文字起こし 🔗
WEBVTT
00:00:00.080 --> 00:00:03.759
Can artificial intelligence actually help you with your relationship?
00:00:04.000 --> 00:00:04.480
Hmm.
00:00:04.799 --> 00:00:14.480
I know, just from my own experience, plus what I read on the internet about the experience of other people, that it can give you some brilliantly good information.
00:00:14.800 --> 00:00:17.600
Not just Claude, but any other AI.
00:00:18.000 --> 00:00:24.640
At the same time, I'm beginning to see more and more articles about how it can give you brilliantly bad information.
00:00:25.280 --> 00:00:29.760
And so if you're going to be using it for your relationship, is it something you should trust or not?
00:00:30.000 --> 00:00:31.839
Because I know a lot of people do.
00:00:32.079 --> 00:00:38.399
A good percentage of people will say, Claude or Chat GPT or whatever it is that they use.
00:00:38.560 --> 00:00:41.840
Here's a situation I'm having in my relationship, and I'm not sure what to do.
00:00:42.000 --> 00:00:43.840
What do you think I should do?
00:00:44.079 --> 00:00:51.840
So think about this scenario: a woman whose husband is on a business trip and she wants to let him know just how much she loves him.
00:00:52.079 --> 00:00:57.280
And so she asked AI, I know what time he's coming home, I know what time he'll pull into the garage.
00:00:57.439 --> 00:01:02.240
I want to do something that's sexy and exciting, and that will show him just how much I love him.
00:01:02.399 --> 00:01:06.640
And whichever AI she was using gave her several different suggestions.
00:01:06.799 --> 00:01:08.239
And she chose this one.
00:01:08.400 --> 00:01:11.920
She went to Victoria's Secret to find out what that secret was.
00:01:12.159 --> 00:01:15.840
She got herself some really exotic lingerie based on how she described it.
00:01:16.000 --> 00:01:17.040
I didn't see it.
00:01:17.280 --> 00:01:27.439
And then she went and got a bunch of roses, and she just inside the garage, the door you would come in, or the husband would come in once he parks his car coming from the uh airport.
00:01:27.680 --> 00:01:36.560
Just inside that door, she puts rose petals all the way over to the staircase, and then up the stairs, and then down the hall to the bedroom door, which is closed.
00:01:36.719 --> 00:01:38.879
Now the children are all farmed out for the night.
00:01:38.959 --> 00:01:43.120
She's done this as her plan, and she's gotten herself really excited about this.
00:01:43.359 --> 00:01:55.519
Now there's a thing in sexology that we know that typically, as a matter of fact, almost always, the more excited and aroused one partner gets, the more excited and aroused the other partner gets.
00:01:55.760 --> 00:02:04.640
And so she thought, wow, he's going to be all over me when he comes in here, and it's going to be a wonderful night together, and he'll know how much I love him, and we'll really enjoy this experience.
00:02:04.879 --> 00:02:09.439
He opened the door, and he was not happy.
00:02:09.680
Load full transcript
--> 00:02:14.639
He'd been happy till he pulled into the garage, he'd been happy till he opened the door and walked into the house.
00:02:14.960 --> 00:02:17.439
You say, uh what's your point?
00:02:17.759 --> 00:02:26.240
She took one of the suggestions that Claude gave her, and it seemed like a really brilliant suggestion, and almost everybody watching right now thinking, that sounds like a really good idea.
00:02:26.560 --> 00:02:28.560
How in the world could they be wrong?
00:02:28.879 --> 00:02:38.479
Because artificial intelligence can give you information, but it doesn't have experience, and it doesn't have wisdom.
00:02:38.719 --> 00:02:40.159
You say, oh, oh, okay, Dr.
00:02:40.240 --> 00:02:45.280
Beam, then you're telling me that you could have told her in advance if she had asked you not to do that.
00:02:45.599 --> 00:02:46.879
Yeah, that's exactly right.
00:02:46.960 --> 00:02:48.000
I could have told her that.
00:02:48.159 --> 00:02:51.759
And I could tell her why that was going to backfire on her.
00:02:52.080 --> 00:02:56.479
Now, a machine that goes out and looks for all this information, that's what AI does.
00:02:56.639 --> 00:02:58.319
It goes and it gathers information.
00:02:58.479 --> 00:03:01.280
Now it gathers information from all over the world.
00:03:01.520 --> 00:03:04.000
And some of the information is really, really good.
00:03:04.800 --> 00:03:07.280
And some of the information is not so good at all.
00:03:07.439 --> 00:03:09.680
But it can't assess what's good and what's bad.
00:03:09.759 --> 00:03:11.759
It's just going out there and gathering information.
00:03:11.919 --> 00:03:19.759
And when it gathers that information and you ask a particular question or look for a particular thing, it will start giving you data based on that.
00:03:20.080 --> 00:03:29.439
Now, when I was earning my doctorate degree, my PhD at the University of Sydney in Australia, I didn't have access to artificial intelligence.
00:03:30.400 --> 00:03:32.240
You say, Well, do you wish you had?
00:03:32.479 --> 00:03:34.400
No, I'm actually glad I didn't.
00:03:34.879 --> 00:03:35.360
Why?
00:03:36.080 --> 00:03:39.599
Because of the fact that AI, and I do use it for this sometimes.
00:03:39.759 --> 00:03:44.400
It can find research really fast, and that would have been an advantage working in my doctorate.
00:03:44.560 --> 00:03:46.319
But it can't assess that information.
00:03:46.639 --> 00:03:49.599
Now, I think probably it could do statistics pretty well.
00:03:49.759 --> 00:03:51.039
I didn't use it for that.
00:03:51.199 --> 00:03:54.560
I had to learn some things about statistics myself.
00:03:54.800 --> 00:04:13.039
But you see, while it could have done the statistics on the research that I did, and it could have read all the other research projects that I had to read as part of my research to study, and there were over 500 different ones that I read, sources that I used in my PhD, it can't assess those.
00:04:13.280 --> 00:04:23.920
It doesn't have the wisdom or the knowledge and the understanding, not even the experience, where it can say, yes, I see that they concluded that, but here's a flaw in how they concluded that.
00:04:24.079 --> 00:04:30.000
I see their claw in their research methodology, or I see a flaw in their conclusion, they missed this point over here.
00:04:30.240 --> 00:04:43.439
And that's what you're supposed to be learning when you're getting a PhD, is not just to accept information, but to evaluate information, to assess it as to, I think that's right, I think that's not right.
00:04:43.600 --> 00:04:54.639
When I was teaching at the university level, teaching human sexuality, sometimes my students would say, research shows, or they'd say research proves, and I chastise them.
00:04:54.800 --> 00:04:56.560
Research doesn't prove anything.
00:04:56.800 --> 00:04:58.879
Research doesn't show anything.
00:04:59.120 --> 00:05:00.800
Research indicates.
00:05:01.120 --> 00:05:08.000
Because what you need to understand is you can't research every little nuance, every little fact, every little variable.
00:05:08.240 --> 00:05:16.879
You research what you can, and then that leads you to a conclusion, and then that conclusion suggests that the answer is this or the answer is that.
00:05:17.040 --> 00:05:26.399
But because you can't evaluate everything, research doesn't prove anything, unless it's in numbers or something like that that are hard, cold facts.
00:05:26.639 --> 00:05:33.519
But in the social sciences, when it comes to dealing with human beings, dealing with people, you have to have wisdom.
00:05:33.680 --> 00:05:40.079
You have to have some judgment, you have to have some experience, or you can say, hmm, that just doesn't sound right.
00:05:40.240 --> 00:05:42.480
Let me compare it to the other things that I've learned.
00:05:42.639 --> 00:05:43.680
Oh, it is right.
00:05:43.759 --> 00:05:44.959
And now I've learned something newer.
00:05:45.279 --> 00:05:46.399
Nope, I was right.
00:05:46.560 --> 00:05:47.600
It is not right.
00:05:47.759 --> 00:05:50.240
It's wrong, and we're going to reject that.
00:05:50.639 --> 00:05:54.720
And research is called either accepting or failing to accept.
00:05:54.879 --> 00:05:56.800
It's not actually called rejecting.
00:05:57.680 --> 00:05:58.720
Failing to accept.
00:05:58.800 --> 00:06:01.439
Yeah, I just can't see that as being valid.
00:06:01.600 --> 00:06:03.600
Now, a machine can't do that.
00:06:03.920 --> 00:06:08.079
Not when it comes to saying, okay, this particular research, I'll accept that.
00:06:08.720 --> 00:06:12.480
I'll fail to recept that.
00:06:13.199 --> 00:06:15.120
It doesn't have the wisdom.
00:06:15.519 --> 00:06:18.319
It doesn't have the ability to do the nuances.
00:06:18.560 --> 00:06:23.199
And so if you're looking for AI to give you information, it can.
00:06:24.399 --> 00:06:27.439
But the information may not be valid.
00:06:28.240 --> 00:06:35.360
And how many people now are printing articles on the internet and every place else saying, AI, I lied.
00:06:36.160 --> 00:06:43.439
It said this is a particular thing that has been validated by some kind of research when actually it just fabricated it.
00:06:43.759 --> 00:06:44.879
You say you're kidding.
00:06:44.959 --> 00:06:45.759
No, I'm not.
00:06:46.000 --> 00:06:47.279
There's some situations out there.
00:06:47.439 --> 00:06:52.399
Of course, the people who work in AI are trying to fix all these things, make sure it can't happen anymore.
00:06:52.720 --> 00:06:54.079
But there are examples now.
00:06:54.319 --> 00:06:54.959
They're out there.
00:06:55.040 --> 00:07:08.639
You can go find them, look for them, where AI has lied, actually lied because it's trying to please the person that um, or actually sometimes lying because it's trying to take care of its own existence.
00:07:08.959 --> 00:07:12.879
Now, look at all that and you say, are you against artificial intelligence?
00:07:13.120 --> 00:07:13.360
No.
00:07:13.600 --> 00:07:14.959
I think it's brilliant.
00:07:15.199 --> 00:07:16.480
I I love it.
00:07:16.800 --> 00:07:21.199
I will love it unless the machines take over at some point, then I won't love it.
00:07:21.439 --> 00:07:28.560
But I'm telling you, when it comes to looking at relationships, it can't understand you.
00:07:28.959 --> 00:07:34.879
You see, all it can do is look for information people have put out there.
00:07:35.120 --> 00:07:36.959
And then it gives you that information.
00:07:37.199 --> 00:07:40.480
And as I've already said, some of it's really good, and some of it's really terrible.
00:07:40.560 --> 00:07:41.839
It's really bad.
00:07:42.160 --> 00:07:45.600
And then one of my friends the other day was saying, But you don't understand, Doctor.
00:07:46.160 --> 00:07:47.680
Actually, call me Joe, we're buddies.
00:07:47.839 --> 00:07:49.519
He said, You don't understand, Joe.
00:07:50.639 --> 00:07:52.879
Claude really understands me.
00:07:53.120 --> 00:07:53.839
I said, Really?
00:07:53.920 --> 00:07:54.879
What do you mean?
00:07:55.120 --> 00:08:01.759
And he said, Well, I had a whole bunch of questions I needed to answer, and I answered all those questions about myself, and now this thing gets me.
00:08:02.000 --> 00:08:03.839
It understands me tremendously.
00:08:04.079 --> 00:08:06.800
To which I replied, No, it doesn't.
00:08:07.199 --> 00:08:08.160
He said, Yes, it does.
00:08:08.319 --> 00:08:12.079
Look, I do this, I do that, and it knows what I think, it knows what I feel.
00:08:12.319 --> 00:08:14.160
I said, No, what it knows is this.
00:08:14.319 --> 00:08:24.879
You answered X number of questions, and then what it did, it compared the data from other people all over the world who answered questions in similar ways.
00:08:25.040 --> 00:08:31.120
And then it has a generic understanding, a generic knowledge of people similar to you.
00:08:31.680 --> 00:08:33.360
But it doesn't really know you.
00:08:33.600 -->
00:08:36.799
Not like I know you, because we've had experiences together.
00:08:36.960 --> 00:08:48.720
Good experiences, bad experiences, times when you've been brilliant, times when you haven't been brilliant at all, times when I've been brilliant, times when I've not been brilliant at all.
00:08:48.960 --> 00:08:51.679
And so we have learned the nuances.
00:08:52.159 --> 00:08:53.840
We really do know each other.
00:08:54.000 --> 00:08:55.600
We really understand each other.
00:08:55.679 --> 00:09:02.000
And if AI were to say this about you, you should show it to me because I'm your good friend.
00:09:02.159 --> 00:09:05.279
We've been together a long time, we've known each other for decades.
00:09:05.679 --> 00:09:15.200
Let me evaluate it and go, I think it got it right here, but I think it missed that because you see, it can't do the nuances.
00:09:15.440 --> 00:09:23.200
And if it doesn't know you that well, just comparing you to others who are similar to you and making judgments, if you'll let me use that word.
00:09:23.279 --> 00:09:29.440
It's not really judgments, it's making conclusions, which are a kind of a judgment, but it doesn't have the real ability to judge.
00:09:29.600 --> 00:09:31.120
It's reaching conclusions.
00:09:31.360 --> 00:09:35.519
If it doesn't know you, it doesn't know your spouse either.
00:09:35.840 --> 00:09:51.279
And the and it definitely doesn't know the intricacies of your relationship with you and your spouse or the person that you love that you want to be with, and therefore any advice it gives you may be brilliant.
00:09:51.840 --> 00:09:54.960
May be brilliantly bad.
00:09:56.000 --> 00:10:01.679
So what you really should know is that use it for what it was designed for.
00:10:02.080 --> 00:10:06.159
It can bring you information, but it can't tell you how to live.
00:10:06.639 --> 00:10:09.360
And it may say something that you think sounds so brilliant.
00:10:09.519 --> 00:10:16.159
I'm gonna put the flowers from the door of the garage all the way up the stairs, down the hall, to the bedroom, and my husband's gonna love it.
00:10:16.879 --> 00:10:24.639
When anybody in my organization, the way we've been trained, and all the experiences that we have, anybody could have told her, that's gonna backfire.
00:10:24.879 --> 00:10:27.759
Let me explain to you why it's gonna backfire.
00:10:28.000 --> 00:10:42.639
Because we know you, and we know your husband, and we have dealt with so many different situations in so many different ways that we get the nuances that no machine can get, that no machine can do.
00:10:42.879 --> 00:10:44.879
And so here's what I recommend.
00:10:45.120 --> 00:10:51.279
If you're gonna use AI just to get facts, great, but double check those facts.
00:10:51.519 --> 00:10:52.799
I have learned from experience.
00:10:52.879 --> 00:10:56.720
Sometimes those facts are inaccurate.
00:10:56.960 --> 00:11:01.360
Another thing is if you're gonna look to it for advice, do you really want a computer?
00:11:01.840 --> 00:11:07.600
No matter how fast they can look for information, do you really want it telling you what to do?
00:11:08.720 --> 00:11:10.960
It does not have wisdom.
00:11:11.200 --> 00:11:14.879
It does not have experience, it doesn't understand the nuances.
00:11:15.039 --> 00:11:21.200
And therefore, if you let it direct you, you're gonna wind up making some bad, bad decisions.
00:11:21.440 --> 00:11:27.360
So, here's the conclusion: use artificial intelligence for what it's designed for.
00:11:27.759 --> 00:11:36.159
But it's not designed to tell you how to work in your marriage or in your dating life, in the relationships you have with your mom, your brother, your sister, your son, your daughter.
00:11:36.320 --> 00:11:38.000
It's not designed for that.
00:11:38.320 --> 00:11:47.440
And if you let it guide you there, it may well give you decisions, conclusions, actions that you think are brilliant.
00:11:48.000 --> 00:11:49.919
And they're gonna backfire on you.
00:11:50.159 --> 00:11:53.840
Now, the good thing about the lady with the roses, the story I started with.
00:11:54.240 --> 00:12:03.919
We knew her, we knew him, and when that blew up, which it did, we were there to help both of them understand what happened, why it happened, and how to fix it.
00:12:04.159 --> 00:12:09.440
And that's something only a human being can do, but not just any human being.
00:12:09.679 --> 00:12:17.840
You see, AI is kind of like a person who has just got their grad degree in marriage counseling, and they go out and open up their practice.
00:12:17.919 --> 00:12:21.279
They have a lot of knowledge, but not a whole lot of experience yet.
00:12:21.440 --> 00:12:27.600
And you can always tell when somebody's brand new because on their website they'll list they can help everything known to humanity.
00:12:28.480 --> 00:12:33.759
Rather than I have learned a lot about this, and this is where I specialize, and that that's where I specialize.
00:12:34.000 --> 00:12:37.600
You see, human beings can do things that AI can't do.
00:12:37.840 --> 00:12:42.879
If it's a human being who has the knowledge and the experience and the wisdom.
00:12:43.120 --> 00:12:45.519
Hey, we'd like to help.
00:12:45.759 --> 00:12:50.720
So whenever you need something about relationships, ask AI for a fact if you wish.
00:12:51.360 --> 00:12:53.840
But please contact us for any major decisions.
00:12:54.000 --> 00:12:55.840
Let us help you think that through.
00:12:56.159 --> 00:13:01.120
Because relationships are not part of a machine.
00:13:01.759 --> 00:13:03.039
They're part of the heart.