Exploring Artificial Intelligence: Energy Requirements, Potential Lawsuits, and Impact on Journalism
November 4, 2023 • 40 mins
Are you ready to dive deep into the complex world of artificial intelligence? As we all know, AI is a double-edged sword, bringing both innovation and challenges. Today, we're going to unravel the power requirements of AI technologies and how they're pushing for cleaner energy sources. We're also going to chat about the New York Times' potential lawsuit against OpenAI and Chat GPT, which is stirring up legal implications worldwide. If that wasn't enough, we'll also cover the recent changes in the New York Times' terms of service that could disrupt Chat GPT's knowledge base. Stick around, because you won't want to miss this!
Transitioning to a heavier topic, we'll address the elephant in the room: AI's environmental footprint. With climate change making headlines daily, we're going to discuss the role AI technologies, server farms, and Bitcoin mining play in this global crisis. As we pivot to our final discussion point, we'll contemplate the implications of AI becoming a primary source of information and the potential consequences of relying less on traditional search engines. This episode is going to open your eyes to AI's undeniable influence on our lives and the future of our world. So, buckle up for a thrilling ride through the realm of AI!
Let's get into it!
Follow us!
Email us: TheCatchupCast@Gmail.com
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
So we're talking this week about the struggles of AI.
Right, we're back on AI train,a thing that has to find the
season of catch up podcasts,right, mm?
Hmm, so we're talking about theold AI issues really because
one of them they're both bigissues One of them seems to be
(00:24):
more immediate than the other.
First of all, chat, gpt andother big AI like it.
Right, they draw a lot of powerbecause they have multiple,
multiple computers forprocessing that provide
processing horsepower for thisAI, right, mm?
hmm, and that's a big dream onpower, big dream of resources.
(00:47):
So we're going to talk aboutthat.
We're also going to talk aboutthe New York Times possibly I
want to report accurate releasesfor target matter journalistic
media possibly is gearing up tosue AI, specifically open AI,
(01:08):
sorry, and chat GPT at thatwould be.
You know, potentially whatthey're saying is we'll come
back to this further.
I'm not explaining this wellall right now, but we'll come
back to this further.
What potentially could happenis this lawsuit, if it happens,
could go to a federal judge thatapparently, if it favors the
(01:31):
New York Times, chat GPT wouldhave to scrub all of its
knowledge and start over, whichis crazy it is, and so it's
definitely a good talking point.
We want to discuss that also.
A side note the New York Timesactually changed its terms of
service.
That prevents any AI from, youknow, scrubbing or not scrubbing
(01:58):
, but you know, gatheringinformation from that site.
You know, I don't really knowquite how you prevent that.
I guess you have to be in theAPI, right, because it would be.
Speaker 2 (02:13):
It wouldn't be in the APIs.
It would be similar to, kind oflike, what has happened with
the social media platformformerly known as Twitter.
Speaker 1 (02:25):
I know as X.
Speaker 2 (02:26):
Exactly when they.
They changed the API ruling.
I think Reddit also didsomething similar, where they
changed how their API is used.
So, yeah, I believe you can doit in the API and so that way
the AI is not able to easily goin there and grab specific data
(02:49):
points and stuff.
Speaker 1 (02:52):
Right, and what is API stand for?
Speaker 2 (02:55):
You know, I don't know, we're going to have to
Google that one.
I actually don't know that one.
Let's find out, let's let'spull it up I see it all the time
.
Speaker 1 (03:07):
I see it all the time because I work with these
social media platforms a lot.
I've always wanted you know youwant you see it all the time
because, oh yeah, I do.
Multiple reasons, yeah,application program interface.
Speaker 2 (03:23):
Programming.
Speaker 1 (03:25):
Programming interface .
You're right, seems a littleredundant.
Not really it's good.
So yeah, that is going to beinteresting, because I do know.
So chat GPT has been withoutInternet surfing abilities for
over a month now.
(03:45):
They seemingly have not beenable to get the issue that
they've had fixed, and sothey've been without Internet
abilities for, like I said, overa month.
Speaker 2 (03:56):
Or at least that plug in.
You know, is that right Plug in?
Speaker 1 (03:59):
Yeah, exactly, but you know I can't search anything
that's current on chat GPTright now.
But, yeah, I totally understandwhat you're saying.
Yes, but one thing I asked, Ithink I wanted to look at
(04:20):
Something I'd posted before,right, because it was posted
something for my music.
And so I wanted to look atsomething I posted before and it
said sorry, but you know,facebook's API, which is also
the same as Instagram's,prevents me from searching for
this stuff.
And I was like, hey, yeah, mm,hmm, so that's kind of the
reason for that.
(04:41):
So yeah, that's interesting, man.
We got a lot to talk about.
We sure do, man.
So I just say we go ahead andget into it.
What's going on?
Everybody, I'm John and I'mDenison, and this is the catch
up.
All right, before we get backinto our main topic.
(05:23):
I just wanted to discuss thethree best ways.
Let me say, when I saydiscussed, right, I just want
you to sit down on the couch,talk with me for a second.
I'm just trying to connect withyou, man, exactly, you and I
have been going our separateways for so long.
You know I try to sit down withconversation.
You want to go play video games?
So, listener, whoever you are.
(05:46):
This was not directed at Denison, in case you were confused.
The listener, I just want tohave a conversation with you,
man.
I want to let you know thethree best ways to support this
show.
You know, there they are flyingacross your screen.
Number one oh man, it's alwaysthe build up and in case RSX
(06:09):
type are nice dude, pretty slick.
They were like really cool ifit was red.
You know, for historicalreasons, number one is Please
leave us rain in review.
Wherever you're listening,wherever you're watching.
It really helps us to grow,helps us to know what we can do
to do better or what we'realready doing well, but it also
(06:32):
gets us down from more peopletoo.
Speaker 2 (06:33):
So it's a big help.
Speaker 1 (06:37):
Please just take a few seconds to leave five stars
or whatever you're feeling,right?
I feel like I preach this weekafter week.
I'm really serious.
It would help us a lot.
You know I'm not going to be afan of it.
It would help us a lot, youknow.
So please just take a fewseconds to help us out.
Number two is give us a follow.
Oh my gosh, oh, I spoke toosoon.
(07:00):
There's a red model right there.
Dang, that was crazy.
I shouldn't have.
Should have held my, held mytongue.
Number two is follow us onFacebook or on YouTube at our
podcast name and we go liveevery Thursday night talking
about our topics in real time.
(07:20):
If you don't already watch usthis way, it allows you to jump
in the discussion with us inreal time and leave your
comments, and we want to have adiscussion with all of you, and
you're getting the format, therecord version like that in real
time.
But it's more.
You know, as cool as that is, wewant to be able to have a
really great conversation withyou and expand this discussion
(07:44):
as a whole.
And number three oh man, youreally pulled that PS5
controller out of the bag.
All right, there is your threeright there.
No way it's an Xbox controller.
I was, I was duped, I wasbamboozled.
All right, we're going to showyou a little Something,
(08:08):
something you might say.
Speaker 2 (08:10):
This is not on show and tell.
Speaker 1 (08:13):
Yeah, this is not on a shirt, right?
Just picture it.
Let's go ahead and get into itwith a nice big pair of
headphones.
You know, I'm saying that issomething Looks real clean,
right?
So that's something that youcould have wrapped around your
body, you know round your nogginman.
(08:34):
Not to make it sound weird oranything, but I think actually I
can give you all a betterexample.
Give me two, give me twoseconds.
You store Perfect.
Speaker 2 (08:48):
One, two.
Speaker 1 (08:50):
Hey, I wasn't ready.
I thought I was.
Speaker 2 (08:53):
Time out man.
Speaker 1 (08:55):
Yeah, I know time flies when you're trying to show
off your merch.
Speaker 2 (08:58):
You are man.
I mean, you're right, you'reright, time does fly, but it is
what it is.
Yeah, yeah, we're out heretrying to show the people you
know show and tell, just likeyou know when you were, when you
were younger, you know, and youwent to class and you always
brought your favorite toy rightand you got to bring it up to
the front of the class and showthem your favorite toy right.
(09:20):
We were just trying to show ourfavorite clothing that we have,
you know, our favorite merch Toour audience, you know, so you
guys can check it out and, youknow, if you feel so inclined,
go ahead and snack.
You know a couple of things.
Speaker 1 (09:37):
And I just want to say too I feel so disappointed
in myself because this is not onour shop page and it needs to
be, you know, Mm.
Hmm, so let's see no, we don'twant black.
You can't see the design.
Speaker 2 (09:59):
No, you can't, no, you can't.
So, it does look good, no itdoes.
Speaker 1 (10:04):
We'll just, we'll leave it at that.
It looks clean, because I forsome reason can't.
Oh, there we go, let's show itto the people.
Speaker 2 (10:14):
Show it to the people .
Speaker 1 (10:15):
All right guys.
Sorry for the long wait on this.
I forgot how to use the websitethat we do this on Little
little gray action, right?
Just a standard gray.
I know a lot of people likegray around here.
Good hard work.
Look are you sharing.
Speaker 2 (10:30):
There we go, I can see that.
That is pretty done clean.
Speaker 1 (10:35):
It's actually me modeling it, you know.
Speaker 2 (10:38):
And you grew out your beard a lot.
Speaker 1 (10:41):
Yeah, yeah me.
Me not being able to is just apart of my character.
Dude, I actually have a fullbeard.
Speaker 2 (10:48):
I see I see Full black beard.
Yeah, yeah, I tend to, I tendto, you had to dye your hair as
well.
Speaker 1 (10:59):
That's.
I can't leave it.
It's all about the lighting,you know.
Ok, ok, that's interestinglighting, hey, but you didn't
ask about the gains, though.
Let's go.
Yeah, so you know, good, cleanshirt.
I'll get this added to our shopfor you.
I think you'll like this one.
This is the guaranteed Favorite, for sure, without doubt.
(11:23):
So please check that out.
That's over at our shop.
By the time you hear this,you'll be at our shop in the
link below.
So, yeah, please check that out.
I know you'll like it.
All right, let's get into thistopic here, this here topic.
Why don't you start us off, man, talking about the AI and the
(11:48):
power that's required for it,right?
Speaker 2 (11:50):
Yeah, of course.
So, as you all may know, wehave been going through a
sweltering, hot summer, you know, and throughout all of that I'm
sure everyone's kind of heardof, like you know, this is
becoming a more a new normal, asit were, due to climate change,
(12:13):
you know, just because of howmany or how much CO2 that we're
pumping into the atmosphere andall the all the things that come
with that.
Speaker 1 (12:25):
Right, which I actually have a question for you
, right, mm-hmm, saturday.
What's your high?
Speaker 2 (12:36):
Let's see Texas High in Texas.
Let's see what we got here.
Speaker 1 (12:42):
Give me a second here , the and surprising me, hours
on Saturday is one oh seven.
Speaker 2 (13:04):
Well, that's crazy.
Speaker 1 (13:06):
Sunday it's one oh seven and Monday is really crazy
and Monday it's one oh five.
Tuesday one oh three.
Wednesday one oh one.
Thursday one oh one man, itdoesn't drop below a hundred
till a week from tomorrow that'scrazy, it's not, it's right.
Speaker 2 (13:24):
Yeah, I mean, we've been in the hundreds for a while
, man, we are expected to get alittle bit of a cold front here,
though you know good old coldfront.
It's going to be 97 degrees,man, cold front.
Gotta get the jacket out.
Speaker 1 (13:39):
I'm about to say you got that north face ready, man,
mm-hmm, yeah.
So I just wanted.
I was curious and brought thatup because I'm feeling it.
I'm very much feeling it, man.
We've already had a couple ofone.
Oh, seven days up here wealready had 10 days straight of
100 plus degrees.
You know I'm saying so mm-hmm.
I feel that.
(14:00):
So anyway, go ahead, sorry noworries.
Speaker 2 (14:03):
So, yeah, I was really gonna say is you know,
since, as everyone said, youknow, due to these changes in
our the amount of CO2 that wehave in our atmosphere, as well
as some of the changes that aregoing on and how that is
affecting and heating us up alot of pressure has been started
(14:28):
to grow on finding cleanerenergy sources and reducing the
energy that we do consume andstuff like that.
And all of that being said goestowards what we're, you know,
one of the topics that we'regoing to be talking about, which
is chat, gpt or even other.
(14:49):
Ai is like chat GPT, using ahuge amount of power, because
you know, although we love to,you know, use these ai is these
these chat bots and use them foryou know, a wide variety of
tasks and stuff like that.
And it feels like magic, right,you know you're, you're typing
(15:10):
in there like, hey, what do I do?
Blah, blah, blah, and it's justboom, we get you this whole
clean answer, really wellthought out, and stuff like that
.
But what we're not reallyseeing or talking about much
behind the scenes is that youknow, all these ai models are
having to be powered via datacenters, right, data centers
(15:34):
with big server farms, so it'sjust tons and tons of very high
power computers and GPUs thatare crunching this data and then
pushing it out to us, theconsumers, the users right of
(15:55):
these products, to toessentially, you know, get all
of our magical experience.
But in these big, gigantic,massive data farms or server
farms, you know, data centers.
They consume a lot of power andthe more users that you have
(16:18):
means that there's their.
There are more instances ofthis ai needing more computing
power, which would also meanthat it needs more servers,
which also means more power.
Servers cost power and if,especially if, you're already
running them at the, you know,the highest amount that you can,
they're consuming a huge more,or a huge amount more power than
(16:43):
they were before.
And, depending on how ourmovement towards cleaner energy
sources, most of the time thesedata centers, since they're
using so much power and it's a24 seven job they're not getting
their energy from a clean, muchmore sustainable energy source,
(17:08):
right.
They're going to be most likelygetting them from coal plants,
right?
Or or natural gas, or differenttypes of fossil fuel, different
types of fossil fuels, becausethose currently at least, since
we haven't really gone through a, you know, a huge process of
(17:32):
trying to overhaul how ourenergy has been created or is
being created.
Our electricity most are stillin that coal power.
Coal powered gas, natural gaspowered.
Speaker 1 (17:48):
More pollutant, more fossil fuel intensive power
generation right, so it's greatmore of a climate issue, right
correct, correct, yeah it's.
Speaker 2 (18:03):
It's something that actually, when Bitcoin was
really high, oh yeah, oh dude,we okay, go ahead.
I want to cut you off no, noworries, I was just going to say
, yeah, like when Bitcoin wasreally big, bitcoin mining
became a really big deal, but italso became kind of a climate
(18:24):
thing, because again you'regoing to be the more people,
more consumers right, becausethat's more people instead of
these big server farms.
Now you're having, like,individuals consuming more
powers to their home, whichagain can lead to more fossil
fuels being burned, which meansmore CO2 being pumped into our
(18:45):
atmosphere, or a greenhousegases.
It doesn't even have to just beCO2, sometimes it can be others
and overall, creating a higherenvironmental impact than it
once previously was.
And it took a couple of lawsand sanctions and stuff like
(19:07):
that that happened againstdifferent individuals, not just
like actual individuals, likeconsumers, but other different
stuff like that that was outthere to try to lower the amount
of people trying to do thatkind of stuff, right, you know
(19:28):
yeah, yeah, no, I feel that, um,you know, it's interesting.
Speaker 1 (19:32):
One thing I was thinking about this is kind of
like a different direction, Iguess.
But you know, ai is kind oflike a thing that can lead to so
many new innovations becauseit's advanced and it's seen like
human thinking, right, like thecollaboration you know, and, um
, what's interesting about that,right, is one of those things
(19:56):
you would think are new sourcesof energy, right, but instead
it's causing such a giganticproblem for that.
Yeah, I mean, how do you thinkthis would compare to, say, like
Google, right, like Google'sfacility, like, say, chat GBT,
(20:19):
right, mm-hmm, what it takes topower that, how does that
compare to?
I mean, I don't know if youknow this, I'm totally open for
peer speculation on this right,but the amount of power in GPUs,
cpus, like you said, that takesto power chat GBT, for example,
(20:40):
what about like Google's searchengine?
How would those two compare,right?
Speaker 2 (20:47):
Um, you know that's a great question and I'm going to
say that this is, for sure,more speculation on my end, but
I would say that it is using apretty decent amount of energy
comparably.
But I almost feel like chat GBTwould end up using more Now.
(21:11):
Of course, over time I think itwould get a little bit less.
But the reason why I'm sayingthis is that Google has had
years and years of time toperfect the algorithm that is
used for their Google searches,which has also most likely
created a lot of performancegains as well, so that a regular
(21:33):
Google search, or most people'sGoogle search, is going to use
a fraction of the power that itused to use when they first
implemented Google search andall sorts of other stuff like
that.
Um, because, you know, not onlyare you getting efficiency
gains when it comes to justoverall hardware right, as
hardware advances, it can domore tasks with uh, with uh
(21:58):
using less power, good stufflike that but then you also have
the fact, too, of just a muchmore tighter, better um
algorithm better, better andmore efficient algorithm to be
able to product or conduct tothose searches uh without any
(22:20):
issues or with far less issues,as well as far less power.
Speaker 1 (22:25):
Yeah.
Speaker 2 (22:29):
I'm sorry, man, that's all good man.
Speaker 1 (22:33):
There's some Daffy duck going there for a second.
Speaker 2 (22:35):
I did, man, I really did, I did.
Speaker 1 (22:40):
I'm sorry, that was just funny.
Um, oh man, um, okay, so, uh,well, I guess one of my, one of
my thoughts would be um,couldn't it help develop, uh,
(23:02):
the hardware that would takeless size and energy to power it
right, yeah, yeah, I mean likethe snake eating its tail, but
reversed, you know.
Speaker 2 (23:14):
Yeah, I mean I, I do feel as though eventually we're
going to get to that kind ofprocess where it's it's the
amount of good that it'sactually putting out is far
outweighs the amount of energythat it's using.
Right, I, I can see it as akinto, um what we're current.
(23:38):
You know, scientists arecurrently trying to continuously
replicate with uh fusion powerright and fusion where it's
currently.
We can create a fusion reaction, but the amount of power that
it takes for us to continue thatreaction not only create it,
but continue that reaction umoutweighs the amount of energy
(24:02):
that we are actually getting outof the equation.
Um.
Only recently, right, we've hadbreakthroughs, and this has
been decades and decades um ofus researching and, and you know
, getting closer to making moreefficient things Um.
But have we gotten to a pointwhere, at least um relatively
(24:24):
recently, we've been able tocreate a nuclear reaction, that
a nuclear fusion reaction I needto be very specific there um
that is able to produce moreenergy than what it's required
to be put into it?
Um, and so I think AI for sureis going to get to that point,
(24:44):
right.
But I think it's that it's thattipping glass right.
It's figuring out what is theright?
How long will it take?
The cool part about it is AI isvery quick and especially
generative AI that we have,where it's learning consistently
and growing, and the amount ofdata that we can continue to
(25:05):
feed into it, the faster thatit'll learn and faster that
it'll grow, and so it's only amatter of time of being able to
figure out ways to make it farmore efficient and how it does
its processes, as well as farmore efficient in, like other
alternatives, to make sure thatit can produce, right in some
(25:26):
ways, more, um, energy than whatit takes to, uh, maintain it or
increase it.
Speaker 1 (25:36):
Yeah, well.
Well, I guess one thought washaving while you're talking to
was.
You know, the possibility isone day these become the main
thing, right?
Yeah, like AI is the main thing.
There's not search engines,right?
Well, not in the status thatthey're in now, correct.
Speaker 2 (25:59):
Correct, yeah, and the instance that we see it now,
I think it'll be a verydifferent landscape, kind of
some sort of an amalgamation oflike a search engine and a chat
bot and other thing like that.
Right, it's kind of like whatwe first, at least what my first
envision of like a GoogleAssistant would be, or a Siri
(26:21):
would be yeah, or your iPhoneusers, we're out of here.
But, yeah, I feel like it wouldbe something akin to that where
you're not going to any ofthese anymore.
You're just like, hey, what isthis?
You?
Speaker 1 (26:37):
know, help me find whatever Right Exactly, or do
this task or whatever so ofcourse you always want to go
more and more green, but I mean,would it be a problem if that's
like the biggest power drawerthat we have from a technology
standpoint someday in the future?
You know?
Speaker 2 (26:58):
I mean, sure, I guess that wouldn't be the worst
thing, but it is one of thosethings that we still need to be
cognizant of.
I mean honestly, if anything,this should be one of the
biggest reasons why we shouldpush more towards more green
(27:20):
energy creation methods, greenerenergy creation methods.
So that way we are not, youknow, that way we can have our
cake and eat it too.
You know, you can use a hugeamount of power, but you have
these tons and tons of redundantgreen energies that are not
(27:42):
polluting and are zero carbon,footprint-wise or whatever like
that and not like zero carbon asin.
I'm buying carbon credits,because that is a thing, but
zero carbon as in.
I am literally not producingany greenhouse gases to go up
(28:02):
into our atmosphere.
So I think it does have thatpossibility of really pushing us
towards that.
Speaker 1 (28:10):
Yeah, yeah, I agree, that's interesting, that's
really cool.
Well, no, that's awesome.
That's awesome, man, yeah.
Speaker 2 (28:23):
I mean sorry, Well, go ahead.
Speaker 1 (28:25):
No, I was just thinking of something silly.
But no, what were you going tosay?
Because I was about to move onto the next topic.
Unless you want to add more, no, no.
Speaker 2 (28:34):
Actually, I was going to say I was like this leads us
great, perfectly into our nexttopic and I'm going to let you
take it over, yeah.
Speaker 1 (28:41):
That's what I'm talking about.
We've been doing a podcast fora while, yeah, so you know.
I kind of laid it out at theintro here, but the New York
Times, reportedly, is thinkingabout suing chat GPT.
They added to their terms ofservice just a few weeks ago
that AIs could not scrub theirwebsite for information, which I
(29:07):
find that fascinating, becauseI'm sure there already have been
others that have done it, butI'm sure there'll be more.
You know to follow this.
That's going to be.
It's going to hinder its growth.
But the real big thing is, ifit sues chat GPT, this could go
to a place where it's in thefederal court and if New York
(29:32):
Times wins, then chat GPT wouldbe completely scrubbed clean.
Like you know, Mr Clean wouldcome in there and just be, like
you know, and it would have tostart completely over.
But see, the difference thistime is so many websites would
(29:54):
be blocked from it studyingRight.
So this is interesting.
What I have a hard time gettingon board with the journalism
issues.
Speaker 2 (30:11):
You know what I mean?
No, I get, that, is it?
Speaker 1 (30:15):
because most of the stuff's behind the paywall.
Speaker 2 (30:18):
Like I would say that right, because I mean, think
about it right.
We already know that when a lotof these journal, these news
papers, as well as just thesenews stations moved online, they
started to really strugglebecause, you know, a lot of
(30:39):
their user base was from youknow, the paper, right, getting
the paper in the mail orwhatever like that.
And so when we started movingmore towards a digital age,
people were less likely, even ifthe subscription was really
cheap, but they were less likelyto pay that subscription model.
(31:00):
They were more likely to useother means to be able to, yeah,
get the news that they wantedand all that kind of stuff.
Correct, correct, right.
So I can't understand wherethey're coming from.
Where now you're going to havean AI that's going to do this
(31:21):
right without ever seeing thearticle right or deliver the
information without ever likedelivering?
Who created this article?
Who wrote this article?
Where did it come from?
Blah, blah.
Of course, you know, putting instandards for AIs to like, cite
their sources and all sorts ofother stuff is great, but it
(31:42):
still is.
There's very few people who aregoing to go back and actually
check right.
Think about it.
When people Google searchsomething, they're very it's
very rare for them to go backand be like, hmm, all right, I
got that one opinion, let mejump to the next one.
Let me jump to the next one andI read a couple of different
articles.
Now they're going to stick withthat first one.
(32:03):
That sounds right to them andthey're going to keep going.
Speaker 1 (32:08):
I agree, man, that's interesting.
So, yeah, I don't know.
This is interesting to me.
I wonder as to what it is thatthey're concerned about having
been scrubbed, because if it'slike historical details, I mean
(32:28):
that's going to be everywhere.
But if it's like exact thingsthat they wrote, I don't know.
I mean, I don't know, man.
Yeah, the scope is kind ofbroad.
It's not plagiarism.
Speaker 2 (32:42):
What's that?
Oh yeah, the scope's kind ofbroad.
Speaker 1 (32:46):
Right, because you know again, I've never had chat
GBT.
Give me something that'sverbatim, something else right,
mm-hmm?
Yeah, I just don't know, man,That'd be really interesting for
it to be completely plagiarized.
I mean, we talked about this afew weeks ago.
That's what some of thesepeople think is that their work
(33:06):
is, just like a lot of authorsand artists and stuff, their
works is straight up gettingplagiarized.
You know, mm-hmm, and I haveyet to see that.
I have seen, like you know,using information gains to write
its own thing.
You know, but I just haven'tseen that.
Maybe that's why this lawsuitwon't go anywhere.
(33:29):
You know what I mean.
But also, my biggest questionwas this I wanted to ask you
this because we've talked aboutthis before Chat GBT wasn't
taught by going to website towebsite, to website to website.
It was taught sorry, in thismassive database, right?
Speaker 2 (33:50):
Yeah, yeah, essentially, yeah, it was a
massive data set that it had orthat they purchased from another
company, most likely ofinformation that that company
was able to obtain via the orsorry, either via scraping or,
you know, asking permission orwhatever, and then subsequently
(34:12):
selling that information thatthey were able to obtain to you
know different organizationslike OpenAI, who would use that
then to train their AI model offof this, off of that data right
.
So you know, it is veryimportant and it does kind of it
(34:39):
can cause issues later on downthe road because we don't know
exactly how.
Like I said before, it's prettybroad, so it can have broader
implications than just AI.
Speaker 1 (34:54):
I agree, yeah, yeah, that's a good point.
So I was just thinking aboutthat because, you know, that
kind of makes information evenmore of a valuable than it was
before.
You know, and now all these AIis like some of the eyes are
going to be made behind thescenes and they're not going to
(35:16):
do the best things, right yeah?
Speaker 2 (35:18):
yeah, exactly, you know, it's one thing that's kind
of interesting that I thoughtof, too, is that I feel like for
the longest time we had alreadywe had started to get to the
point of where a lot of thesepeople who were housing so much
(35:42):
data, so much user dataespecially, didn't know what to
use it for.
And then, long comes, you know,here comes along AI.
Right, you can feed that exactthat user data that you're just
holding on to to theseorganizations that are creating
AI's and AI models, and you cannow have a new use for that data
(36:08):
that you've been collecting,that data and that information
that you've been using for.
Speaker 1 (36:13):
That's a good point.
Speaker 2 (36:15):
Yeah, data brokers are.
Speaker 1 (36:19):
It's a very lucrative field because of that you know
Right, which is that's not aneasy field either.
But man that's been going onfor a long time and now it's
probably more profitable of ause than ever, you know.
Speaker 2 (36:33):
Yeah, exactly.
Speaker 1 (36:35):
Yeah Well, I think we had a good discussion about
this.
That was more of a quickmention on my part, no real
reason to dive into it deeperbecause it's not official yet,
you know.
Yeah, so unless you hadsomething you wanted to add, bro
.
Speaker 2 (36:53):
No, I mean, I think the only thing that I wanted to
add is that it's reallyinteresting to see these first
few lawsuits or even politicalhearings that are going to come
up with, not just OpenAI, butall these other companies that
(37:16):
are trying to jump into the AIgame.
It's interesting to see wherethese rules, rulings of these
cases can, how that can changeand shape how our AI is and how
it works in the future.
In some ways, maybe this is thetinfoil hat to me, but this is
(37:44):
one way where one way to slowdown or stop the progression of
AI.
Yeah, it's definitely that,because at least, even if you
wanted to do a pause, this isone of the best ways to do it.
If you can get a judge to sidewith your lawsuit, then you can
(38:11):
set AI creation, as well as AIprogression, back by maybe years
worth.
Speaker 1 (38:23):
Yeah, absolutely yeah , they have to find another way
to gain information.
That's a good point.
Yeah, there could even besituations where, depending on
what the court asks for, theycould ask for a stay or a hold
on chat GPT's services.
They could be blocked outbefore the trial even starts or,
(38:47):
I guess, while the trial isgoing on.
It's interesting.
Well, I think we had a gooddiscussion on this man.
Speaker 2 (38:58):
Also.
Speaker 1 (38:59):
I just went full screen and I basically just
realized it looks like I have anoutline of hair around this
giant baby face.
It's like when you shadow theedges.
That's what it looks like.
Anyway, on that note, thank youguys so much for tuning in.
We have a great time every weekdoing this.
(39:19):
Just because we're best friends, we're brothers, we get to
catch up.
You guys make this even better.
So thank you so much for beinghere, for being here every week.
Please remember to like, shareand subscribe.
Smash the follow subscribebutton and we'll catch up with
you next week.
So we're talking this week about the struggles of AI.
Right, we're back on AI train,a thing that has to find the
season of catch up podcasts,right, mm?
Hmm, so we're talking about theold AI issues really because
one of them they're both bigissues One of them seems to be
(00:24):
more immediate than the other.
First of all, chat, gpt andother big AI like it.
Right, they draw a lot of powerbecause they have multiple,
multiple computers forprocessing that provide
processing horsepower for thisAI, right, mm?
hmm, and that's a big dream onpower, big dream of resources.
(00:47):
So we're going to talk aboutthat.
We're also going to talk aboutthe New York Times possibly I
want to report accurate releasesfor target matter journalistic
media possibly is gearing up tosue AI, specifically open AI,
(01:08):
sorry, and chat GPT at thatwould be.
You know, potentially whatthey're saying is we'll come
back to this further.
I'm not explaining this wellall right now, but we'll come
back to this further.
What potentially could happenis this lawsuit, if it happens,
could go to a federal judge thatapparently, if it favors the
(01:31):
New York Times, chat GPT wouldhave to scrub all of its
knowledge and start over, whichis crazy it is, and so it's
definitely a good talking point.
We want to discuss that also.
A side note the New York Timesactually changed its terms of
service.
That prevents any AI from, youknow, scrubbing or not scrubbing
(01:58):
, but you know, gatheringinformation from that site.
You know, I don't really knowquite how you prevent that.
I guess you have to be in theAPI, right, because it would be.
Speaker 2 (02:13):
It wouldn't be in the APIs.
It would be similar to, kind oflike, what has happened with
the social media platformformerly known as Twitter.
Speaker 1 (02:25):
I know as X.
Speaker 2 (02:26):
Exactly when they.
They changed the API ruling.
I think Reddit also didsomething similar, where they
changed how their API is used.
So, yeah, I believe you can doit in the API and so that way
the AI is not able to easily goin there and grab specific data
(02:49):
points and stuff.
Speaker 1 (02:52):
Right, and what is API stand for?
Speaker 2 (02:55):
You know, I don't know, we're going to have to
Google that one.
I actually don't know that one.
Let's find out, let's let'spull it up I see it all the time
.
Speaker 1 (03:07):
I see it all the time because I work with these
social media platforms a lot.
I've always wanted you know youwant you see it all the time
because, oh yeah, I do.
Multiple reasons, yeah,application program interface.
Speaker 2 (03:23):
Programming.
Speaker 1 (03:25):
Programming interface .
You're right, seems a littleredundant.
Not really it's good.
So yeah, that is going to beinteresting, because I do know.
So chat GPT has been withoutInternet surfing abilities for
over a month now.
(03:45):
They seemingly have not beenable to get the issue that
they've had fixed, and sothey've been without Internet
abilities for, like I said, overa month.
Speaker 2 (03:56):
Or at least that plug in.
You know, is that right Plug in?
Speaker 1 (03:59):
Yeah, exactly, but you know I can't search anything
that's current on chat GPTright now.
But, yeah, I totally understandwhat you're saying.
Yes, but one thing I asked, Ithink I wanted to look at
(04:20):
Something I'd posted before,right, because it was posted
something for my music.
And so I wanted to look atsomething I posted before and it
said sorry, but you know,facebook's API, which is also
the same as Instagram's,prevents me from searching for
this stuff.
And I was like, hey, yeah, mm,hmm, so that's kind of the
reason for that.
(04:41):
So yeah, that's interesting, man.
We got a lot to talk about.
We sure do, man.
So I just say we go ahead andget into it.
What's going on?
Everybody, I'm John and I'mDenison, and this is the catch
up.
All right, before we get backinto our main topic.
(05:23):
I just wanted to discuss thethree best ways.
Let me say, when I saydiscussed, right, I just want
you to sit down on the couch,talk with me for a second.
I'm just trying to connect withyou, man, exactly, you and I
have been going our separateways for so long.
You know I try to sit down withconversation.
You want to go play video games?
So, listener, whoever you are.
(05:46):
This was not directed at Denison, in case you were confused.
The listener, I just want tohave a conversation with you,
man.
I want to let you know thethree best ways to support this
show.
You know, there they are flyingacross your screen.
Number one oh man, it's alwaysthe build up and in case RSX
(06:09):
type are nice dude, pretty slick.
They were like really cool ifit was red.
You know, for historicalreasons, number one is Please
leave us rain in review.
Wherever you're listening,wherever you're watching.
It really helps us to grow,helps us to know what we can do
to do better or what we'realready doing well, but it also
(06:32):
gets us down from more peopletoo.
Speaker 2 (06:33):
So it's a big help.
Speaker 1 (06:37):
Please just take a few seconds to leave five stars
or whatever you're feeling,right?
I feel like I preach this weekafter week.
I'm really serious.
It would help us a lot.
You know I'm not going to be afan of it.
It would help us a lot, youknow.
So please just take a fewseconds to help us out.
Number two is give us a follow.
Oh my gosh, oh, I spoke toosoon.
(07:00):
There's a red model right there.
Dang, that was crazy.
I shouldn't have.
Should have held my, held mytongue.
Number two is follow us onFacebook or on YouTube at our
podcast name and we go liveevery Thursday night talking
about our topics in real time.
(07:20):
If you don't already watch usthis way, it allows you to jump
in the discussion with us inreal time and leave your
comments, and we want to have adiscussion with all of you, and
you're getting the format, therecord version like that in real
time.
But it's more.
You know, as cool as that is, wewant to be able to have a
really great conversation withyou and expand this discussion
(07:44):
as a whole.
And number three oh man, youreally pulled that PS5
controller out of the bag.
All right, there is your threeright there.
No way it's an Xbox controller.
I was, I was duped, I wasbamboozled.
All right, we're going to showyou a little Something,
(08:08):
something you might say.
Speaker 2 (08:10):
This is not on show and tell.
Speaker 1 (08:13):
Yeah, this is not on a shirt, right?
Just picture it.
Let's go ahead and get into itwith a nice big pair of
headphones.
You know, I'm saying that issomething Looks real clean,
right?
So that's something that youcould have wrapped around your
body, you know round your nogginman.
(08:34):
Not to make it sound weird oranything, but I think actually I
can give you all a betterexample.
Give me two, give me twoseconds.
You store Perfect.
Speaker 2 (08:48):
One, two.
Speaker 1 (08:50):
Hey, I wasn't ready.
I thought I was.
Speaker 2 (08:53):
Time out man.
Speaker 1 (08:55):
Yeah, I know time flies when you're trying to show
off your merch.
Speaker 2 (08:58):
You are man.
I mean, you're right, you'reright, time does fly, but it is
what it is.
Yeah, yeah, we're out heretrying to show the people you
know show and tell, just likeyou know when you were, when you
were younger, you know, and youwent to class and you always
brought your favorite toy rightand you got to bring it up to
the front of the class and showthem your favorite toy right.
(09:20):
We were just trying to show ourfavorite clothing that we have,
you know, our favorite merch Toour audience, you know, so you
guys can check it out and, youknow, if you feel so inclined,
go ahead and snack.
You know a couple of things.
Speaker 1 (09:37):
And I just want to say too I feel so disappointed
in myself because this is not onour shop page and it needs to
be, you know, Mm.
Hmm, so let's see no, we don'twant black.
You can't see the design.
Speaker 2 (09:59):
No, you can't, no, you can't.
So, it does look good, no itdoes.
Speaker 1 (10:04):
We'll just, we'll leave it at that.
It looks clean, because I forsome reason can't.
Oh, there we go, let's show itto the people.
Speaker 2 (10:14):
Show it to the people .
Speaker 1 (10:15):
All right guys.
Sorry for the long wait on this.
I forgot how to use the websitethat we do this on Little
little gray action, right?
Just a standard gray.
I know a lot of people likegray around here.
Good hard work.
Look are you sharing.
Speaker 2 (10:30):
There we go, I can see that.
That is pretty done clean.
Speaker 1 (10:35):
It's actually me modeling it, you know.
Speaker 2 (10:38):
And you grew out your beard a lot.
Speaker 1 (10:41):
Yeah, yeah me.
Me not being able to is just apart of my character.
Dude, I actually have a fullbeard.
Speaker 2 (10:48):
I see I see Full black beard.
Yeah, yeah, I tend to, I tendto, you had to dye your hair as
well.
Speaker 1 (10:59):
That's.
I can't leave it.
It's all about the lighting,you know.
Ok, ok, that's interestinglighting, hey, but you didn't
ask about the gains, though.
Let's go.
Yeah, so you know, good, cleanshirt.
I'll get this added to our shopfor you.
I think you'll like this one.
This is the guaranteed Favorite, for sure, without doubt.
(11:23):
So please check that out.
That's over at our shop.
By the time you hear this,you'll be at our shop in the
link below.
So, yeah, please check that out.
I know you'll like it.
All right, let's get into thistopic here, this here topic.
Why don't you start us off, man, talking about the AI and the
(11:48):
power that's required for it,right?
Speaker 2 (11:50):
Yeah, of course.
So, as you all may know, wehave been going through a
sweltering, hot summer, you know, and throughout all of that I'm
sure everyone's kind of heardof, like you know, this is
becoming a more a new normal, asit were, due to climate change,
(12:13):
you know, just because of howmany or how much CO2 that we're
pumping into the atmosphere andall the all the things that come
with that.
Speaker 1 (12:25):
Right, which I actually have a question for you
, right, mm-hmm, saturday.
What's your high?
Speaker 2 (12:36):
Let's see Texas High in Texas.
Let's see what we got here.
Speaker 1 (12:42):
Give me a second here , the and surprising me, hours
on Saturday is one oh seven.
Speaker 2 (13:04):
Well, that's crazy.
Speaker 1 (13:06):
Sunday it's one oh seven and Monday is really crazy
and Monday it's one oh five.
Tuesday one oh three.
Wednesday one oh one.
Thursday one oh one man, itdoesn't drop below a hundred
till a week from tomorrow that'scrazy, it's not, it's right.
Speaker 2 (13:24):
Yeah, I mean, we've been in the hundreds for a while
, man, we are expected to get alittle bit of a cold front here,
though you know good old coldfront.
It's going to be 97 degrees,man, cold front.
Gotta get the jacket out.
Speaker 1 (13:39):
I'm about to say you got that north face ready, man,
mm-hmm, yeah.
So I just wanted.
I was curious and brought thatup because I'm feeling it.
I'm very much feeling it, man.
We've already had a couple ofone.
Oh, seven days up here wealready had 10 days straight of
100 plus degrees.
You know I'm saying so mm-hmm.
I feel that.
(14:00):
So anyway, go ahead, sorry noworries.
Speaker 2 (14:03):
So, yeah, I was really gonna say is you know,
since, as everyone said, youknow, due to these changes in
our the amount of CO2 that wehave in our atmosphere, as well
as some of the changes that aregoing on and how that is
affecting and heating us up alot of pressure has been started
(14:28):
to grow on finding cleanerenergy sources and reducing the
energy that we do consume andstuff like that.
And all of that being said goestowards what we're, you know,
one of the topics that we'regoing to be talking about, which
is chat, gpt or even other.
(14:49):
Ai is like chat GPT, using ahuge amount of power, because
you know, although we love to,you know, use these ai is these
these chat bots and use them foryou know, a wide variety of
tasks and stuff like that.
And it feels like magic, right,you know you're, you're typing
(15:10):
in there like, hey, what do I do?
Blah, blah, blah, and it's justboom, we get you this whole
clean answer, really wellthought out, and stuff like that
.
But what we're not reallyseeing or talking about much
behind the scenes is that youknow, all these ai models are
having to be powered via datacenters, right, data centers
(15:34):
with big server farms, so it'sjust tons and tons of very high
power computers and GPUs thatare crunching this data and then
pushing it out to us, theconsumers, the users right of
(15:55):
these products, to toessentially, you know, get all
of our magical experience.
But in these big, gigantic,massive data farms or server
farms, you know, data centers.
They consume a lot of power andthe more users that you have
(16:18):
means that there's their.
There are more instances ofthis ai needing more computing
power, which would also meanthat it needs more servers,
which also means more power.
Servers cost power and if,especially if, you're already
running them at the, you know,the highest amount that you can,
they're consuming a huge more,or a huge amount more power than
(16:43):
they were before.
And, depending on how ourmovement towards cleaner energy
sources, most of the time thesedata centers, since they're
using so much power and it's a24 seven job they're not getting
their energy from a clean, muchmore sustainable energy source,
(17:08):
right.
They're going to be most likelygetting them from coal plants,
right?
Or or natural gas, or differenttypes of fossil fuel, different
types of fossil fuels, becausethose currently at least, since
we haven't really gone through a, you know, a huge process of
(17:32):
trying to overhaul how ourenergy has been created or is
being created.
Our electricity most are stillin that coal power.
Coal powered gas, natural gaspowered.
Speaker 1 (17:48):
More pollutant, more fossil fuel intensive power
generation right, so it's greatmore of a climate issue, right
correct, correct, yeah it's.
Speaker 2 (18:03):
It's something that actually, when Bitcoin was
really high, oh yeah, oh dude,we okay, go ahead.
I want to cut you off no, noworries, I was just going to say
, yeah, like when Bitcoin wasreally big, bitcoin mining
became a really big deal, but italso became kind of a climate
(18:24):
thing, because again you'regoing to be the more people,
more consumers right, becausethat's more people instead of
these big server farms.
Now you're having, like,individuals consuming more
powers to their home, whichagain can lead to more fossil
fuels being burned, which meansmore CO2 being pumped into our
(18:45):
atmosphere, or a greenhousegases.
It doesn't even have to just beCO2, sometimes it can be others
and overall, creating a higherenvironmental impact than it
once previously was.
And it took a couple of lawsand sanctions and stuff like
(19:07):
that that happened againstdifferent individuals, not just
like actual individuals, likeconsumers, but other different
stuff like that that was outthere to try to lower the amount
of people trying to do thatkind of stuff, right, you know
(19:28):
yeah, yeah, no, I feel that, um,you know, it's interesting.
Speaker 1 (19:32):
One thing I was thinking about this is kind of
like a different direction, Iguess.
But you know, ai is kind oflike a thing that can lead to so
many new innovations becauseit's advanced and it's seen like
human thinking, right, like thecollaboration you know, and, um
, what's interesting about that,right, is one of those things
(19:56):
you would think are new sourcesof energy, right, but instead
it's causing such a giganticproblem for that.
Yeah, I mean, how do you thinkthis would compare to, say, like
Google, right, like Google'sfacility, like, say, chat GBT,
(20:19):
right, mm-hmm, what it takes topower that, how does that
compare to?
I mean, I don't know if youknow this, I'm totally open for
peer speculation on this right,but the amount of power in GPUs,
cpus, like you said, that takesto power chat GBT, for example,
(20:40):
what about like Google's searchengine?
How would those two compare,right?
Speaker 2 (20:47):
Um, you know that's a great question and I'm going to
say that this is, for sure,more speculation on my end, but
I would say that it is using apretty decent amount of energy
comparably.
But I almost feel like chat GBTwould end up using more Now.
(21:11):
Of course, over time I think itwould get a little bit less.
But the reason why I'm sayingthis is that Google has had
years and years of time toperfect the algorithm that is
used for their Google searches,which has also most likely
created a lot of performancegains as well, so that a regular
(21:33):
Google search, or most people'sGoogle search, is going to use
a fraction of the power that itused to use when they first
implemented Google search andall sorts of other stuff like
that.
Um, because, you know, not onlyare you getting efficiency
gains when it comes to justoverall hardware right, as
hardware advances, it can domore tasks with uh, with uh
(21:58):
using less power, good stufflike that but then you also have
the fact, too, of just a muchmore tighter, better um
algorithm better, better andmore efficient algorithm to be
able to product or conduct tothose searches uh without any
(22:20):
issues or with far less issues,as well as far less power.
Speaker 1 (22:25):
Yeah.
Speaker 2 (22:29):
I'm sorry, man, that's all good man.
Speaker 1 (22:33):
There's some Daffy duck going there for a second.
Speaker 2 (22:35):
I did, man, I really did, I did.
Speaker 1 (22:40):
I'm sorry, that was just funny.
Um, oh man, um, okay, so, uh,well, I guess one of my, one of
my thoughts would be um,couldn't it help develop, uh,
(23:02):
the hardware that would takeless size and energy to power it
right, yeah, yeah, I mean likethe snake eating its tail, but
reversed, you know.
Speaker 2 (23:14):
Yeah, I mean I, I do feel as though eventually we're
going to get to that kind ofprocess where it's it's the
amount of good that it'sactually putting out is far
outweighs the amount of energythat it's using.
Right, I, I can see it as akinto, um what we're current.
(23:38):
You know, scientists arecurrently trying to continuously
replicate with uh fusion powerright and fusion where it's
currently.
We can create a fusion reaction, but the amount of power that
it takes for us to continue thatreaction not only create it,
but continue that reaction umoutweighs the amount of energy
(24:02):
that we are actually getting outof the equation.
Um.
Only recently, right, we've hadbreakthroughs, and this has
been decades and decades um ofus researching and, and you know
, getting closer to making moreefficient things Um.
But have we gotten to a pointwhere, at least um relatively
(24:24):
recently, we've been able tocreate a nuclear reaction, that
a nuclear fusion reaction I needto be very specific there um
that is able to produce moreenergy than what it's required
to be put into it?
Um, and so I think AI for sureis going to get to that point,
(24:44):
right.
But I think it's that it's thattipping glass right.
It's figuring out what is theright?
How long will it take?
The cool part about it is AI isvery quick and especially
generative AI that we have,where it's learning consistently
and growing, and the amount ofdata that we can continue to
(25:05):
feed into it, the faster thatit'll learn and faster that
it'll grow, and so it's only amatter of time of being able to
figure out ways to make it farmore efficient and how it does
its processes, as well as farmore efficient in, like other
alternatives, to make sure thatit can produce, right in some
(25:26):
ways, more, um, energy than whatit takes to, uh, maintain it or
increase it.
Speaker 1 (25:36):
Yeah, well.
Well, I guess one thought washaving while you're talking to
was.
You know, the possibility isone day these become the main
thing, right?
Yeah, like AI is the main thing.
There's not search engines,right?
Well, not in the status thatthey're in now, correct.
Speaker 2 (25:59):
Correct, yeah, and the instance that we see it now,
I think it'll be a verydifferent landscape, kind of
some sort of an amalgamation oflike a search engine and a chat
bot and other thing like that.
Right, it's kind of like whatwe first, at least what my first
envision of like a GoogleAssistant would be, or a Siri
(26:21):
would be yeah, or your iPhoneusers, we're out of here.
But, yeah, I feel like it wouldbe something akin to that where
you're not going to any ofthese anymore.
You're just like, hey, what isthis?
You?
Speaker 1 (26:37):
know, help me find whatever Right Exactly, or do
this task or whatever so ofcourse you always want to go
more and more green, but I mean,would it be a problem if that's
like the biggest power drawerthat we have from a technology
standpoint someday in the future?
You know?
Speaker 2 (26:58):
I mean, sure, I guess that wouldn't be the worst
thing, but it is one of thosethings that we still need to be
cognizant of.
I mean honestly, if anything,this should be one of the
biggest reasons why we shouldpush more towards more green
(27:20):
energy creation methods, greenerenergy creation methods.
So that way we are not, youknow, that way we can have our
cake and eat it too.
You know, you can use a hugeamount of power, but you have
these tons and tons of redundantgreen energies that are not
(27:42):
polluting and are zero carbon,footprint-wise or whatever like
that and not like zero carbon asin.
I'm buying carbon credits,because that is a thing, but
zero carbon as in.
I am literally not producingany greenhouse gases to go up
(28:02):
into our atmosphere.
So I think it does have thatpossibility of really pushing us
towards that.
Speaker 1 (28:10):
Yeah, yeah, I agree, that's interesting, that's
really cool.
Well, no, that's awesome.
That's awesome, man, yeah.
Speaker 2 (28:23):
I mean sorry, Well, go ahead.
Speaker 1 (28:25):
No, I was just thinking of something silly.
But no, what were you going tosay?
Because I was about to move onto the next topic.
Unless you want to add more, no, no.
Speaker 2 (28:34):
Actually, I was going to say I was like this leads us
great, perfectly into our nexttopic and I'm going to let you
take it over, yeah.
Speaker 1 (28:41):
That's what I'm talking about.
We've been doing a podcast fora while, yeah, so you know.
I kind of laid it out at theintro here, but the New York
Times, reportedly, is thinkingabout suing chat GPT.
They added to their terms ofservice just a few weeks ago
that AIs could not scrub theirwebsite for information, which I
(29:07):
find that fascinating, becauseI'm sure there already have been
others that have done it, butI'm sure there'll be more.
You know to follow this.
That's going to be.
It's going to hinder its growth.
But the real big thing is, ifit sues chat GPT, this could go
to a place where it's in thefederal court and if New York
(29:32):
Times wins, then chat GPT wouldbe completely scrubbed clean.
Like you know, Mr Clean wouldcome in there and just be, like
you know, and it would have tostart completely over.
But see, the difference thistime is so many websites would
(29:54):
be blocked from it studyingRight.
So this is interesting.
What I have a hard time gettingon board with the journalism
issues.
Speaker 2 (30:11):
You know what I mean?
No, I get, that, is it?
Speaker 1 (30:15):
because most of the stuff's behind the paywall.
Speaker 2 (30:18):
Like I would say that right, because I mean, think
about it right.
We already know that when a lotof these journal, these news
papers, as well as just thesenews stations moved online, they
started to really strugglebecause, you know, a lot of
(30:39):
their user base was from youknow, the paper, right, getting
the paper in the mail orwhatever like that.
And so when we started movingmore towards a digital age,
people were less likely, even ifthe subscription was really
cheap, but they were less likelyto pay that subscription model.
(31:00):
They were more likely to useother means to be able to, yeah,
get the news that they wantedand all that kind of stuff.
Correct, correct, right.
So I can't understand wherethey're coming from.
Where now you're going to havean AI that's going to do this
(31:21):
right without ever seeing thearticle right or deliver the
information without ever likedelivering?
Who created this article?
Who wrote this article?
Where did it come from?
Blah, blah.
Of course, you know, putting instandards for AIs to like, cite
their sources and all sorts ofother stuff is great, but it
(31:42):
still is.
There's very few people who aregoing to go back and actually
check right.
Think about it.
When people Google searchsomething, they're very it's
very rare for them to go backand be like, hmm, all right, I
got that one opinion, let mejump to the next one.
Let me jump to the next one andI read a couple of different
articles.
Now they're going to stick withthat first one.
(32:03):
That sounds right to them andthey're going to keep going.
Speaker 1 (32:08):
I agree, man, that's interesting.
So, yeah, I don't know.
This is interesting to me.
I wonder as to what it is thatthey're concerned about having
been scrubbed, because if it'slike historical details, I mean
(32:28):
that's going to be everywhere.
But if it's like exact thingsthat they wrote, I don't know.
I mean, I don't know, man.
Yeah, the scope is kind ofbroad.
It's not plagiarism.
Speaker 2 (32:42):
What's that?
Oh yeah, the scope's kind ofbroad.
Speaker 1 (32:46):
Right, because you know again, I've never had chat
GBT.
Give me something that'sverbatim, something else right,
mm-hmm?
Yeah, I just don't know, man,That'd be really interesting for
it to be completely plagiarized.
I mean, we talked about this afew weeks ago.
That's what some of thesepeople think is that their work
(33:06):
is, just like a lot of authorsand artists and stuff, their
works is straight up gettingplagiarized.
You know, mm-hmm, and I haveyet to see that.
I have seen, like you know,using information gains to write
its own thing.
You know, but I just haven'tseen that.
Maybe that's why this lawsuitwon't go anywhere.
(33:29):
You know what I mean.
But also, my biggest questionwas this I wanted to ask you
this because we've talked aboutthis before Chat GBT wasn't
taught by going to website towebsite, to website to website.
It was taught sorry, in thismassive database, right?
Speaker 2 (33:50):
Yeah, yeah, essentially, yeah, it was a
massive data set that it had orthat they purchased from another
company, most likely ofinformation that that company
was able to obtain via the orsorry, either via scraping or,
you know, asking permission orwhatever, and then subsequently
(34:12):
selling that information thatthey were able to obtain to you
know different organizationslike OpenAI, who would use that
then to train their AI model offof this, off of that data right
.
So you know, it is veryimportant and it does kind of it
(34:39):
can cause issues later on downthe road because we don't know
exactly how.
Like I said before, it's prettybroad, so it can have broader
implications than just AI.
Speaker 1 (34:54):
I agree, yeah, yeah, that's a good point.
So I was just thinking aboutthat because, you know, that
kind of makes information evenmore of a valuable than it was
before.
You know, and now all these AIis like some of the eyes are
going to be made behind thescenes and they're not going to
(35:16):
do the best things, right yeah?
Speaker 2 (35:18):
yeah, exactly, you know, it's one thing that's kind
of interesting that I thoughtof, too, is that I feel like for
the longest time we had alreadywe had started to get to the
point of where a lot of thesepeople who were housing so much
(35:42):
data, so much user dataespecially, didn't know what to
use it for.
And then, long comes, you know,here comes along AI.
Right, you can feed that exactthat user data that you're just
holding on to to theseorganizations that are creating
AI's and AI models, and you cannow have a new use for that data
(36:08):
that you've been collecting,that data and that information
that you've been using for.
Speaker 1 (36:13):
That's a good point.
Speaker 2 (36:15):
Yeah, data brokers are.
Speaker 1 (36:19):
It's a very lucrative field because of that you know
Right, which is that's not aneasy field either.
But man that's been going onfor a long time and now it's
probably more profitable of ause than ever, you know.
Speaker 2 (36:33):
Yeah, exactly.
Speaker 1 (36:35):
Yeah Well, I think we had a good discussion about
this.
That was more of a quickmention on my part, no real
reason to dive into it deeperbecause it's not official yet,
you know.
Yeah, so unless you hadsomething you wanted to add, bro
.
Speaker 2 (36:53):
No, I mean, I think the only thing that I wanted to
add is that it's reallyinteresting to see these first
few lawsuits or even politicalhearings that are going to come
up with, not just OpenAI, butall these other companies that
(37:16):
are trying to jump into the AIgame.
It's interesting to see wherethese rules, rulings of these
cases can, how that can changeand shape how our AI is and how
it works in the future.
In some ways, maybe this is thetinfoil hat to me, but this is
(37:44):
one way where one way to slowdown or stop the progression of
AI.
Yeah, it's definitely that,because at least, even if you
wanted to do a pause, this isone of the best ways to do it.
If you can get a judge to sidewith your lawsuit, then you can
(38:11):
set AI creation, as well as AIprogression, back by maybe years
worth.
Speaker 1 (38:23):
Yeah, absolutely yeah , they have to find another way
to gain information.
That's a good point.
Yeah, there could even besituations where, depending on
what the court asks for, theycould ask for a stay or a hold
on chat GPT's services.
They could be blocked outbefore the trial even starts or,
(38:47):
I guess, while the trial isgoing on.
It's interesting.
Well, I think we had a gooddiscussion on this man.
Speaker 2 (38:58):
Also.
Speaker 1 (38:59):
I just went full screen and I basically just
realized it looks like I have anoutline of hair around this
giant baby face.
It's like when you shadow theedges.
That's what it looks like.
Anyway, on that note, thank youguys so much for tuning in.
We have a great time every weekdoing this.
(39:19):
Just because we're best friends, we're brothers, we get to
catch up.
You guys make this even better.
So thank you so much for beinghere, for being here every week.
Please remember to like, shareand subscribe.
Smash the follow subscribebutton and we'll catch up with
you next week.
Popular Podcasts
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.