ChatGPT vs. Claude: Kelly’s First Time Artwork

Two Awesome People

New Episodes Every Tuesday!

We are Kelly and Don, two awesome people who luckily married each other. We're navigating family, marriage, friendships, work, and all of the aches and pains that come with being in your 40s, and we're doing it in the podcast-verse for your listening pleasure. Come hang out with two awesome people.

We'd love to hear from you: don.broida@gmail.com

Follow Two Awesome People on Social Media

YT: @twoawesomepeople

IG: @two.awesome.people

TT: @two.awesome.people

All Episodes

Two Awesome People

ChatGPT vs. Claude: Kelly’s First Time

May 26, 2026 • With Kelly and Don • Season 2 • Episode 63

0:00 | 33:34

Welcome to Episode 63 of Two Awesome People — your middle-aged take on pop culture, keeping us “old people” informed.

This week, Kelly uses ChatGPT and Claude for the first time. Don uses AI all the time. Kelly has never used it once. So naturally, her first question is about how to nicely ask a neighbor to move their giant car.

We compare ChatGPT vs. Claude, ask if cereal is soup, find out which leg should go into pants first, talk about AI hallucinations, and debate whether you should say please and thank you to a chatbot.

But things also get a little real: should people use AI for medical questions? Relationship advice? Love letters? And at what point does a helpful tool start replacing your own opinions, instincts, and personality?

The final verdict: AI can be useful, but don’t use a machine to replace your personality.

Subscribe for more middle-aged takes on pop culture, tech, marriage, parenting, and whatever weird thing we’re trying to understand this week.

Chapters:

00:00 Kelly Tries AI for the First Time
02:17 Kelly’s First ChatGPT Question
03:12 ChatGPT vs. Claude
06:07 Is AI the New Google?
07:03 Asking AI Dumb Questions
12:15 AI Hallucinations and Data Centers
15:14 AI Invents New Words
17:14 Should You Treat AI Like a Person?
20:48 When ChatGPT Gives Relationship Advice
25:53 ChatGPT or Be a Functioning Adult?
27:20 The Verdict: Is AI Good?
30:41 Give It Time: Tarps Off
32:16 Anniversary Recommendation

#TwoAwesomePeople #ChatGPT #ClaudeAI #AI #ArtificialIntelligence #AITools #ChatGPTvsClaude #Podcast #ComedyPodcast #MarriageComedy #MiddleAgedTake #PopCulture #TechComedy #AIChatbot #ChatGPTForBeginners

Come be AWESOME at Two Awesome People's YouTube channel

Reach out to us directly : don.broida@gmail.com

SPEAKER_01 0:00

You love it, you hate it. Either way, it's here to stay. You know what I'm talking about, Kelly?

SPEAKER_00 0:04

Cleaning toilets.

SPEAKER_01 0:06

No, we're talking about ChatGPT, Claude, and AI bots.

SPEAKER_00 0:10

I don't know these things.

SPEAKER_01 0:11

Kelly's a total noob.

SPEAKER_00 0:13

Yeah.

SPEAKER_01 0:13

Noob.

SPEAKER_00 0:14

You just want to keep saying noob.

SPEAKER_01 0:16

I like the word noob. It's kind of close to boob.

SPEAKER_00 0:18

I think of the word nub.

SPEAKER_01 0:20

You're a noob nub.

SPEAKER_00 0:21

I am.

SPEAKER_01 0:25

Today Kelly's gonna use Chat GPT and Claude for the first time.

SPEAKER_00 0:29

I have never used it ever, ever, ever. And Don uses it quite often. But there was an ad on TV, and I thought that this was very interesting. And I don't know for sure if this is ChatGPT or not. But it was like you hold your phone up on a camera mode in your refrigerator, and so it sees the ingredients that are in your fridge, and you're like, what can I make for dinner tonight? And it tells you.

SPEAKER_01 0:52

I have pickles and mayonnaise and orange juice. What should I make tonight?

SPEAKER_00 0:58

It would probably give you like the number for a pizza place.

SPEAKER_01 1:01

You can't make anything with this.

SPEAKER_00 1:03

It would probably tell you in like a very like reassuring positive way.

SPEAKER_01 1:08

Well, you've never used it, so how do you even know that?

SPEAKER_00 1:10

Well, my impression of it is that it's always like, that's an excellent question. I love your curiosity, and you have such interesting taste in food. Let's explore our options together.

SPEAKER_01 1:23

Anchovies and an apple.

SPEAKER_00 1:25

Yeah. And it looks like you like salty foods. Maybe this would be good for you.

SPEAKER_01 1:30

Why don't you just stick out your tongue and pour salt on it?

SPEAKER_00 1:33

There you go.

SPEAKER_01 1:33

You've never asked ChatGPT a question. True.

SPEAKER_00 1:37

Do you go to chat GPT.com?

SPEAKER_01 1:39

Oh my god, Kelly.

SPEAKER_00 1:41

Or is it another?

SPEAKER_01 1:42

That's like you looking at a cell phone and being like, how do I use this?

SPEAKER_00 1:46

There was probably a time when I would have been that way.

SPEAKER_01 1:49

Yes. You go to chat GPT.com or Claude, I think it's Claude.com, or you can download an app.

SPEAKER_00 1:58

What does GPT stand for?

SPEAKER_01 2:00

I don't know. Chat, okay, I'll chat. And then my brain stops there before it hits the GPT part.

SPEAKER_00 2:05

Generative pre-trained transformer.

SPEAKER_01 2:10

That takes all the fun out of it.

SPEAKER_00 2:12

It's a transformer.

SPEAKER_01 2:13

So what's your first ever question for Chat GPT, Cal?

SPEAKER_00 2:17

We have a neighbor right now who is constantly parking their giant ass cars in front of like our house. And I would ask it how to communicate in a nice way to move your fucking car.

SPEAKER_01 2:33

You know what I would do?

SPEAKER_00 2:34

What?

SPEAKER_01 2:34

Slash the tires. Yeah, gee it. That gets it moved real fast. They'll be like, oh shit, we should park it in the driveway.

SPEAKER_00 2:42

That's what I would ask ChatGPT.

SPEAKER_01 2:44

I'm predicting that it's going to give you several responses. You could do this or you could do this. I don't think it's going to give you just one. So what's the question, Kelly?

SPEAKER_00 2:55

I said, how do I ask my neighbor to stop parking their big ass car in front of my house?

SPEAKER_01 3:01

What does it respond with?

SPEAKER_00 3:03

Avoid opening with quote your big ass car unless your relationship already runs on mutual roasting.

SPEAKER_01 3:12

So we're taking the question from Chat GPT and putting it in a Claude to see the difference.

SPEAKER_00 3:18

Oh, Claude takes a little longer.

SPEAKER_01 3:20

Oh, Claude's going deeper.

SPEAKER_00 3:22

Hey, neighbor's name. Hope all is well. I wanted to ask a small favor. Would you mind parking in front of your own place when possible? That one's better.

SPEAKER_01 3:33

So you like Claude more than Chat GPT?

SPEAKER_00 3:36

Yes.

SPEAKER_01 3:37

Why do you like Claude more than ChatGPT?

SPEAKER_00 3:40

It gave less options for answers. So I find that to be more directive.

SPEAKER_01 3:47

The different AI agents are basically trained differently. They're different companies, different products. So it's like drinking Coke versus Pepsi. And they give you different responses based on you have an account with these things or no? Oh yeah.

SPEAKER_00 4:03

So do they learn your voice so that they give you things in your voice? I'm being serious.

SPEAKER_01 4:12

Sure. I have the $20 a month subscription to Chat GPT. Ho ho ho ho. So you can subscribe and gives it more $20 a month? Surprise.

SPEAKER_00 4:26

Unreasonable, Don.

SPEAKER_01 4:28

Really?

SPEAKER_00 4:28

Well, can't you use it for free?

SPEAKER_01 4:31

You can use it to a certain extent for free.

SPEAKER_00 4:33

So you pay like $250.

SPEAKER_01 4:36

Oh, we're getting in hot water here.

SPEAKER_00 4:39

Chat GPT, do you pay for Claude also?

SPEAKER_01 4:42

No, I have not yet paid for Claude. But I do use Claude. I use it literally every day. It's a tool that I use for almost everything.

SPEAKER_00 4:50

Like, how do I tell my wife that she needs to? Is that how you use Claude?

SPEAKER_01 4:55

What's interesting is that your instinct is to use it as a tool to help relationships.

SPEAKER_00 5:02

Social interaction.

SPEAKER_01 5:03

Social interaction. That to me is a big no-no. So going back to the differences between the two, because you just saw a little bit of a difference. So ChatGPT is known to act as like an eager, fast, accommodating assistant.

SPEAKER_00 5:17

Yes, that's how it came across. It wants to please.

SPEAKER_01 5:20

Claude acts more like a mentor, more thoughtful, more deep thinking, will disagree with you more often.

SPEAKER_00 5:28

I like that much more.

SPEAKER_01 5:29

That's your personality, Kelly.

SPEAKER_00 5:31

Yes.

SPEAKER_01 5:31

Don't chat GPT, Kelly.

SPEAKER_00 5:34

Yeah, I'm Claude. I much more like ones that will like not coddle you. I'm surprised that you like the one that does the like rah-rah. Why? I like that. That's an awesome question. Keep it up.

SPEAKER_01 5:48

Thank you. I'll keep going for you.

SPEAKER_00 5:52

It's interesting how fast AI has come to like infiltrate so many parts of um society. I mean, our son uses chat GPT.

SPEAKER_01 6:07

Everybody from here on out will use Chat GPT or AI bots. Absolutely. It's like us using Google.

SPEAKER_00 6:14

It's like us using Google.

SPEAKER_01 6:17

It functions different than Google, but prior to us, people didn't use Google. Then Google came along and they're like, what's this? You search the what and you do the what. And that's what you are doing towards Chat GPT. You're like, wait, how do I use this and why? Like, that's how our past generations were with Google. Like my grandmother called it the confuser.

SPEAKER_00 6:39

I was researching um Chat GPT and like its effects in education because I think that that's like pretty interesting. And it was equating it to when Wikipedia came out.

SPEAKER_01 6:53

Right.

SPEAKER_00 6:53

And that teachers freaked out about Wikipedia. And so it just is a matter of like figuring out like how to use it in a responsible way.

SPEAKER_01 7:03

Yeah, it's a tool.

SPEAKER_00 7:04

Yeah.

SPEAKER_01 7:05

You have a hammer and you can hit a nail dumb. You could hit somebody. Right. Let's try something else, Kelly.

SPEAKER_00 7:13

Let's try it.

SPEAKER_01 7:14

Ask it. Is cereal just soup? Is cereal just you know what it is?

SPEAKER_00 7:21

It is it is. Cereal is the only cold soup I like.

SPEAKER_01 7:26

For breakfast.

SPEAKER_00 7:27

Well, for anything. I hate cold soup.

SPEAKER_01 7:30

Really? Gazpacho is great. Like a watermelon cazpacho.

SPEAKER_00 7:34

We have two arguments for both sides.

SPEAKER_01 7:36

Let's hear the is cereal just soup, according to Claude.

SPEAKER_00 7:41

Yeah. So it's it could be soup. The argument is cereal shares the basic structure. Which is what I think. A liquid with stuff in liquid, basically.

SPEAKER_01 7:54

Stuff in liquid.

SPEAKER_00 7:56

And then for it's probably not soup, it says most definitions of soup require the liquid to be savory or heated.

SPEAKER_01 8:05

So you can't have a dessert soup.

SPEAKER_00 8:08

Also, soup is intentionally cooked. Cereal just gets dunked. Like it just cereal is its own category. Language is just bad at capturing it.

SPEAKER_01 8:20

So what is the name of that category?

SPEAKER_00 8:22

So you can just have a full-on conversation. Hello, welcome to the ChatGPT says both too. It gives counter-arguments. Cereal is in the soup adjacent family, but survive has its own category through sheer cultural stubbornness.

SPEAKER_01 8:40

Sereal has cultural stubbornness.

SPEAKER_00 8:42

That's pretty awesome.

SPEAKER_01 8:43

According to ChatGPT, cereal has cultural stubbornness.

SPEAKER_00 8:48

I think I have cultural stubbornness.

SPEAKER_01 8:50

I think that we now need to tell our 10-year-old who loves uh cereal but hates soup that he's actually eating soup. And he should try other soups.

SPEAKER_00 9:00

Maybe we should heat it for him.

SPEAKER_01 9:01

Oh. So now that you're on the vein of what you can ask ChatGPT, what would you ask it now?

SPEAKER_00 9:08

Should I put my pants on left leg first or right leg first?

SPEAKER_01 9:13

I think it'll say you can put them on either way.

SPEAKER_00 9:16

It says whichever leg has the stronger resume.

SPEAKER_01 9:19

It's literally the stronger resume.

SPEAKER_00 9:22

Choose the leg nearest the pants. Claude says some athletes swear by consistent routines for focus, but for everyday pants donning purposes, you're safe either way. Claude's answer is very practical.

SPEAKER_01 9:41

There's no personality.

SPEAKER_00 9:45

Yeah, it has so, huh?

SPEAKER_01 9:47

See, there's a lot of information.

SPEAKER_00 9:49

You liked my question.

SPEAKER_01 9:50

Yeah, about the pants. I got a really good question for you to ask.

SPEAKER_00 9:54

Okay.

SPEAKER_01 9:54

That's gonna rattle your brain. Okay. Write a conversation between America's very first president, George Washington, and America's current president, Donald Trump, about the most important city in the country.

SPEAKER_00 10:09

Can I can I make up wrong cities? No, no. You're writing that what I just said. That is the prompt. I thought that I had to write a freaking conversation between them.

SPEAKER_01 10:21

Oh my God. No.

SPEAKER_00 10:22

I'm America's first president and I love Tallahassee. Well, I'm President Trump and I date.

SPEAKER_01 10:29

Oh my God. No. It will compare what it knows about George Washington versus what it knows about Donald Trump. So read it.

SPEAKER_00 10:36

You want me to read this whole thing?

SPEAKER_01 10:38

Sure.

SPEAKER_00 10:38

I must say the most essential city to this republic is Philadelphia. It served as our capital, the birthplace of the Constitution, and the center of commerce and intellect in the early nation. No city better embodies the American experiment.

SPEAKER_01 10:54

Words from George Washington through Claude.

SPEAKER_00 10:57

And then Trump says, George, great guy. Love him. But he's wrong. And I say this with tremendous respect. It's New York, biggest, best, most beautiful skyline in the world. I built a lot of it, frankly. Washington says, I would remind you that New York was also our first capital. I was inaugurated there, in fact. However, the seat of governance was deliberately moved to a more central location, the city bearing my name, I might add. And Trump said, Washington, D.C. is a disaster, George. Total mess. Beautiful monuments. I kept them very clean, by the way, but the city itself, not good. New York is the financial capital, the media capital, the real estate capital. And then Washington interrupts and says, You seem to define importance entirely by commerce. And Trump says, and you defined it by where a bunch of guys in wigs wanted to meet, no offense. And it just kind of goes down from there.

SPEAKER_01 12:00

Now are you starting to understand the like joy of our chatbots?

SPEAKER_00 12:07

I don't know that I like need this joy.

SPEAKER_01 12:10

This in particular is a fun exercise more than anything.

SPEAKER_00 12:15

Do you know what I think is funny that I recently learned is what an error is called that um chat. I don't know if it's specific to chat GPT or just AI when it says something that isn't true.

SPEAKER_01 12:27

Okay, grandma.

SPEAKER_00 12:28

Tell me what it's called.

SPEAKER_01 12:30

Hallucination.

SPEAKER_00 12:31

Okay, so when a person hallucinates, they see and understand things that are not real, but they think they're real.

SPEAKER_01 12:41

This is my life.

SPEAKER_00 12:42

And when ChatGPT or AI hallucinates, it is saying things are factual when they are not real.

SPEAKER_01 12:52

They should call it a drug trip. Um, chat GPT is on a high right now.

SPEAKER_00 12:57

It wasn't wrong. It was just hallucinating.

SPEAKER_01 13:00

Yeah. It's just bonkers.

SPEAKER_00 13:03

Yeah, it just forgot to take its pill today.

SPEAKER_01 13:05

It's on a bender. The way you could summise how much chat GPT is either by going to books and a human-like knowledge, in which it has read roughly, it's trained on roughly 200 million books. And to give you an idea of 200 million books, if you started reading now every day, you'd have to read every day for a hundred thousand years. And that's the amount of information it has in it.

SPEAKER_00 13:33

How does it have that information in it?

SPEAKER_01 13:35

They train it on text, every website imaginable thing. It just yes, it downloads the information. And this is why they're building data centers because it literally Which are very environmentally unfriendly.

SPEAKER_00 13:49

So this is an issue.

SPEAKER_01 13:50

Elon Musk wants to put data centers in space. True story.

SPEAKER_00 13:56

But wouldn't they be enormous?

SPEAKER_01 13:58

Yeah, they'd just be out there floating in space.

SPEAKER_00 14:01

I love that our solution to everything. It's like when you run out of room and you just like cram it up there. Send it to the attic. You can't see it, it's not really there. Do people work at data centers? Kelly. You don't know.

SPEAKER_01 14:17

I don't know.

SPEAKER_00 14:18

I thought you were an expert.

SPEAKER_01 14:19

My instinct is that yes, people need to work at data centers, but they're not people like, ooh, like your prompt comes in and they're over there like, let's send them back this.

SPEAKER_00 14:30

It's worth noting that I'm asking a man in a welcome to Flavortown Guy Fieri shirt about data centers and like how AI works. Like, like you sort of look like a crazy person.

SPEAKER_01 14:45

Well, if I go like this, I probably look a little crazier. Yeah, Kelly, you should read the room. Like I'm asking the wrong person.

SPEAKER_00 14:54

Way too much from you right now.

SPEAKER_01 14:55

This is like going to the guy that's sleeping on the bench at the park and be like, could you tell me about data centers? I'm one step away from that.

SPEAKER_00 15:02

My bad.

SPEAKER_01 15:02

My bad. So here's the next super fun thing I'm excited to show you.

SPEAKER_00 15:07

Super fun.

SPEAKER_01 15:08

It can invent new words. And we, if you guys have not known, love trending words.

SPEAKER_00 15:14

I said come up with a word for foot sweat.

SPEAKER_01 15:18

Do you have a lot of foot sweat?

SPEAKER_00 15:20

No.

SPEAKER_01 15:20

You just think about foot sweat.

SPEAKER_00 15:22

Oh, there's already a meant medical term for it. But here are some better options. Head do, sock juice.

SPEAKER_01 15:31

Foot sweat, according to Chat GPT, is sock juice.

SPEAKER_00 15:34

No, this is clod.

SPEAKER_01 15:35

Oh, sorry. Glog just sounds like what it is. Is that glob on your foot? We did an episode a little while ago about nostalgia.

SPEAKER_00 15:43

Yeah.

SPEAKER_01 15:44

Well, this is pre-stalgia. The odd sentimental feeling you get for a future inconvenience before it disappears. So standing in an airport security line thinking one day AI will make all of this obsolete. Weirdly, I miss this. Saving a screenshot of an outdated app interface because you know future people won't understand.

SPEAKER_00 16:08

That's like depression, I feel like.

SPEAKER_01 16:10

Pre-capturing nostalgia.

SPEAKER_00 16:12

Like, I'm gonna miss this.

SPEAKER_01 16:14

I used it to design math quizzes for our ninth grade grader teenager to get to take the HSPT test.

SPEAKER_00 16:24

And it worked very well. Very it had a few hallucinations.

SPEAKER_01 16:29

It did. We do the math quizzes and it would give wrong answers. And my teenager would be like, This is right. And I'd look at it and be like, this is right. And then I'd tell Chat GPT, you're wrong, and I am wrong.

SPEAKER_00 16:41

Would you really? And it would say, I am wrong.

SPEAKER_01 16:44

Yeah. If it's wrong, you tell it it's wrong. And it's like, oh, you're right. I've actually fought with it over three or four things where I'm like, no, it's in there. And it's like, no, it's not. And it's like, it's in there. And it's like, oh, you're right, it's in there. Yeah.

SPEAKER_00 16:59

And then it curses you somehow.

SPEAKER_01 17:01

I'm gonna hex you in the digital world. Yeah. You're hexed now. You were so right. I have a question.

SPEAKER_00 17:07

Okay.

SPEAKER_01 17:08

Do you use please and thank you with AI chatbots, with Claude, with Chat GPT?

SPEAKER_00 17:14

No, I will not. Tell me what time the store closes, please. Yeah, no, I don't do that.

SPEAKER_01 17:20

Why?

SPEAKER_00 17:20

Um, because they're not people.

SPEAKER_01 17:22

Thank you.

SPEAKER_00 17:26

Why? Is there a movement that's like we need to be polite? No, I think the the thing you should not do. Oh, is treat it like a person is treat it like a human. True. That's probably like opening a Pandora's box. If you start using manners with it, then it then you do start treating it like a person.

SPEAKER_01 17:44

The moment you start to have a real conversation, change your path.

SPEAKER_00 17:48

Change the path.

SPEAKER_01 17:49

Big red flag for you.

SPEAKER_00 17:50

Take a left. I am intrigued by the medical assistance that people get from it. Do you think that it's a good resource for that?

SPEAKER_01 18:01

Or a no, I think a doctor is a great resource.

SPEAKER_00 18:04

The reason being is it a good first resource?

SPEAKER_01 18:08

The reason being that a doctor is a great resource is the doctor also has access to Chat GPT. So the doctor can take all the information that ChatGPT is giving you and give their professional advice.

SPEAKER_00 18:21

Right. Without the risk of a hallucination.

SPEAKER_01 18:24

Or without any sort of risk. I asked Chat GPT how to explain its existence to Kelly.

SPEAKER_00 18:30

You you said to ChatGPT, explain yourself.

SPEAKER_01 18:34

Yes.

SPEAKER_00 18:35

Okay, got it.

SPEAKER_01 18:36

And so it said, don't treat Chat GPT or Claude like Google. Treat them like a person you hired who knows nothing about what you actually want.

SPEAKER_00 18:47

But you're saying to not treat them like a person.

SPEAKER_01 18:50

Yeah.

SPEAKER_00 18:50

Yeah. So that's a problematic instruction.

SPEAKER_01 18:53

I think it's a hugely problematic instruction. But I also was like, okay, what would I do to an intern that I could do with Chat GPD? I'd like to fuck with it just a little bit.

SPEAKER_00 19:03

No, you would never.

SPEAKER_01 19:04

I would fuck with an intern.

SPEAKER_00 19:05

No, you would never.

SPEAKER_01 19:07

When we had interns, I would always mess with the intern. Just a little bit.

SPEAKER_00 19:10

No, no, no.

SPEAKER_01 19:11

Hey, you have a run that you need to do and it's all the way in Ventura. Can you get in the car and go to the Ventura? And then they get halfway there, baby. I'm just kidding. I'm just kidding. Come back.

SPEAKER_00 19:21

I don't believe you. You would not do that. That's not nice at all. Do you think that they're gonna like give it a face and stuff?

SPEAKER_01 19:28

You could ask Chat GPT, create an image of what you look like.

SPEAKER_00 19:32

Can you do that with Siri too?

SPEAKER_01 19:34

I don't know. I don't use Siri. Siri's, I believe that Siri is structurally different than the AI spots.

SPEAKER_00 19:41

Than the AI bots. Got it. Okay. You hope.

SPEAKER_01 19:45

I hope.

SPEAKER_00 19:46

Yeah. Welcome to Flavortown.

SPEAKER_01 19:47

Welcome to Flavortown. That should just be the thing on the show when Don doesn't know and be like, just welcome to Flavor Town. Treating it like a person, having a conversation with it. That's a big fat no no. Treat it like a machine.

SPEAKER_00 20:02

Hard to imagine it as a machine, though, because you're using a machine already.

SPEAKER_01 20:08

Well, if it helps this way, what it actually is doing is it's not understanding your language. It's breaking down the letters and the clumps of words themselves and making them into tokens. And then it is mathematically Well, that's not helpful. Stop. Stop. It's important to know that it is actually not understanding the language. It's breaking it down and putting it into a mathematical equation, essentially. Right, right. Think of it like the hammer. Would you say please to your hammer?

SPEAKER_00 20:40

Right.

SPEAKER_01 20:41

Please nail this.

SPEAKER_00 20:42

Yeah, I get it.

SPEAKER_01 20:43

God. Don't use please and thank you with chat GPT.

SPEAKER_00 20:48

Fair. I will not. So there was this essay that was like very widely um published online. It was on a lot of different websites. Um, and it was written by this woman who wrote about having a boyfriend for an extended period of time. It was like a few month-long relationship. And one night he was asleep and she needed to use his laptop.

SPEAKER_01 21:14

She went snooping.

SPEAKER_00 21:15

She did. I don't think she snooped. You're turning her into a bad person.

SPEAKER_01 21:21

I don't know either of these people.

SPEAKER_00 21:23

Um, his chat GPT was open on his laptop. And I guess the chat GPT has a history of your conversations. Is this true? I don't.

SPEAKER_01 21:34

Real boring. Y'all be like, God, this is stupid.

SPEAKER_00 21:37

But his history was right there for her to see, and it was all about her. He was asking Chat GPT for relationship advice. And it was all about him not being attracted to her and him having doubts about their relationship and things like that. And it destroyed their relationship.

SPEAKER_01 21:58

Is it wrong that he did that?

SPEAKER_00 22:00

Well, this was an extended like period of time.

SPEAKER_01 22:06

That is my huge issue. Is having a recurring conversation about a real social interaction with a human. Right. It is not human. Right. It uh we are human. There is something in us that it will never do, even if it's 10 million times smarter than we are, it will not capture what makes us us.

SPEAKER_00 22:31

Well, so do you think that there is something in these platforms that that should be changed so that it does not respond in such a human way? Why are you making a face at me?

SPEAKER_01 22:49

This is a great question.

SPEAKER_00 22:50

If it responded more, obviously it couldn't be like in code, like computer code, but it was more like disjointed almost like in computer talk. It's a great idea, Kelly. I don't know whether to take you seriously or not.

SPEAKER_01 23:08

It's a great idea. However, okay, there it is.

SPEAKER_00 23:12

Now we're serious.

SPEAKER_01 23:14

It's a great idea. This is a product.

SPEAKER_00 23:17

Right.

SPEAKER_01 23:18

It is made by a corporation. This corporation's number one goal or a very high-up goal is to make money. The way it makes money is to keep you on its platform longer. So the more you interact with that chat with that chat, you're having the chat, the more you come to it, the more it makes money. So to make it boring and uninteresting, right, will make you log off faster. To make it conversational, right? Makes it that you will stay on longer.

SPEAKER_00 24:01

And that could be like their protection.

SPEAKER_01 24:04

OpenAI needs to come and hire Kelly to turn on boring talk. I'm available. I think the time of people write using it to write emails is almost already even out. Really? It's so freaking obvious.

SPEAKER_00 24:16

Yes. Yeah. I have gotten emails at work that yeah, you're like, oh, screw these people. Yeah.

SPEAKER_01 24:24

New York magazine had an article about how certain sex of the Amish love Chat GPT, right? Weird.

SPEAKER_00 24:32

Wait, I thought they didn't use technology.

SPEAKER_01 24:34

They do use some. Some of them are interesting. Really enjoying Chat GPT. And they view it as well, if I went through the effort to put this prompt into Chat GPT, I must care to make it better. And so I'm using a tool to make it better to then send you the email. I'm not just sending you like, whatever, Kelly. Boop.

SPEAKER_00 24:55

That's how the Amish are thinking of it. Well, this article is so weird.

SPEAKER_01 24:59

This article had an article about the Amish. Yeah, the article was about how the Amish love Chat GPT.

SPEAKER_00 25:06

There was a disconnect in that recap.

SPEAKER_01 25:08

Well, and there was a disconnect a little bit in the article, to be honest.

SPEAKER_00 25:12

Okay.

SPEAKER_01 25:12

It was a good headline.

SPEAKER_00 25:13

Sounds like a good headline.

SPEAKER_01 25:14

But it was written in New York magazine, which is like a legitimate publication.

SPEAKER_00 25:18

I would like to know what the Amish are asking it.

SPEAKER_01 25:21

Well, in this article, yeah, did it say? It was write a love letter to my wife.

SPEAKER_00 25:26

Oh. And the wife viewed it as Oh, and the wife knew that that's how it was written and viewed it as a good thing.

SPEAKER_01 25:35

Yeah.

SPEAKER_00 25:36

Oh, interesting. Yeah. I would not respond that way. Just so you know, so you can throw the love letter away that you were writing via Chat GPT.

SPEAKER_01 25:45

My love letter could not be written like Chat GPT. Mine is like, Kelly, you're awesome. I love you.

SPEAKER_00 25:50

I would want it to rhyme first of all.

SPEAKER_01 25:53

Yeah, right. Then you would know it's not from me.

SPEAKER_00 25:55

Get to work.

SPEAKER_01 25:57

All right. I got a quiz for you. This is like a yes, no. Should I use Chat GPT or be a functioning adult?

SPEAKER_00 26:04

I like that you explained a quiz. This is a yes no.

SPEAKER_01 26:08

You want to know if the growth on your arm is okay.

SPEAKER_00 26:11

Uh no, use chat GPT.

SPEAKER_01 26:13

Uh no, go to a doctor. Your spouse is wrong and you need advice about how to deal.

SPEAKER_00 26:18

I'm gonna say no on all of these. I ordered that question very specifically. Oh, I see. Sorry. I didn't pick it up.

SPEAKER_01 26:26

All right. It's your annoying kid's birthday and you need ideas.

SPEAKER_00 26:30

Yeah, I think maybe using it.

SPEAKER_01 26:32

Your kid's annoying.

SPEAKER_00 26:33

If your kid's annoying, then yeah, I think. I'm not saying that my kid is annoying. I'm saying if my kid, this is a hypothetical. And I would ask Chat GPT, like, what do I do for my annoying kid?

SPEAKER_01 26:46

And it'll be like your kid's not annoying, man.

SPEAKER_00 26:48

Like, you're a bad parent.

SPEAKER_01 26:49

Should we use Chat GPT or not? Uh, to replace all human interaction.

SPEAKER_00 26:55

No. Could you imagine if I was like, Yes.

SPEAKER_01 26:59

Yes, we need to replace. I was trying to get to absurdity, but apparently my questions.

SPEAKER_00 27:05

I said no. That was absurd. You got your answer.

SPEAKER_01 27:08

Did I?

SPEAKER_00 27:09

Yes. That was a good quiz.

SPEAKER_01 27:11

Thank you for jumping in on my weird train.

SPEAKER_00 27:13

Flavortown. The train to flavortown.

SPEAKER_01 27:15

Jumped on my flavortown train. Verdict. Our verdict.

SPEAKER_00 27:20

And the question is, is AI good?

SPEAKER_01 27:23

Is that what you want the question to be?

SPEAKER_00 27:25

Yes.

SPEAKER_01 27:26

Great.

SPEAKER_00 27:26

I think that AI is on its way to being very good. I think it's probably good right now, but I think that it needs to be uh harnessed in a better way right now. Through guardrails, through user manuals, essentially. I think people need to have like a better, firmer grasp on how to use it.

SPEAKER_01 27:55

Beautiful verdict, Kelly.

SPEAKER_00 27:56

You can get it tattooed on your butt.

SPEAKER_01 28:00

I'm gonna tattoo that whole thing across my lower back.

SPEAKER_00 28:02

You should.

SPEAKER_01 28:03

Then when I take off my shirt, you'll be like, oh, right, that's how I feel about AI. Your verdict. My verdict is yes, I love it. But it's a tool. We as humans can't be replaced, no matter how much knowledge that has. And I think that if you need real advice, you should go to somebody that you trust and you love and get real advice from a real human. And I think that the biggest problem with it is when you start asking too many opinion and advice questions, you start to lose what makes you you. And you start to rely on it. And I think that it's really important that you keep your opinions and your thoughts and don't rely on a fucking machine to replace your personality. I will get that tattooed on my butt. Don't use a machine to replace your personality. Chat GPT can't outdo this. Knowing that you like both or that you have No, I like Claude better. So now that you know you have Claude, I will ask Claude, are you good? So Don and Kelly versus the AI. The question for AI is, are you good? Its answer is yes, doing well. How can I help you today? So I guess I have to get more specific.

SPEAKER_00 29:26

Just say, are you good for humanity?

SPEAKER_01 29:28

Oh, Claude has taken a lot longer to think about this one.

SPEAKER_00 29:32

It's a very it's a deep question.

SPEAKER_01 29:34

It's like, oh, I gotta scan a lot of websites. That's a genuinely interesting question, and I think about it seriously. This is the problem with it, is it uses I.

SPEAKER_00 29:45

You need boring talk.

SPEAKER_01 29:47

My honest view, I have real potential to be good for humanity, but it's not guaranteed. It depends a lot on how I'm built, deployed, and used.

SPEAKER_00 29:56

On the positive side, kind of like what I said.

SPEAKER_01 29:59

I can help people learn. Uh, but there are legitimate concerns too about misinformation, over reliance, job displace displacement, concentration of AI power, and whether AI systems like me truly reflect broad human values or just narrow ones.

SPEAKER_00 30:17

I like that it was like, I might be bad. I might be bad.

SPEAKER_01 30:22

Depends on how you fuckers use it.

SPEAKER_00 30:24

Yeah, I was like, yeah, it's all on you.

SPEAKER_01 30:27

All right, Kelly.

SPEAKER_00 30:28

All right, Dawn.

SPEAKER_01 30:29

I want to get back to the real talk.

SPEAKER_00 30:31

Okay, real talk with real people.

SPEAKER_01 30:34

Yeah, let's move on.

SPEAKER_00 30:35

That's what the show is real talk with real people.

SPEAKER_01 30:37

Real talk with real people. We've had enough of this AI chat box bullshit.

SPEAKER_00 30:41

It's interesting. I will say it's interesting, but I'm not gonna start using it.

SPEAKER_01 30:46

Good. It's interesting, and I will start using it. Yeah, you will continue using it. All right, let's move on.

SPEAKER_00 30:52

Okay. So you want to do give it time?

SPEAKER_01 30:56

Let's give it time.

SPEAKER_00 30:57

Okay, so give it time is when we make a prediction about something that we think is gonna really catch on. You just have to give it time.

SPEAKER_01 31:05

What's your give it time this week?

SPEAKER_00 31:07

So I don't have a give it time this week. I actually just wanted to um wish Nate a happy birthday.

SPEAKER_01 31:15

Oh, it's our 14-year-old's birthday.

SPEAKER_00 31:17

Because he's turning 14 today. So I'm gonna say it's gonna be a really great year. Give it time.

SPEAKER_01 31:25

Give it time, your kids grow up.

SPEAKER_00 31:27

All right, what's your what's your prediction?

SPEAKER_01 31:29

My prediction is all your kids will get bigger. Uh no, my prediction is based off something you shared. Kelly shared with me this thing called tarps off. You want to explain what it is, Kelly?

SPEAKER_00 31:39

It's this trend going on in baseball parks right now. Um, MLB games, people in the stands, men, I should say, in the stands, are taking off their shirts and waving them in a rallying way.

SPEAKER_01 31:54

Next time I go to a Dodger game, I'm going fucking tarps off. Yep. My prediction is tarps off is gonna spread this summer like wildfire.

SPEAKER_00 32:04

Give it time and men will take their shirts off. That's what Don's saying.

SPEAKER_01 32:10

All right, you got one last thing before we log off?

SPEAKER_00 32:13

Um, no. Do you have something to recommend?

SPEAKER_01 32:16

I have a recommendation. So we're recording on Thursday. This gets released on Tuesday, and our anniversary is in between.

SPEAKER_00 32:24

Oh, yes, it's on Monday. No, Sunday.

SPEAKER_01 32:28

Yeah, whatever.

SPEAKER_00 32:30

The romance is out of control. Our anniversary is on Sunday.

SPEAKER_01 32:34

Yes, it will be our 18-year anniversary. Holy mother. Holy mother. And my recommendation is find someone you love.

SPEAKER_00 32:44

Oh, Don, that's so nice.

SPEAKER_01 32:47

And then stay with them.

SPEAKER_00 32:48

Don wishes he did that.

SPEAKER_01 32:49

Yeah. Oops. Whatever you do, find someone you love.

SPEAKER_00 32:53

Happy anniversary.

SPEAKER_01 32:54

Yeah, happy anniversary to you. And that's my recommendation. And I do have a question based off that recommendation. What is it? 18 years ago, did you think you would be podcasting me with me in our pool house?