Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
I have something that's not really f worthy, but something
that I find to be somewhat interesting. And I'm just
not exactly sure how to handle this because again I'm
maybe not the right guid to ask. Because I like
using AI. I like using artificial intelligence. It helps me,
It inspires me on some things. It is just fun
(00:20):
to kind of noodle around with, you know, I'll talk
to it about different things. You know, I'll have it
like create like a bracket tournament and like simulate what
something could happen if all these things happen together, rewrite history,
and just kind of read about it. It's it's actually
a lot of fun if you're using it for for
just giggles. It's actually quite useful if you have good
(00:41):
reasons to use it for the reasons that were intended theoretically.
And so with that all being said, artificial intelligence is
coming into our workplace, our workplace like Matt and I's workplace. Matt,
how much do you use artificial intelligence?
Speaker 2 (00:56):
I don't use it all that often. I under like
I anticipate I probably well more and more as it
gets more incorporated into every piece of what we do.
Speaker 1 (01:04):
Artificial intelligence as in like CHAGYPT is very helpful for me,
so I do use it quite a bit. But there
are programs out there, and I'm sure you've seen some
that can make things sound the way a person actually sounds,
These deep fakes videos that are altered ever so slightly
to make it look like it's legitimate, but it actually
is not. Yeah, that's it's scary stuff, Yeah it is.
(01:27):
And it actually I've told this story before. I've had
a person and they made multiple accounts because it got
blocking them when I was working in Des Moines. Use
a deep fake app essentially, and I don't know why
they would pay for this just to use it like this,
but would make it sound like my voice. I mean,
it sounded like me, and it's like, oh, listen to
(01:48):
this incredibly racist guy, Emery songer. And then it's a
video it's so like of the audio of what like
my show would sound like, and it's me saying like
incredibly hurtful and hateful things that I never would say
to somebody, especially on the radio. And it's scary, Like
what do we do about that? Right? How do I
protect myself? How do I protect my reputation? Well there's
(02:09):
another thing there, right, I can promise you, because I'm
sitting here that I'm not artificial intelligence. You're listening to
a real, live person. You know what media you can't
exactly ever really know as real these is print when
you're reading something from a newspaper or a magazine. And
we talked about this when Sports Illustrated kind of flipped
over a little bit. Using artificial intelligence to basically come
(02:32):
up with lists and write stories and be able to
put these things together without really having a legitimate editing process,
something that honestly I don't think should be legal, but
it's something that you know, I guess the cat is
out of the bag on it. There were are, or
there were books on a list saying heat Index your
(02:52):
guide to the best of Summer. It was a section
in the Chicago Sun Times and the Philadelphia Inquirer last week.
It was a list of books that were reading suggestions
for the summer. Over half of the books on the
list did not actually exist. Now the author of the piece,
his name is Marco Buscaglia. He said he helped use
(03:14):
AI for research and didn't double check what it produced.
So the writer of this and again he used artificial
intelligence not to put the whole thing together, but to
help him formulate the list, and he said a really
stupid error on my part. So it makes you wonder
what do people actually do when they're looking for help
from artificial intelligence? So are they just taking it verbatim?
(03:34):
Are they acting like these things are actually legitimate and real?
How much more work would it take for somebody who
you know, you're a journalist or you're at least a
writer for a publication. The least you could do is
maybe Google search to make sure that you have at
least an idea of what these books may be, instead
of just trusting an artificial intelligence engine to tell you
(03:56):
books that you think are new and would be recommended.
Over the summer, it made up over half of the
books on the list. So here's my warning to heed,
not just for people who are using artificial intelligence for
their job, but people who are consuming artificial intelligence on
accident or on purpose. Understand that it's not always one
(04:18):
hundred percent ac can remember that time I asked it
what the most influential people in the world.
Speaker 2 (04:21):
Were, Oh, and it took the didn't it take all
of its data? From twenty twenty four?
Speaker 1 (04:25):
It didn't have Donald Trump because it said, if Donald
Trump wins the election, something he had done months earlier,
he would be on the list.
Speaker 2 (04:33):
That's kind of concerning to me that you didn't tell
it to pull from old stuff, but it still did.
Speaker 1 (04:39):
I asked it, what are the foremost influential people in
the world, and where does the pope fit? This was
when the pope was getting selected.
Speaker 2 (04:46):
Yeah, and it.
Speaker 1 (04:48):
Didn't have the President of the United States there, And
I asked it why and it said, well, Donald Trump
hasn't won election yet, and that's it was not true. Now, see,
that's an obvious mistake that I can catch and I
can correct. But it tells you that these engines are
still not perfect. There are flaws, they can be mistakes.
A lot of the times that I'm talking to it,
I kind of am testing it a little bit. I'll
(05:09):
have it talk to me about what you know something
that's obscure enough that I happen to know a lot
about for one reason or another, that it doesn't know about,
Like what are the best songs by a musical artist? Right,
What's what's an underneath musical artist that you don't think
super mainstream at that you're a big fan of, Like
that you could name, you know, five or six songs
(05:30):
of their catalog and be like, yeah, this isn't a
band that I heard a lot on the radio, but
I was really into them.
Speaker 2 (05:36):
I don't know, well, so not ray Well. Mainstream is
an interesting conversation right now because there are so many
different streams. I feel like there's a lot of medium streams.
What do you consider mainstream? I guess probably what KISSFM plays, right.
Speaker 1 (05:49):
That would be what I would say.
Speaker 2 (05:50):
I don't listen to anything on there, no disrespect, but.
Speaker 1 (05:52):
So you wouldn't know, okay, So, like, give me a
band that you think isn't like a local band it has,
you know, just one hundred listens on their songs on
social media. Give me a band that you know people
around the country would at least be aware of.
Speaker 2 (06:10):
That they'd be okay. So a band that people are aware.
Speaker 1 (06:13):
Of, but you would think is probably not a band
that would be obviously mainstream in any way, shape or form,
wouldn't be on a lot of people's radar unless you're
looking for them.
Speaker 2 (06:21):
I don't know. There's a lot of places to go
with this.
Speaker 1 (06:23):
Well, I just want to test AI and see if it.
Speaker 2 (06:25):
Oh, I got you? Yeah, Yeah, the answer is not
important to the as much to the conclusion.
Speaker 1 (06:30):
Right, Just get give me a band that maybe people
listening wouldn't necessarily have heard of.
Speaker 2 (06:35):
Let's say Radiohead. Radiohead, now I think most people have
probably heard of them, but yeah, but they're off the
radar enough. Maybe so five best. I asked it for
the five best songs and it comes up, and it's
very detailed. It says, here are my five top five
Radiohead songs. Number one Paranoid Android number one. Interesting that
that's number one.
Speaker 1 (06:55):
It says it's from Okay computer nineteen ninety seven.
Speaker 2 (06:57):
Sounds like robotic bias there to me, considering it's Paranoid Android.
Speaker 1 (07:01):
Oh so you're saying that it's brought up simply because
of references a computer.
Speaker 2 (07:06):
I just based on people and what they know about Radiohead.
I'm shocked.
Speaker 1 (07:10):
It's not creep not even in the top five, you're kidding,
not even in the honorable mentions.
Speaker 2 (07:14):
It's not one of mine. It's not a top five
favorite of mine personally, But as far as like Okay,
it's based on all the criteria, based on critical acclaim,
fan favorites, cultural impact, and artistic achievement.
Speaker 1 (07:26):
Here's a sadless of five that could be considered their best.
Number one was Paranoid Android and says a sprawling, multi
part epic Oh, it's a beautiful compared to Bohemian Rhapsody.
Speaker 2 (07:34):
Without a doubt, you know what? I okay? What just
depends on the criteria. If they're saying objectively, what's their
quote unquote best song. I'm not gonna mess with the criteria.
I was thinking more popular. Most people like. The only
song they know about from Radiohead is the song cree Creep.
Everyone knows that song, Yeah creep right, everyone can sing
along to that. They've heard it before, It's in movies
(07:54):
all the time. Paranoid Android a legitimately great song. You're
saying absolutely, yeah, it's a beautiful song. It's Radiohead is
I don't know. I'm a fan and they're brilliant musicians.
Speaker 1 (08:04):
So you're you're impressed with AI that that's number one.
Speaker 2 (08:07):
I'm intrigued to see what I'm not. I don't know,
could you be impressed with AI number time? Intrigued to
see what they're the data they used to collect this.
Speaker 1 (08:14):
Karma Police Number two is Karma Police.
Speaker 2 (08:17):
Yeah, solid and based on Paranoid Android be number one.
I'm not surprised if that's number two. Number three was Idiotech.
Idiotech doesn't surprise me either. Based on this list. Now,
this criteria is making more sense, and that is definitely
a top five for me. I love all these songs.
And number four is Fake Plastic Trees. Yeah yeah, and
number five.
Speaker 1 (08:36):
Was Street Spirit Street Spirit Okay, so yeah, it's this dark,
hypnotic and one of the band's most emotional songs. And
then the honorable mentions was everything in Its Right Place? Yeah,
they're there, Reckoner Pyramid song, No surprises in how to
Disappear completely.
Speaker 2 (08:50):
Yeah, it's a street Spirit. Man, that's really interesting. That's
from the Bens. That's ninety five. I'm surprised. I would eh, okay, hey,
you know what, I'm interested to see the criteria.
Speaker 1 (09:02):
I let me ask it what exactly its criteria is
and I will report on that when we return. Because
this is what I'm saying, right like, is this accurate?
Did they pull correct? This could be a fun conversation
you could have with AI or you could pull from it.
But they did this with books that are coming out
this summer. This was in the Inquirer in Philly and
on the sometimes in Chicago, and half of the books
that it listed were not They don't they don't exist,
(09:24):
the big books.
Speaker 2 (09:25):
I had that. I had that for my you get
an f Sorry, I used I had it yesterday. We
would have time. In the segment, I saw it's hilarious.
Speaker 1 (09:32):
I mean, how does that happen?
Speaker 2 (09:34):
Right?
Speaker 1 (09:34):
So, but I will say what the radio Head best
Song criteria is and we can try to understand why
it thinks this way and why it pulled actual legitimate
songs and it wasn't incorrect. We'll do that next on
news radio eleven to ten kfab. I asked Chad Gpt,
that's the model I'm using for the five best songs.
I didn't say what's the highest selling songs? It's information
I could find on Wikipedia, What are the best albums
(09:57):
or none of that. I said, what are the best songs?
And I did not define what that is. I let
it come up with its own definition. The number one
song it said Matt was Paranoid Android, and you said
could be the best song.
Speaker 2 (10:07):
It makes sense to me. It made me curious to
the criteria. I was initially surprised that it didn't say creep,
because that's certainly the most well known. But then if
you put it in the context of Okay, if this
was the same thing, but the question was what's the
greatest movie of all time, you probably wouldn't say one
of the Marvel movies, even though it's the most popular
and made the most money. Right now, that would be
the song Creep. It's the most popular and certainly has
(10:29):
made the band the most money and is probably the
reason why they're still relevant today. Otherwise they'd probably be
playing in a tavern in Liverpool somewhere and never have
made this worldwide acclaim because that song was huge.
Speaker 1 (10:41):
But well, let me tell you what I said about
Paroid Android here and because I asked, give me a
reason why you had this, listen to this order, and
it said why Paranoid Android is number one? Is Structure Ambition,
a four part suite essentially Radioheads version of prog rock.
The shifts in time, signature, tone, and energy are bold
and masterfully executedim It frequently appears in the Best Songs
(11:02):
of All Time and lists, and it's a standout track
on an album called one of the best of the
nineties in Okay, Computer Live Favorite It's a show stopper
at its concerts, showing off the band's range and cohesion
and legacy symbolic of radioheads leap from alt rock to
something more experimental and cerebral, which you agree with that assessment.
Speaker 2 (11:19):
I would anyone out there who likes prog rock rush,
name your artists like that. Certainly Bohemian Rhapsody Queen fits
into that conversation. Anybody who likes songs like that, who
likes to be taken on a ride and an experience,
check it out. The only thing you may not like
the way the lead singer sings tom yorke I do,
but that's kind of the barrier to entry for a
lot of people.
Speaker 1 (11:39):
But you might be surprised they It says summary of
criteria used artistic innovation, critical praise, fan legacy, and emotional
and philosophical impact, and then it graded it on a
scale of one to three check marks and all three
of those things and then just basically added it up.
So Paranoid Android had three for artistic innovation, two for
critical praise, two for fan legacy, and two for emotional
(12:00):
philosophical impact. And nothing had more than that score across
the board. And that's why I had it at number one,
So I don't know. It's interesting to look in and
peer in because it can be so specific when you're
asking it for information like this and for discussion, but
(12:20):
when you're using it for work, you got to be
like ready to double check the work because I'm not
going to be able to intelligently talk about this. Even
after I looked this up. I couldn't talk to you
about paranoid Androdi. I've never heard it, I never studied it.
I can't just use AI and be like, hey, you
should listen to this.
Speaker 2 (12:35):
It's right. So if you're not an expert to go
in with, you just have to trust AI, and obviously
you can't sometimes.
Speaker 1 (12:41):
And that's what happened to these guys with this book review,
that half the books weren't even real More on the
Way coming up on news radio eleven ton kfab