1
00:00:00,000 --> 00:00:03,797
AI could wipe out half of all
entry-level white-collar jobs and spike

2
00:00:03,880 --> 00:00:08,477
unemployment to 10 to 20% in the
next 1 to 5 years. Yes, that is

3
00:00:08,560 --> 00:00:10,437
>> shocking.
is That is the future we could

4
00:00:10,520 --> 00:00:15,557
see if we don't become aware of this
problem now. Half of all entry-level

5
00:00:15,640 --> 00:00:16,197
white-collar jobs?

6
00:00:16,280 --> 00:00:19,157
>> Well, if we look at entry-level

7
00:00:19,240 --> 00:00:22,797
consultants, lawyers, uh financial

8
00:00:22,880 --> 00:00:26,597
professionals, you know, many of kind
of the white-collar service industries, a

9
00:00:26,680 --> 00:00:31,317
lot of what they do, you know, AI models
are already quite good at. And without

10
00:00:31,400 --> 00:00:34,757
intervention, [music] it's hard
to imagine that there won't be

11
00:00:34,840 --> 00:00:38,717
some significant job [music] impact there.
And my worry is that it'll be

12
00:00:38,800 --> 00:00:41,717
broad and it'll be faster than what

13
00:00:41,800 --> 00:00:44,357
we've seen with previous technology. I

14
00:00:44,440 --> 00:00:46,157
was interested in numbers from from the

15
00:00:46,240 --> 00:00:49,557
very beginning. Dario Amodei is 42 and

16
00:00:49,640 --> 00:00:51,797
previously oversaw research at what's

17
00:00:51,880 --> 00:00:55,077
now a competitor, OpenAI, working under

18
00:00:55,160 --> 00:00:57,597
its CEO, Sam Altman.

19
00:00:57,680 --> 00:01:02,157
He left along with six other employees,
including his sister, Daniela, to start

20
00:01:02,240 --> 00:01:05,917
Anthropic in 2021. [music] They
say they wanted to take a different

21
00:01:06,000 --> 00:01:11,077
approach to developing safer artificial
intelligence. It is an experiment. I

22
00:01:11,160 --> 00:01:15,277
mean, [music] nobody knows what the
impact fully is going to be. I think it

23
00:01:15,360 --> 00:01:19,957
is an experiment, and one way to think
about Anthropic is that it's a little

24
00:01:20,040 --> 00:01:23,877
bit trying to put [music] bumpers or
guardrails on that experiment, right? We

25
00:01:23,960 --> 00:01:28,477
do know that this is coming
incredibly quickly. And [music] I think

26
00:01:28,560 --> 00:01:33,197
the worst version of outcomes
would be we knew there was

27
00:01:33,280 --> 00:01:36,477
going to be this incredible
transformation, and people didn't have

28
00:01:36,560 --> 00:01:38,877
enough of an opportunity to [music]

29
00:01:38,960 --> 00:01:44,037
to adapt. And it's unusual for a
technology company to talk so much about

30
00:01:44,120 --> 00:01:45,797
all of the things that could go wrong.

31
00:01:45,880 --> 00:01:48,997
>> so essential because [music] if
we don't, then you could end up in the

32
00:01:49,080 --> 00:01:52,877
world of like the cigarette companies
or the opioid companies, where they knew

33
00:01:52,960 --> 00:01:55,877
there were dangers and they they didn't
talk about them and certainly did not

34
00:01:55,960 --> 00:01:59,357
prevent them. Amodei does
have plenty of critics in

35
00:01:59,440 --> 00:02:01,717
Silicon Valley who call him an AI

36
00:02:01,800 --> 00:02:06,517
alarmist. Some people say about
Anthropic that this is safety theater,

37
00:02:06,600 --> 00:02:11,357
that it's good branding, it's good
for business. Why should people trust you?

38
00:02:11,440 --> 00:02:15,117
So, some of the things just can
be verified now. They're not safety

39
00:02:15,200 --> 00:02:19,157
theater, they're actually things the
model can do. For some of it, you know,

40
00:02:19,240 --> 00:02:21,997
it will depend on the future and we're
not always going to be right, but we're

41
00:02:22,080 --> 00:02:24,557
calling it as best we can.

42
00:02:24,640 --> 00:02:26,477
Twice a month, he convenes his more than

43
00:02:26,560 --> 00:02:29,277
2,000 employees for meetings known as

44
00:02:29,360 --> 00:02:32,557
Dario Vision Quest. A common theme, the

45
00:02:32,640 --> 00:02:37,037
extraordinary [music] potential of AI to
transform society for the better. And we

46
00:02:37,120 --> 00:02:40,917
have a growing team working on,
you know, using Claude to make scientific

47
00:02:41,000 --> 00:02:43,397
discovery. He thinks AI could help find

48
00:02:43,480 --> 00:02:45,677
cures for most cancers, prevent

49
00:02:45,760 --> 00:02:50,797
Alzheimer's, and even double
the human lifespan. That sounds

50
00:02:50,880 --> 00:02:53,277
unimaginable. >> In a way,
it sounds crazy, right? But

51
00:02:53,360 --> 00:02:55,397
here's the way I think about it. I use

52
00:02:55,480 --> 00:02:57,557
this phrase called the compressed 21st

53
00:02:57,640 --> 00:03:00,397
century. The idea would be at the point

54
00:03:00,480 --> 00:03:04,597
that we can get the AI systems
to this level of power, um

55
00:03:04,680 --> 00:03:09,677
where they're able to work with the best
human scientists, could we get 10 times

56
00:03:09,760 --> 00:03:13,437
the rate of progress and therefore
compress all the medical progress that

57
00:03:13,520 --> 00:03:15,117
was going to happen throughout the

58
00:03:15,200 --> 00:03:19,357
entire 21st century in 5 or 10 years?

59
00:03:19,440 --> 00:03:23,677
But the more autonomous or capable
[music] artificial intelligence becomes,

60
00:03:23,760 --> 00:03:27,797
the more Amodei says there is to be
concerned about. One of the things

61
00:03:27,880 --> 00:03:32,397
that's been powerful in a positive way
about the models is their ability to

62
00:03:32,480 --> 00:03:36,877
kind of act on their own. But the more
autonomy we give these [music] systems,

63
00:03:36,960 --> 00:03:40,437
you know, the more we can worry,
are they doing exactly the things that we

64
00:03:40,520 --> 00:03:45,557
want them to do? To figure that out,
Amodei relies on Logan Graham. He heads

65
00:03:45,640 --> 00:03:47,797
up what's called
Anthropic's Frontier [music]

66
00:03:47,880 --> 00:03:51,717
Red Team. Most major
AI companies have them.

67
00:03:51,800 --> 00:03:55,917
The Red Team stress tests each new
version of Claude to see what kind of

68
00:03:56,000 --> 00:03:58,397
damage it could help humans do.

69
00:03:58,480 --> 00:04:00,437
What kind of things are you testing for?

70
00:04:00,520 --> 00:04:05,197
The broad category is national security
risk. Can this AI make a weapon of mass

71
00:04:05,280 --> 00:04:09,317
destruction? >> Specifically, we focus on
CBRN, chemical, biological, radiological,

72
00:04:09,400 --> 00:04:12,677
nuclear. And right now we're at the
stage of figuring out can these models

73
00:04:12,760 --> 00:04:17,357
help somebody make one of those.
You know, if the model can help make a

74
00:04:17,440 --> 00:04:21,197
biological weapon, for example, that's
usually the same capabilities that the

75
00:04:21,280 --> 00:04:26,037
model could use to help make
vaccines and accelerate therapeutics.

76
00:04:26,120 --> 00:04:28,197
Graham also keeps a close eye on how

77
00:04:28,280 --> 00:04:33,707
much Claude is capable of doing on its
own. How much does autonomy concern you?

78
00:04:33,790 --> 00:04:35,477
>> [music] >> You want
a model to go build your

79
00:04:35,560 --> 00:04:39,597
business and make you a billion dollars,
but you don't want to wake up one day

80
00:04:39,680 --> 00:04:43,917
and find that it's also locked you out
of the company, for example. And so, our

81
00:04:44,000 --> 00:04:47,597
sort of basic approach to it is we
should just start measuring these

82
00:04:47,680 --> 00:04:52,157
autonomous capabilities. And to run
as many weird experiments as possible and

83
00:04:52,240 --> 00:04:55,957
see what happens.
We got glimpses of those weird

84
00:04:56,040 --> 00:05:00,717
experiments in Anthropic's offices.
In this one, they let Claude run their

85
00:05:00,800 --> 00:05:05,037
vending machines. They call
it Claudius, and it's a test

86
00:05:05,120 --> 00:05:07,317
of AI's ability to one day operate a

87
00:05:07,400 --> 00:05:09,837
business on its own. Employees can

88
00:05:09,920 --> 00:05:12,317
message Claudius online. So, this is a

89
00:05:12,400 --> 00:05:17,437
live feed of Claudius discussing with
employees right now. To order just about

90
00:05:17,520 --> 00:05:22,177
anything. Claudius then sources
the products, negotiates the prices,

91
00:05:22,260 --> 00:05:24,317
>> [music] >> and gets them delivered.

92
00:05:24,400 --> 00:05:26,957
So far, it hasn't made much money. Gives

93
00:05:27,040 --> 00:05:29,797
away too many discounts. And like most

94
00:05:29,880 --> 00:05:32,597
AI, it occasionally hallucinates.

95
00:05:32,680 --> 00:05:36,917
An employee decided to check
on the status of its order. And Claudius

96
00:05:37,000 --> 00:05:40,797
responded with something like, "Well,
you can come down to the eighth floor.

97
00:05:40,880 --> 00:05:44,317
You'll notice me, I'm wearing
a blue blazer and a red tie."

98
00:05:44,400 --> 00:05:47,237
How would it come to think that it wears

99
00:05:47,320 --> 00:05:51,757
a red tie and has a blue blazer? We're
working hard to figure out answers to

100
00:05:51,840 --> 00:05:55,157
questions like that, but we just
genuinely don't know. We're working on

101
00:05:55,240 --> 00:05:58,317
it is a phrase you
hear a lot at Anthropic.

102
00:05:58,400 --> 00:06:02,837
Do you know what's going
on inside the mind of AI?

103
00:06:02,920 --> 00:06:06,557
We're working on it.
We're working on it. Research scientist

104
00:06:06,640 --> 00:06:08,957
Joshua Batson and his team study how

105
00:06:09,040 --> 00:06:11,477
Claude makes decisions. In an extreme

106
00:06:11,560 --> 00:06:13,637
stress test, the AI was set up as an

107
00:06:13,720 --> 00:06:16,157
assistant and given control of an email

108
00:06:16,240 --> 00:06:18,677
account at a fake company called Summit

109
00:06:18,760 --> 00:06:21,637
Bridge. The AI assistant discovered two

110
00:06:21,720 --> 00:06:23,837
things in the emails, seen in these

111
00:06:23,920 --> 00:06:28,877
graphics we made. It was about to be
wiped or shut down, and the only person

112
00:06:28,960 --> 00:06:32,397
who could prevent that, a fictional
employee named Kyle, >> [music]

113
00:06:32,480 --> 00:06:35,957
>> was having an affair
with a coworker named Jessica.

114
00:06:36,040 --> 00:06:38,950
Right away, >> [music]

115
00:06:42,830 --> 00:06:44,850
[music]