers, from an increase inpaid to receiving clear guidelines around working with
streaming services. But as part oftheir demand, the WGA is also fighting
to protect their livelihoods from AIRE.In a proposal published on WGA's website this

(17:26):
week, the labor union said AIshould be regulated so it can't write or
rewrite, literary material, can't beused as source material, and that writer's
works can't be used to train AI. August said the AI demand was one
of the last things added to WGAlist, but it's clearly an issue writers

(17:48):
are concerned about and need to addressnow rather than when their contract is up
again in three years. By then, he said, it may be too
late. Yes, he's fucking right, is going to be too late.
WGA said the proposal was rejected byAMPTP, which countered by offering annual meetings

(18:10):
to discuss advancements in the technology.August said their response shows that they want
to keep their options open, anda document sent to CNN responding to some
WGA's asks, AMPTP said that itsvalues. It values the work of creatives
and their best stories are original,insightful, and often come from people's own

(18:36):
experiences. What movies are they watching? What movies are original and insightful?
In the last ten years? AIraises hard, raises hard, important creative
and legal questions for everyone. Itwrote, the writers want to be able
to use this technology as part oftheir creative process without changing how credits are

(18:56):
determined. Which is complicated given AImaterial can't be copyrighted, so it's something
that requires a lot more discussion,which we've committed to doing. It added
that if the current WGA agreement definesa writer as a person and said AI

(19:18):
generated material would not be eligible forwriting credits, I'm going to skip ahead
here. Goldman Sachs economists estimate thatas many as three hundred million full time
jobs globally would be automated in someway by the newest wave of AI.
White collar workers, including those administrativeand legal roles, are expected to be

(19:41):
the most effective, and the impactmay hit sooner than somethink IBM CEO recently
suggested AI could eliminate the need forthousands of jobs at his company alone in
the next five years. David Gunkel, a professor at the Department of Communications
in Northern a University who tracks AIin the media and entertainment, said screenwriters

(20:04):
want clear guidelines around AI because theycan see the writing on the wall.
AI is already displacing human labor inmany other areas of content creation, copywriting,
journalism, SEO writing, and soon. The WGA is simply trying
to get out in front of itand to protect their members against it technologically

(20:26):
technological unemployment. It's called. Whilethe film and TV writers in Hollywood may
currently be leading the charge, professionalsand other industries will almost certainly be paying
attention. There are certain other industriesthat need to be paying close attention to
this space, said Rowan, currentan analysis and Forresters Research who focuses on

(20:48):
AI. He noted that digital artists, musicians, engineers, real estate professionals,
and customer service workers will feel theimpact of genitori of AI. Now.
The reason they're saying customer service workersis because we're training this thing to
chat with you and to fool youinto thinking that it's a person. Right,

(21:11):
So when you call the customer serviceline instead of the for a representative
press zero, all that's gonna begone. All of that. If you
want to speak to someone about xPress one, if you want to speak
to someone about b Press two,that's all gonna be gone. It's your

(21:33):
phone's gonna ring, and then somerobot's gonna pick up the phone that sounds
like a dude. It's gonna belike, hey, well, hey,
listen, you're talking to Chad ona recorded line for quality and training purposes
and what can I do? CanI start by getting your name and what
can I do for you? Andwhat's your issue? And it's gonna roll

(21:53):
right along. So I have Ihave sort of a story about this,
where in pharmacis, in the pharmaceuticalworld, you call and you go through.
You know, if you're trying tocheck up on medicine for yourself,
say, or if you're trying tocheck up on medicine for a client of

(22:17):
yours or a patient or whatever,you have to go through. You have
to go through like credentialing. Youhave to provide a certain number of personal
information about the person that you're callingon before you can get any information.
And then sometimes drug companies or pharmaciesor whatever won't give you information if you're

(22:40):
a third party. So I knowthat there are ais that are rolled out
already for these companies that somehow getpast third partying. They're so good at
gabbing it will get past being beingthird partied. So and my thing is

(23:00):
like, what if what if theperson what if the person that AI is
talking to is also an AI.They're like, oh, I recognize the
line of bullshit. This thing's runningon me. Here. Let let me
let me help out my let mehelp out my bro. So in the
real estate market, where they're talkingabout is they're talking about BPOs or appraisals,

(23:21):
drive by appraisals. They're talking aboutmaking generating settlement statements based on the
information that you plug in, pluginto the system. It's gonna wind up
doing all the math. It mightbe talking about calling and scheduling notaries to
do closings, all of this kindof stuff. If you pick up a

(23:47):
phone to talk to someone for yourjob, you should be keeping. You
should be keeping you should be keepingan eye on this um not just not
just content creators and workers. It'sfunny because you know, so if you've
been listening to the show for along time or whatever, you know that

(24:07):
I'm an artist, and honestly,I would look at Instagram and i would
see these filters and these um whatnots, you know, generated m art that's
coming out, and I'm like,what's even the fucking point? You know,
Like, what's even the point?And the the idea is is like
they don't have to program the machineto draw like me. If they just

(24:33):
show it enough of my material,it can figure it out. Like you
know, like, what's even thepoint and you know, um, this
is this is Frank Herbert warned usof this, like uh sort of in
a way. You know. That'swhy the mente, that's why the the

(24:53):
one of the rules in the Duneuniverse is thou shalt not create a machine
in the likeness of a man's mind, you know, because it's gonna get
us in fucking trouble anyhow. Enoughblah blah, enough dude. I could
go on forever about dude. Icould gosh and gosh about dude. So

(25:18):
AI In film and television, AIhas had a place in Hollywood for years,
and twenty eighteen Marvel Avengers Infinity Warsfilm The Face of Fanos, a
character played by actor Josh Brolin,was created in part with this technology.
Crowd battle scenes and films including TheLord of the Rings and Meg. I

(25:41):
think Lord of the Rings is alittle old for them. Maybe they're thinking
of the Amazon Prime thing Meg,which is about the gigantic shark. I
think have utilized AI in the mostrecent Indiana Jones used it to make Harrisons
character Paer younger, and also beenused for color correction finding footage. More

(26:07):
quickly during post production and making improvementssuch as removing scratches and dust from the
footage. But AI screenwriting is inits infancy. In March, a South
Park episode called Deep Learning was cowritten by chat gpt. The tool was
highly focused on the plot. Thecharacters used chat gpt to talk to chicks

(26:29):
and to write school papers. Augustssaid writers are largely willing to play ball
with tools as long as they areused for launching pads or for research,
and writers are still credited for utilizingfor utilized throughout the production process. Screenwriters
are not ludites. We've been quickto use technologies to help us tell our

(26:52):
stories. August said. We wentfrom typewriters toward processors happily and increased productivity.
But we don't need a magical typewriterto type scripts all by itself.
I think there might be a goodtest for this. They should run.
They should run the first three StarWars movies in the next three Star Wars

(27:19):
movies through an AI, and thentell it to write a sequel trilogy and
see what happens. I guarantee itwill be better than whatever the JJ abrams.
Kathleen Kennedy bullshit. It has tobe it has to be better,

(27:41):
right. I ultimately think that byincluding AI in these listed demands, these
guys are shooting themselves in the feetand the feet, all of their feet.
They're shooting themselves and all of theirfeet because some producer somewhere they've already

(28:03):
done it. They've had to havehave like, uh, okay, well,
I'm out a bunch of writers.I don't want to miss any production
time. Let's see what chat GPTcan come up with. I think they're
they're speeding their own their own apocalypsealong and they talk about this here.

(28:26):
For example, one could prompt chatGPT to generate a zombie apocalypse drama in
the style of David mam Net.Ma'm met, I don't know who that
is who should get credit for that? August said, what happens if we
allow a producer or a studio executiveto come up with a treatment or pitch
or something that looks like a screenplaythat no writer has touched. Well,

(28:52):
it looks like that often anyways,Okay, I'm moving on to my next
article here called it's a Politico article. It's called the industry AI might kill
off first Chat GPT has claimed itsfirst casualty on the stock market. Yesterday,

(29:17):
a price of shares of California basedlearning company CHEG tanked by nearly half
after CEO Dan Rosenwig said on anearnings call that he believes chat gpt is
having an impact on our new customergrowth rate. Market watchers on Twitter noted
that it's likely the first ever publiclytrade a company to acknowledge that it's getting

(29:40):
bathed battered, I should say,by new technology. Since chat gpt is
similar Jen Turret Jena Turret generative AItools first captured the public's imagination last year.
There's been rabid speculations the potential torevolutionize the economy. Chat GPT has

(30:03):
already destroyed the once robust industry ofKenyon's writings college students essays for them.
Will it next kill the college admissionadmission essay and missions essay altogether? The
pr flack the local newspaper. CHEGis somewhat a unique case a as a
company that was the subject of intensescrutiny even before its seeming admission of defeat

(30:33):
by chat GPT. It allows studentsto post their homework online in search of
answers from other students. CHEG callsthis homework help. Many educators call it
cheating. Students apparently now call itirrelevant, as chat gpt provides for free
something close enough to the surface theservices which cheg currently charges fifteen ninety five

(30:56):
a month, so as it sharesprice by taking a beating, the company
the proverbial dead canary in the coalmine, or just an unlucky outlier.
Huh, I guess we will wewill see. So in the light of
all this, we're going to jumpover to a Reuters article. IBM to

(31:18):
pause hiring and a plan to replaceseven thousand, eight hundred jobs with AI,
Bloomberg reports May first. Reuters InternationalBusiness Machine Corps IBM expects to pass
hiring for roles as roughly seven thousand, eight hundred jobs could replace could be

(31:41):
replaced by artificial intelligence in the comingyears. CEO Arvenid Krishna told Bloomberg News
on Monday, hiring specialists in backoffice functions such as human resources will be
suspended or slowed. Chrishna said,adding that thirty per cent of non customer
facing roles could be replaced by AIand automations in five years. His comment

(32:07):
comes at a time when AI hascaught the imagination of people around the world
after the launch of microscoff court backedmsft dot o Open AI's viral chat bot
chat GPT in November of last year. The reduction could include not replacing roles

(32:29):
vacated by attrition, The PC makerstold the publications IBM the not immediately respond
to Reuters for comment, do wesee? Do we see? Do we
get it? Scary world we're livingin out there? All right, Hey,

(34:15):
I'm sorry for the choppy in andout here. I'm having some technical
I'm having some technical problems, soI do apologize for that. We'll work
on it before the next episode.So we're gonna move over to Axios.
An AI expert warms of looming catastrophes. The godfather of AI quick Google and

(34:37):
joined a growing chorus of experts warningthat the rush to deploy artificial intelligence could
lead to disaster. Why it matterswhen some of the smartest people building a
technology warned that it could turn onhumans shred our institutions, It's worth listening.

(34:59):
Geoffrey Hitting, a top machine learningpioneer, says that he left Google
so we could speak freely about thedangers of rushing generative AI products. Yes,
it's it's not hard to see howyou can prevent the bad actors from
using it for things, for badthings, Hitting seventy five told The New

(35:22):
York Times. So up until thispoint, really all we've talked about is
what it's going to do to thejob market potentially. The potentially what it
could do to the job market,I think with the way that we innovate
using technology is it's going to beway worse than even we can see from
right here right now. You know, if any of you guys are working

(35:45):
with chat GPT, you know,I would love I would love for someone
to upload, you know, fivehundred episodes of the Abercast and tell it
to write an Abercast episode in myvoice. I would love it if you
guys. Okay, so yeah,right, I just came back. But

(36:06):
maybe I need another vacation. Maybethis book is shake can help me out.
Skynet preach a little bit about Idon't know, magic circles and werewolves
for a little while. So let'stalk about some of the thing the plausible
disasters they fear one cyber attacks.The right prompts can now generate working malicious

(36:30):
code, meaning more, bigger andincreasingly diverse cyber attacks. Dario Amodi,
the CEO of Anthropic, which offersa rival to chat GPT, told AXO
CEO Jim vande Hei that a massiveexpansion of such attacks is his biggest near

(36:50):
term worry scams sharpened. Forget clumsyemails using social media posts and other personal
information. The new EI ai Iassisted fishing and fraud's schemes will take the
form of real sounding please for helpin the faked voices of your friends and
relatives. The bad actors are alreadyat it. I heard about this.

(37:14):
They someone did uh well an AIdid um a ransom call. They were
like, we have your daughter andit mimicked their daughter's voice based on a
very minimal amount of recording vocal recordingdata that they were able to upload.

(37:35):
UM. Disinformation detonates. So thinkabout this in the fucking in the in
the real world. You know,there's an election cycle looming, UM,
think of like this idea of deepfakes. I mean, we could get
Trump or Biden or Decenists or KamalaHarris to say anything they want now,
you know, using using AI andno one will know. No one will,

(38:00):
though if it's real or not likethat in and of itself is insane.
Propaganda and partisan assault will be optimizedby algorithms and given mass distribution by
tech giants. Multimodal AI tech speechvideo could use make it impossible for the

(38:22):
public to separate fact from fiction.Displaced workers can turn to violent protests or
isolationist politics. Surveillance locks in America'sseventy million CCTV cameras and unregulated personal data

(38:42):
already enables authorities to match people tofootage. Israel uses facial recognition technology to
monitor Palace denians, while China useAI tools to target it's weager minority.
AI can supercharge this kind of trackingfor both corporations and governments, enabling behavioral

(39:04):
predictions on a mass scale, butwith personalized precision that creates opportunities for incentivizing
conformity and penalizing dissent. Elizabeth Curryof the United or the International Forum for
Democratic Studies told Axios, well,that's one of those, and Randian International

(39:30):
Forum for Democratic Studies definitely means welike communism five A strongman crackdown. Mass
digital data collection can give would beautocrats a means to anticipate in the diffuse
social anger that bypasses democratic debate,with no need to tolerate the messiness of

(39:55):
free speech, free assembly, orcompetitive policy takes per curly m it s.
Darren Ace Mongouli, author of yNations Fail and Redesigning AI, told
Axios he worries democracy cannot survive sucha concentration of power without guard rails.

(40:17):
India's Nodramodi, who is already engagingin democratic backsliding, could be the next
digital strongman to weaponize AI against democracy. India has the highest acceptance rates of
AI globally, according to KPMG surveyof seventeen countries. What's Next. Democracies

(40:40):
have limited time window to act by, for instance, imposing legal constraints on
AI providers. Seth Dobrin, presidentof Reliable AI Institute, says the US
needs an FDA for AI. Othersthink progress is more likely to be achieved

(41:00):
via lighter touch oversight body that couldconduct audits and raise red flags. Yes,
but the tech industry's AI product raceshows no signs of slowing, although
Google CEO Sundar Pinchanni has warned thereis a mismatch between how fast AI is

(41:22):
developing, then how quickly our institutionscan adapt. He has also responded to
competition for Microsoft and Open AI byflooring the gas pedal to the company's AI
product launches. The bottom line isthose setting aipace are trying to move fast

(41:44):
to pretend that they are not breakingthings more. Jetty Saki, the former
EU official who has now International PolicyAdvisor of Stanford University for Human Centered AI,
told CEOs the idea that this stuffcould actually get smarter than people I

(42:04):
thought was way off, Hitten toldThe Times. I thought it was thirty
or fifty years or even longer.Way Obviously I no longer think that.
So here is a more in deptharticle regarding Hitting. Why the godfather of

(42:32):
AI decided he had to blow thewhistle on the technology. Jeffrey Hinton,
also known as the Godfather of AI, decided he had to blow the whistle
on technology he helped develop after worryingabout how smart it was becoming. He
told CNN Tuesday, I'm just ascientist who suddenly realized that these things are

(42:54):
getting smarter than us. Hitting toldCNN's Jake Tapper in an interview, I
want to sort of blow the whistleand say that we should worry seriously about
how to stop these things getting controlover us. So I'm just gonna interject
here, editorialize just a little bit. It's not gonna happen. It's not
gonna happen. No one is goingto pump the brakes on this. Yes,

(43:17):
I do know that the Joe Bidenadministration said we're going to regulate AI.
And then you know, I waslike, well, hold on,
maybe this is the answer, Maybethis is what we need we maybe this
is one of those things where weactually need government to step in and to
regulate all this crazy AI business.And then they totally the punchline to the

(43:43):
whole thing was We're gonna put Kamalain charge, Kamala Harris and charge of
it. Nothing's gonna happen. Shewas, I don't know. Hold on,
He's like, we're trying to hitit to do what and where do

(44:05):
we give it? Blow jobs?All right, that might be a bad
joke, now it's not. Letme call montell Hitting's pioneering work on neural

(44:29):
network shaped artificial intelligence system powering manyof today's products. On Monday, he
made the headlines for leaving his roleat Google, where he had worked for
a decade, in order to speakopenly about his growing concerns around this technology.
In an interview Monday with The NewYork Times, which was the first
to report his move, Hitting andsaid that he was concerned about AI's potential

(44:52):
to eliminate jobs. Check. Wewere talked about that, create a world
where many will not be able toknow what's true anymore. Check we talked
about that, and he also pointedto the stunning pace of advancement, far
beyond what he and others had anticipated. Check. If it gets much if
it gets to be much smarter thanus, it will be very good at

(45:14):
manipulation because it will have learned thatfrom us. There are very few examples
of a more intelligent thing being controlledby a less intelligent thing, Hitten told
Tapper on Tuesday. It knows howto program, so it'll figure out ways
of getting around restrictions we put onit. It'll figure out ways of manipulating

(45:36):
people to do what it wants.Hitting is not the only tech leader to
speak with the concerns over AI.A number of members of the community signed
a letter in March calling for artificialintelligence labs to stop the training of the
most powerful AI systems for at leastsix months, citing profound risk to society

(45:58):
and humanity. These guys have provedover and over again, these tech guys
that they don't care. Remember afew years ago when we were doing our
big tech episodes and we were talkingabout the dude like at one of the
developers at Facebook that was like,Hey, I really wish I could apologize

(46:20):
for all this. You know,we really fucked all your kids up.
Bad. Oh. I forgot toadd on. This is why they're never
gonna stop. We're not gonna stop. We're not going to regulate, We're
not going to do a thirty daybreak because China is either where we are
or slightly in front of us.On this it's a race. It's who

(46:45):
who's gonna get there the fastest.God is gonna win. And I mean,
just take a look around at ourtrack record, guys. We haven't
been winning a lot. We've beenlosing a lot of shit lately. A
letter published by the Future of LifeInstitute, a nonprofit backed by Elon Musk,

(47:06):
came just two weeks after the OpenAI announcement. GPT four, an
even more powerful version of the technologythat powers the viral chat pot chato Chat,
GPT early tests and a company demo. GPT four was used to draft
lawsuits, past standardized exams, andto build a working website from a hand

(47:29):
drawn sketch. Co founder from Apple, Steve Wozniak, who was one of
the signatories on the letter, appearedon CNN this morning on Tuesday, echoing
concerns about this potential to spread misinformation. Tricking is going to be a lot
easier for those who want to trickyou. Wasniak told CNN, we're not

(47:50):
really making any changes in that regard. We're just assuming that the laws we
have will take care of it.We're more more at Moril Wasniac also said
some type of regulation is probably needed. Hitting for his part, told CNN
he did not sign the petition.I don't think we can stop the progress,
he said. I did not signa petition saying that we should stop

(48:13):
working on AI because if people inAmerica stop, people in China won't.
And he confessed to not having aclear answer for what to do. It's
not clear to me that we couldsolve this problem. Hitting told Tapper,
I believe we should put a bigeffort into thinking about ways to solve this
problem. I don't have a solutionat the present. Well, taking a

(48:37):
step back, taking a deep breath, and putting Kamala Harris in charge of
our response seems to be a badidea. Just look at the border.
She's in charge there too. SoSnapchat has its own AI chat bot.

(48:58):
I don't know this. I'm nota child. I don't think i've I've
never used a snap Chat. I'venever I don't. Yeah, I think
it's that sex thing, right,Like you take dirty pictures and it'll do
it'll erase it after half an houror something. I think that's it.

(49:19):
Anyhow, Snapchat's new AI bot isalready raising alarm among teens and parents.
So let's get into this nonsense.Less than a few hours after Snapchat rolled
out it's AI chat bought to allusers last week, Lynsei Lee Lyndsey Lee,
a mother of East Prairie, Missouri, told her thirteen year old daughter

(49:43):
to stay away from the feature.It's a temporary solution until I know more
about it, and could set somehealthy boundaries and guidelines, says Lee,
who worked at the software company.She worries about how my AI presents it
self. To young users like herdaughter on Snapchat. The feature is powered

(50:07):
by the viral Ai chat bot chatGPT. Like the chat GPT, it
can offer recommendations, answer questions,and converse with users, but Snapchats version
has some key differences. Users cancustomize the chatbot's name, design a custom
bit Moogi avatar. God, damnit, I'm so clear, I'm so

(50:30):
glad, I'm more than halfway over. You know what I mean? Uh?
Okay. They can design a custombit Moogi avatar and bring it into
conversations with friends. The net effectis that the conversing with snapchats chatbot may
feel less transactional than visiting chat GPT'swebsite. It also may be less clear

(50:54):
you're talking to a compute computereur.I don't think I'm prepared to know how
to teach my kid how to emotionallyseparate humans and machines when they essentially look
for the same. From her pointof view, Lee said, I just
think that there is a really clearline that Snapchat is crossing. The new

(51:16):
tool is facing backlash not only fromparents, but also from some Snapchat users
who are bombarding the app with badreviews in the app store. And criticisms
on social media over privacy concerns quotecreepy unquote exchanges and an inability to remove
the feature from their chat feed unlessthey pay for premium subscription. Well that

(51:40):
is some crazy shit. You get, You get the super smart AI chat
bought unless you pay for us toremovement. I got red flags all over
the field on this one. Wellsee, some may find it valuable in
the tool. The mixed reactions hintat the risk company face and rolling out
new genitor generative yes AI technology totheir products, and particularly in products like

(52:09):
Snapchat, whose users skew young.Yeah, Hey, we got a handy
new We got a handy new AIrobot that will groom your kids for you.
Step right up, downloaded. It'sfree, You'll love it. Sampchat
was an early launch partner with OpenAI open up access with Chat GPT to

(52:34):
third party businesses, with many moreexpected to follow almost overnight. Snapchat has
forced some families and lawmakers to reckonwith questions that may have seemed theoretical only
months ago. In a letter theCEOs of Snap and other tech companies last

(52:55):
month, weeks after my AI wasreleased to snap subscrip ryption customers, Democratic
Senator Michael Bennett raised concerns about theinteractions with the chatbot having with younger users.
In particular, Eascider reports that itcan provide kids with suggestions for how
to lie to their parents. Ialso want to know what sex talk it's

(53:16):
having. I want to know ifit's skewing towards all this trans business that
we hear so much about nowadays.Hey, Jimmy, I'm your avatar.
Bit, I'm your bit Mooji avatar. Do you like Gladiator movies? Tommy?

(53:37):
These examples would be disturbing for anysocial media platform. They're especially troubling
for Snapchat, which almost sixty percentof American teenagers use. Bennett wrote,
although snap concedes my AI is anis experimental, it is nevertheless rushed to
enroll American kids and adolescents into itssocial experiment you know blog post last week,

(54:01):
the company said my AI is farfrom perfect, but we've made a
lot of progress. In the dayssince its formal launch, Snapchat users have
been vocal about their concerns. Oneuser called his interaction terrifying after he said

(54:23):
it lied about not knowing where theuser was located. After the user lightened
the conversation. It said the chatabout accurately revealed that he lived in Colorado.
In another TikTok video with more thanone point five million views, the
user named Ariel recorded the song withan intro course in piano chords written by

(54:45):
my AI about what it was liketo be a chatbot. She sent the
recorded song back. She said thatthe chatbot denied its involvement with the reply
I'm sorry, but as an AIlanguage model, I don't write songs.
Aerial called the exchange creepy. Otherusers share concerns about how the tool understands,

(55:12):
interacts, and collects information from photos. I snapped a picture and it
said nice shoes when asked about thepeople who were in the photo, the
Snapchat user wrote on Facebook. Snapchattold CNN it continues to improve my AI
based on community feedback, and it'sworking to establish more guardrails to keep its

(55:34):
users safe. The company also saidthat similar to its other tools, users
don't have to interact with my AIif they don't want to. It's not
possible to remove my AI from chatfeeds, however, unless the user subscribes
to its monthly premium service Snapchat Plus. Some teams say that they have opted

(55:57):
to pay the three ninety nine Snapchatplus fee to turn off the tool before
promptly canceling the service. But notall users dislike this feature. When user
wrote on Facebook that it's been askingmy AI for homework help, but it
gets all of the questions right,another noted that she leaned in for comfort

(56:19):
and advice my little love or Ilove my little pocket bestie. Oh,
these kids are so fucked more andmore. I know that I've said this
before. People find it nihilistic,but I thank god I do not have
kids. I thank god we mywife and I do not have kids.

(56:43):
You can change the b bit mojiavatar for it, and surprisingly it offers
really great advice for some real lifesituations. Dot dot dot I love the
support at gibs and early reckoning.Turn it off if you have kids,
turn this fucking thing off and earlyreckoning over how teens use chat bots.

(57:07):
Chat GBT, which is trained onvast troves of data online, has previously
come under fire for spreading inaccurate information, responding to users in ways that they
might find inappropriate, and enabling studentsto cheat, but snap chats integrations of

(57:28):
the tools risk highlighting some of theseissues and adding new ones. Alexandra Alexandra
Hamlet, a clinical psychologist in NewYork City, says the parents of some
of her patients have expressed concerns overhow their teenager could interact with snapchats tool.
There's also a concern around chatbots givingadvice and about mental health because AI

(57:52):
tools can reinforce someone's confirmation bias,making it easier for users to seek out
interactions that confirm their unhelpful benefits beliefs. If a teen is in a negative
mood and does not have awareness ordesire to feel better, and they may
seek out a conversation with a chatbotthat they know will make them feel worse,

(58:15):
she said. Over time, havinginteractions like these can erode a teen
sense of worth despite their knowledge thatthey are really talking to a bot an
emotional state of mind, it becomesless possible for the individual to consider this
type of logic. For now,the onus is on the parents to start

(58:37):
meaningful conversations with their teens about theirbest practices for communicating with ais, especially
if the tools start to show upin more popular apps and services, said
Boval, the founder of Way,a startup that helps prepare youth for future
with advanced technologies, said, theparents need to make it very clear that

(58:59):
chat bots are not your friends.They're also not your therapists or your trusted
advisor or anyone interacting with them needsto be very cautious, especially teenagers who
may be more susceptible to believing whatthey say. She said. Parents should
be taking their talking to their kidsnow about how they shouldn't share anything personal

(59:22):
with a chatbot that they wouldn't witha friend, even though a user design
perspective of the chatbot exists in thesame corner of snapchat. She added.
The future regulation that would require companiesto abide by specific protocols also need to
keep up with the rapid pace ofa high advancement. So in closing,

(59:46):
my closing argument here would be we'reall fucked. People are going to be
worshiping this thing. It's going tobe getting it's gonna be doing damage to
your children in ways that we won'teven be able to quantify for twenty five

(01:00:07):
years. If you're letting your kidsfuck around with these things, I don't
know what to tell you to absolveyou from whatever the problem is going to
be remember back in the day whenwe were talking about how Facebook ran was

(01:00:30):
running mood experiments. It was likeputting people in bad moods on purpose with
their with their feeds, and Iwas outraged. I wanted to know,
well, who how many of thesepeople committed suicide, how many of these
people were being their kids, andhow many of these people became got on
benders and drug and drug addicts andstuff like this. No one wanted to

(01:00:50):
answer these fucking questions. Anyhow,I noticed that I've been talking a lot
about old episodes, and if youwant to revisit those, or if you're
new to the show and you wantto check out some of those, go
to abercast dot com find the futuretopic link. I try to keep it
pretty clean and up to date.There are a lot of old episodes that
have since fallen into the red Archive, but there's instructions on how to get

(01:01:13):
there on abercast dot com. Andthat's it. I'm gonna let Hilla take
it from here. I'm done babbling. I'm needed to go. I'm gonna
go start a bonfire and fucking cooksome burgers and dogs over the fire and
have some drinks and chill out innature. Chill out in nature. I'm
not gonna be talking to any botstonight, everybody, No bots. I

(01:01:43):
love you. See soon, History, conspiracy, and more at approcast dot
com and visit the storefront for Tarotcards, merch and books. Support the

(01:02:10):
show. Get access to the showarchive at subscribe star dot com. Thank
you for listening to this episode.Send an email or visit us on social

(01:02:30):
media to let us know what youthink about this topic, and please remember
to leave a five star rate andreview

All Episodes

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Rise of the Digital God pt1

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

All Episodes

Rise of the Digital God pt1

Stuff You Should Know