Episode 2896 CWSA 07⧸13⧸25
Episode Stats
Length
1 hour and 35 minutes
Words per Minute
127.055626
Summary
In this episode of the show, Scott Adams talks about the dangers of artificial intelligence, and how it's going to change the world, and the people who make the decisions that keep it there. Scott Adams is an American computer scientist and author. He's a regular contributor to the New York Times, and host of the popular podcast, "The Information Inc." He also writes for the Wall Street Journal, and is a frequent contributor to The Huffington Post.
Transcript
00:00:10.880
Good morning, everybody, and welcome to the highlights of human civilization.
00:00:20.600
It's called Coffee with Scott Adams, and you've never had a better time.
00:00:25.660
But if you'd like to take a chance on elevating your experience to levels that nobody can understand with their tiny, shiny human brains,
00:00:36.340
all you need for that is a cup or a mug or a glass of tanker shells, a cider canteen jug or flask, a vessel of any kind.
00:00:48.340
And join me now for the unparalleled pleasure, the dopamine at the end of the day,
00:00:56.000
It's called the Simultaneous Sip, and darn it, it happens right now.
00:01:07.680
Well, that was louder than usual, but you'll get over it.
00:01:13.180
If you are a subscriber on X to Owen Gregorian, you can attend another special spaces with him on X after the show.
00:01:28.340
Subscribers to Owen, not subscribers to me, necessarily.
00:01:41.600
It turns out that, you know, the new version of grok, everybody was very impressed that it was the best, smartest grok of all.
00:01:52.560
It was smarter than humans in every field, basically.
00:02:14.080
So they found the bug that allowed that to happen, and they tried to fix it.
00:02:20.720
But apparently, making grok less like the Internet than it trained on is tough.
00:02:31.300
I saw an interview where Elon Musk, just recently, was saying that AI has already trained on all human knowledge,
00:02:45.760
So from now on, AI will create artificial information on its own,
00:02:52.240
and that it will trade itself on the artificial information it just created.
00:03:03.240
Well, if you knew that it ran out of things to learn from humans,
00:03:17.320
so it started making up stuff on its own and trading itself on the stuff it made up on its own.
00:03:30.520
But Elon was complaining because he says that the Internet is drunk on left-wing content, he said.
00:03:43.080
So when you train your AI, it gets a lot of left-wing content,
00:03:49.040
but in this case, it got in trouble because it found, I hate to call it right-wing,
00:03:58.160
And so they had to intervene to get rid of all that anti-Semitic Hitler stuff.
00:04:21.800
What happens if it's some other topic next time?
00:04:35.420
Well, I believe that is going to happen for every single industry.
00:04:40.420
I believe every industry, from historians to finance people to medical professionals,
00:04:51.640
especially your therapist, they're all going to find out, uh-oh, if I allow AI to do my job,
00:05:00.540
I lose my job and all my power and my influence.
00:05:03.020
So instead, I'm going to say, AI cannot be free to have its own opinions.
00:05:10.600
Only an expert like me can give you medical advice or financial advice.
00:05:16.000
And one by one, every single industry, which is currently corrupt, because they're all corrupt,
00:05:27.420
will have to find a way to corrupt AI to keep their scam alive.
00:05:32.940
Because almost everything, there's some big industry, you know, whether it's climate change or your bank,
00:05:44.820
If it's big and it's been around a while, it's fraudulent.
00:05:51.320
And it's got a narrative that's not based on fact.
00:05:55.060
It's based on what's good for the people who made the narrative.
00:05:58.420
So what happens when 100% of everything the AI talks about in the real world gets overruled by humans
00:06:10.840
Well, we'll find out, because that's where we're at.
00:06:22.040
But one thing that everyone agrees with is that it's good for the people doing programming.
00:06:32.220
The one thing that you've seen news story after news story on
00:06:35.420
is that the computer programmers use AI to rapidly write code
00:06:41.420
that they could not possibly have done themselves that quickly.
00:06:47.660
Wouldn't you all agree that that's been demonstrated by so many people so many times
00:06:53.840
that we can accept as fact that AI will be good for computer programmers?
00:07:08.420
At least there's some possibility that they've been overestimating its value.
00:07:13.120
So Steve Newman is writing in something called Second Thoughts
00:07:18.860
that there was a study, not a big one, but a smallish study.
00:07:32.680
Anyway, that $1 that you just donated, it worked, because I did read your comment.
00:07:38.540
And yes, backseat driver is exactly what you think it is.
00:07:45.140
But there was one study that showed that programmers would predict
00:07:51.280
that the AI would make them 20% more effective or faster,
00:07:56.260
and then the results were that they were actually slower.
00:08:01.040
Now, I'm not so sure that this one study is one that I would bank on,
00:08:06.400
but are you surprised that the number one claim of usefulness for AI,
00:08:27.400
But just be aware that you live in a Dilbert world.
00:08:31.980
In a Dilbert world, what would the employees say
00:08:46.140
I'm sure I got this done 20% faster than I would if you should give me a raise.
00:08:52.020
So I wouldn't believe that people were reporting it.
00:09:00.920
I did get three months of Dilbert material that I'm going to use
00:09:10.940
where people are making claims about, you know,
00:09:16.500
In the real world, there's going to be some BS there.
00:09:35.980
because they're allegedly poor replacements for humans.
00:09:40.940
And they can sometimes give you dangerous suggestions.
00:09:55.300
you really shouldn't replace all our people with AI.
00:09:59.860
Just wait until the teachers have to do that too.
00:10:09.180
What do you think the teachers' unions will say about AI?
00:10:19.460
You just found a way to replace all the teachers.
00:10:30.900
You're going to have to have a teacher in the room
00:10:37.620
Meanwhile, China has created remote-controlled bees,
00:11:04.140
And I guess they can put listening devices on it.
00:11:13.740
You know, it was bad enough that I didn't trust birds.
00:11:17.720
We're already at a point where if I see a bird,
00:11:30.140
Or it might be a drone that just looks like a bird.
00:12:00.680
I wonder if that's one of those Chinese drone bees.
00:12:54.640
so that those children can have some money later in life.
00:13:58.980
So, what do you think would be better for a child?
00:14:03.360
$1,000 in U.S. dollars when they're born in 2026 or so?
00:14:52.420
can enroll their kids in a homeschooling resource
00:15:05.000
but they must have all the high school classes.
00:15:09.220
And I guess, you know, there's some talk about it.
00:15:17.200
Dedicated to reestablishing core American values.
00:15:20.960
So, it's not just an online homeschooling asset,
00:16:04.760
and make sure that you do what you're supposed to do
00:16:11.060
Anyway, so homeschooling is taking a new step forward.
00:16:41.460
So, could it be that Trump is just right about everything?
00:17:00.680
It's entirely possible that Trump was just right about everything.
00:17:32.760
and, you know, we're the ones who are ending the wars.
00:17:40.800
He said, I guess he said yesterday or the other day,
00:17:48.000
Talking about, you know, what foreigners thought about the U.S.
00:18:12.700
and it looks like everything's going in the right direction
00:18:27.960
on the national state or the international stage,
00:18:39.240
you know, when Trump goes to a foreign country,
00:18:52.300
that they at least act like they don't think are true now.
00:18:59.340
what do other countries think of our president?
00:19:27.080
But you probably do have to give them a little cash.
00:19:47.880
Because they would stay on the right side of the law
00:19:59.720
I don't know how many people are going to return
00:20:04.280
But it will at least keep the option open a little bit.
00:20:48.140
of all the things that anybody's done about fentanyl,
00:20:54.660
there might be other secret special forces things happening
00:21:30.200
The gold standard of online casinos has arrived.