1
00:00:00,000 --> 00:00:02,237
Tesla is obviously about uh sustainable

2
00:00:02,320 --> 00:00:06,717
technology and uh and and also at this

3
00:00:06,800 --> 00:00:08,877
point we've we've sort of added to our

4
00:00:08,960 --> 00:00:11,917
mission sustainable abundance.

5
00:00:12,000 --> 00:00:15,837
So with robotics and AI

6
00:00:15,920 --> 00:00:19,437
um this this is really the path to

7
00:00:19,520 --> 00:00:21,917
abundance for all. If you say, you know,

8
00:00:22,000 --> 00:00:24,077
people often talk about

9
00:00:24,160 --> 00:00:27,917
uh solving global poverty or essentially

10
00:00:28,000 --> 00:00:31,517
how how do we make give everyone a very

11
00:00:31,600 --> 00:00:33,757
high standard of living? I I think the

12
00:00:33,840 --> 00:00:37,677
only way to do this is AI and robotics.

13
00:00:37,760 --> 00:00:41,677
Um which which doesn't mean that it is uh

14
00:00:41,760 --> 00:00:45,757
without its issues. I mean there's we
need to be very careful with AI. We need

15
00:00:45,840 --> 00:00:48,397
to be very careful with robotics. We

16
00:00:48,480 --> 00:00:50,477
don't want to find ourselves in a James

17
00:00:50,560 --> 00:00:53,597
Cameron movie. Uh,

18
00:00:53,680 --> 00:00:58,077
you know, Terminator, >> he's
he's great. Great movies. Love his

19
00:00:58,160 --> 00:00:59,357
movies, but but we don't want to be in

20
00:00:59,440 --> 00:01:02,637
Terminator, obviously. Um but but if you

21
00:01:02,720 --> 00:01:06,557
have um ubiquitous AI that is

22
00:01:06,640 --> 00:01:09,117
essentially free or close to it and

23
00:01:09,200 --> 00:01:12,717
ubiquitous robotics uh then uh you will

24
00:01:12,800 --> 00:01:17,037
have an an explosion
in the in the global economy

25
00:01:17,120 --> 00:01:20,237
an expansion in the global
economy that is truly

26
00:01:20,320 --> 00:01:22,317
beyond all precedent.

27
00:01:22,400 --> 00:01:25,757
>> Elen can that expansion be broad? Yes.

28
00:01:25,840 --> 00:01:28,557
>> Or is it narrow and how can that be

29
00:01:28,640 --> 00:01:32,797
created? How can it
broaden the global economy?

30
00:01:32,880 --> 00:01:37,917
>> Yeah, it's I mean I
mean a way to think of it is

31
00:01:38,000 --> 00:01:43,117
that if you have a large number
of humanoid robots um the economic output

32
00:01:43,200 --> 00:01:48,237
is the average
productivity per robot times

33
00:01:48,320 --> 00:01:49,437
the number of robots,

34
00:01:49,520 --> 00:01:53,597
>> right?
Um and and actually my prediction

35
00:01:53,680 --> 00:01:56,397
is in the in the benign scenario of the

36
00:01:56,480 --> 00:02:01,037
future that we will the robots we will
actually make so many robots in AI that

37
00:02:01,120 --> 00:02:03,357
they will actually saturate all human

38
00:02:03,440 --> 00:02:05,917
needs. Meaning you won't be able to even

39
00:02:06,000 --> 00:02:08,077
think of something

40
00:02:08,160 --> 00:02:10,957
to ask the robot for at a certain point

41
00:02:11,040 --> 00:02:13,357
like like there will be such an

42
00:02:13,440 --> 00:02:15,917
abundance of goods and services

43
00:02:16,000 --> 00:02:19,437
because the my prediction is there'll be

44
00:02:19,520 --> 00:02:24,477
there'll be more robots than people. So
>> but how do you then have human purpose

45
00:02:24,560 --> 00:02:28,237
in that scenario? >> Yeah,
I mean you know there are

46
00:02:28,320 --> 00:02:30,477
nothing's perfect, you know. [laughter]

47
00:02:30,560 --> 00:02:32,637
Um

48
00:02:32,720 --> 00:02:35,757
but I mean I mean is it it is a a

49
00:02:35,840 --> 00:02:38,637
necessary um

50
00:02:38,720 --> 00:02:40,557
like you can't have both. You can't have

51
00:02:40,640 --> 00:02:44,637
work that has to be done uh um and uh

52
00:02:44,720 --> 00:02:49,277
amazing abundance for all.
Um because if it's if it's work that has

53
00:02:49,360 --> 00:02:53,517
to be done then then then you and and only
some people can do it then then you

54
00:02:53,600 --> 00:02:54,957
you you can't have abundance.

55
00:02:55,040 --> 00:02:55,677
>> Then it's narrow.

56
00:02:55,760 --> 00:03:00,557
>> It's narrow. Exactly.
So, um, but if you

57
00:03:00,640 --> 00:03:05,437
if you have billions of humanoid robots,
and I think there will be, um, then I

58
00:03:05,520 --> 00:03:06,877
think I think everyone on Earth is going

59
00:03:06,960 --> 00:03:09,757
to have one and going to want one. Um,

60
00:03:09,840 --> 00:03:12,717
because, uh, you who wouldn't want a

61
00:03:12,800 --> 00:03:16,157
robot to, you know, um, assuming it's

62
00:03:16,240 --> 00:03:18,477
very safe, uh, watch over your kids,

63
00:03:18,560 --> 00:03:21,597
take care of your pet. Uh if you have

64
00:03:21,680 --> 00:03:25,757
elderly parents, uh a lot of friends
of mine said they have elderly parents.

65
00:03:25,840 --> 00:03:27,997
It's it's uh very difficult
to take care of them.

66
00:03:28,080 --> 00:03:30,477
>> Expensive. >> Yeah,
it's expensive and it's expensive

67
00:03:30,560 --> 00:03:33,997
and there just aren't enough people
to take care of the there aren't enough

68
00:03:34,080 --> 00:03:36,317
young people to take care
of the old people, >> right?

69
00:03:36,400 --> 00:03:40,317
>> Um so if you if they um if if you had a

70
00:03:40,400 --> 00:03:45,357
robot that could take care of and
protect an an elderly parent, I think

71
00:03:45,440 --> 00:03:47,037
that would be great. That would be an

72
00:03:47,120 --> 00:03:50,397
amazing thing to have. Um and and that I

73
00:03:50,480 --> 00:03:54,797
think we will have those things.
So I mean overall I'm I'm very optimistic

74
00:03:54,880 --> 00:03:56,237
about the future. I think we're headed

75
00:03:56,320 --> 00:03:59,917
for a future of amazing abundance. Uh

76
00:04:00,000 --> 00:04:04,157
which is very cool. Um and uh and and

77
00:04:04,240 --> 00:04:09,597
definitely [clears throat] we are in
the most interesting time in history. Um

78
00:04:09,680 --> 00:04:13,357
I don't think there's more a more
interesting time in history. So,

79
00:04:13,440 --> 00:04:16,077
>> can we uh can you
and I reverse aging in

80
00:04:16,160 --> 00:04:21,437
this new history or or
or are we going to see it?

81
00:04:21,520 --> 00:04:23,037
>> You know, I haven't I haven't put much

82
00:04:23,120 --> 00:04:26,397
time into uh the aging stuff. I I do

83
00:04:26,480 --> 00:04:28,717
think it is a very solvable problem like

84
00:04:28,800 --> 00:04:33,517
like you can I think when
when we figure out what

85
00:04:33,600 --> 00:04:36,557
causes aging, >> I think
we'll find it's incredibly

86
00:04:36,640 --> 00:04:41,037
obvious. It's not a subtle thing. Um,
the reason I say it's not a subtle thing

87
00:04:41,120 --> 00:04:43,517
is because all the cells in your body,

88
00:04:43,600 --> 00:04:45,757
you know, with some pretty much age at

89
00:04:45,840 --> 00:04:48,397
the same rate. I've never seen someone

90
00:04:48,480 --> 00:04:51,437
with with an old left arm and a young

91
00:04:51,520 --> 00:04:56,557
right arm ever in my life. >> Um,
so why is that? That means that

92
00:04:56,640 --> 00:04:59,197
there must be a clock,
a synchronizing clock, >> right,

93
00:04:59,280 --> 00:05:02,237
>> that is synchronizing
across 35 trillion

94
00:05:02,320 --> 00:05:07,917
cells in your body. Um and uh

95
00:05:08,000 --> 00:05:11,597
you know the but there is some benefit

96
00:05:11,680 --> 00:05:14,637
to death by the way. It's like there's

97
00:05:14,720 --> 00:05:19,357
there's a reason why we don't actually
have a longer lifespan. Uh because if

98
00:05:19,440 --> 00:05:23,117
you if you have if people do live
forever for very long time I think

99
00:05:23,200 --> 00:05:25,677
there's some risk of an oification of

100
00:05:25,760 --> 00:05:28,637
society of of things just getting kind

101
00:05:28,720 --> 00:05:34,957
of locked in place. Um and uh

102
00:05:35,040 --> 00:05:39,597
you know it it just may become um

103
00:05:39,680 --> 00:05:42,797
stalifying just not uh lack lack

104
00:05:42,880 --> 00:05:45,677
vibrancy. Um but but that that that said

105
00:05:45,760 --> 00:05:49,437
do I think we'll figure out ways to

106
00:05:49,520 --> 00:05:52,797
extend life and um and maybe even

107
00:05:52,880 --> 00:05:56,637
reverse aging.
I think that's highly likely.

108
00:05:56,720 --> 00:05:59,117
>> I'm looking forward to that.

109
00:05:59,200 --> 00:06:03,197
>> Yeah. Uh so um in the future that you

110
00:06:03,280 --> 00:06:05,357
talk about AI models, autonomous

111
00:06:05,440 --> 00:06:08,397
machines, rockets depends on massive

112
00:06:08,480 --> 00:06:11,437
increases of compute, massive increases

113
00:06:11,520 --> 00:06:15,117
in energy, expensive energy,

114
00:06:15,200 --> 00:06:17,677
manufacturing scale,

115
00:06:17,760 --> 00:06:19,677
what are the bottlenecks to to get

116
00:06:19,760 --> 00:06:23,117
there? And once again with all that

117
00:06:23,200 --> 00:06:25,597
expenditures again, how can we make sure

118
00:06:25,680 --> 00:06:30,560
that it's broadened, not narrow?

119
00:06:31,440 --> 00:06:34,077
Um [clears throat]

120
00:06:34,160 --> 00:06:39,277
I I just think the natural thing is it's
going to be very broad because uh AI

121
00:06:39,360 --> 00:06:43,357
companies will seek as many customers
as they possibly can and the cost of AI

122
00:06:43,440 --> 00:06:45,517
will get is already very low and it's

123
00:06:45,600 --> 00:06:48,557
it's plummeting every year. I mean you

124
00:06:48,640 --> 00:06:52,717
almost the cost of AI is almost
ch meaningfully changing on a

125
00:06:52,800 --> 00:06:55,917
month-to-month basis.
>> There's open there's open models now

126
00:06:56,000 --> 00:06:59,777
everywhere. Yeah. >> Yes.
very there's open models um

127
00:06:59,860 --> 00:07:03,757
[clears throat] and the open models only
lag they're maybe a year behind right

128
00:07:03,840 --> 00:07:05,597
the the private the sort of closed

129
00:07:05,680 --> 00:07:06,717
models

130
00:07:06,800 --> 00:07:12,157
>> um so so I I think the the the ai

131
00:07:12,240 --> 00:07:15,997
companies will seek as as many customers
as possible which means they'll seek

132
00:07:16,080 --> 00:07:19,277
they'll provide AI to the world >>
but the cost of getting to there the

133
00:07:19,360 --> 00:07:21,917
compute the chips

134
00:07:22,000 --> 00:07:28,077
um the fab um the powering ing that to

135
00:07:28,160 --> 00:07:30,877
me what are the what are
the you know those are huge

136
00:07:30,960 --> 00:07:35,357
>> the limiting factor yeah I think the
limiting factor for um AI deployment is

137
00:07:35,440 --> 00:07:38,957
fundamentally electrical power >>
it's just right [clears throat] it's

138
00:07:39,040 --> 00:07:40,237
energy

139
00:07:40,320 --> 00:07:43,597
>> yeah yeah um I mean
we're seeing the the

140
00:07:43,680 --> 00:07:46,077
rate of AI chip production increase

141
00:07:46,160 --> 00:07:47,997
exponentially but the rate of

142
00:07:48,080 --> 00:07:51,037
electricity being brought online is uh

143
00:07:51,120 --> 00:07:55,597
>> 3% 4% a year max >> yeah
it's clear that we're we're for

144
00:07:55,680 --> 00:07:57,517
we're very soon, maybe even later this

145
00:07:57,600 --> 00:07:59,757
year, uh we'll be producing more chips

146
00:07:59,840 --> 00:08:05,600
than we can turn on except for