WEBVTT

1
00:00:00.208 --> 00:00:01.042
- Vlad, welcome.

2
00:00:01.042 --> 00:00:03.545
Now that Gen AI is so accessible,

3
00:00:03.545 --> 00:00:05.505
does that change the way
that CEOs are thinking

4
00:00:05.505 --> 00:00:07.841
about leveraging it in their own business?

5
00:00:07.841 --> 00:00:09.050
- Yeah, I mean they're being asked

6
00:00:09.050 --> 00:00:10.301
about it all the time, right,

7
00:00:10.301 --> 00:00:12.178
from their boards and
from their employees.

8
00:00:12.178 --> 00:00:14.639
Like, one of the most interesting comments

9
00:00:14.639 --> 00:00:17.142
I got from one of the questions was,

10
00:00:17.142 --> 00:00:19.686
she felt pressured by
her employees who said,

11
00:00:19.686 --> 00:00:21.688
"I can use this stuff in my personal life

12
00:00:21.688 --> 00:00:24.482
and I'm not getting those
tools in my hands at work.

13
00:00:24.482 --> 00:00:25.984
When are you going to change this?"

14
00:00:25.984 --> 00:00:27.736
So they're feeling the pressure.

15
00:00:27.736 --> 00:00:29.195
They're also seeing
the art of the possible

16
00:00:29.195 --> 00:00:30.989
with all the latest tools and releases

17
00:00:30.989 --> 00:00:32.531
that played out over the last few months

18
00:00:32.531 --> 00:00:34.451
that everybody had access to.

19
00:00:34.451 --> 00:00:36.327
So it's for sure on top of the agenda.

20
00:00:36.327 --> 00:00:38.705
And they're all trying to
figure out what their plays are.

21
00:00:38.705 --> 00:00:39.706
- Yeah, trying to figure out...

22
00:00:39.706 --> 00:00:43.126
How do CEOs balance
experimentation and risk

23
00:00:43.126 --> 00:00:45.837
with productivity and
competitive advantage?

24
00:00:45.837 --> 00:00:46.963
- It's all over the place.

25
00:00:46.963 --> 00:00:50.550
Your question asserts that
they have all found a balance,

26
00:00:50.550 --> 00:00:52.635
and the reality is that's not the case.

27
00:00:52.635 --> 00:00:53.470
There's only few of them.

28
00:00:53.470 --> 00:00:54.971
And the ones that do find the balance are

29
00:00:54.971 --> 00:00:57.015
the ones that are curious,

30
00:00:57.015 --> 00:01:00.518
and they're seeing it as a
potential benefit and a tool

31
00:01:00.518 --> 00:01:01.728
usually through the companies

32
00:01:01.728 --> 00:01:03.396
that are more of a growth mindset

33
00:01:03.396 --> 00:01:05.607
who feel like they have
a big bold growth agenda

34
00:01:05.607 --> 00:01:07.692
and they're seeing AI as an ability

35
00:01:07.692 --> 00:01:10.361
to give them scale, right,

36
00:01:10.361 --> 00:01:12.447
and to be able to scale
up faster than the others.

37
00:01:12.447 --> 00:01:15.700
Usually, I see within those
companies that they experiment,

38
00:01:15.700 --> 00:01:17.368
they try to do it in an ethical way,

39
00:01:17.368 --> 00:01:18.870
in a responsible way.

40
00:01:18.870 --> 00:01:20.080
And they're thinking proactively

41
00:01:20.080 --> 00:01:22.415
around building the
right data infrastructure

42
00:01:22.415 --> 00:01:24.334
to be able to get access to the right data

43
00:01:24.334 --> 00:01:25.543
to train these models.

44
00:01:25.543 --> 00:01:26.461
And then they're aggressive

45
00:01:26.461 --> 00:01:28.630
about putting their strongest performers

46
00:01:28.630 --> 00:01:31.716
to actually own the agenda
versus outsourcing it to vendors

47
00:01:31.716 --> 00:01:33.802
and hoping that someone
else will do it for them.

48
00:01:33.802 --> 00:01:36.763
- A hard question because this
space is moving so quickly,

49
00:01:36.763 --> 00:01:37.722
so many innovations,

50
00:01:37.722 --> 00:01:38.807
but what are some of the ones

51
00:01:38.807 --> 00:01:40.934
that you are most excited
about at the moment?

52
00:01:40.934 --> 00:01:42.894
- Yeah, for me, it's things
that are in the pipeline.

53
00:01:42.894 --> 00:01:45.188
When I look at the roadmaps
of a lot of these products,

54
00:01:45.188 --> 00:01:49.275
I think we are seeing 1/10th
of what is in the pipeline

55
00:01:49.275 --> 00:01:52.028
where a lot of the

56
00:01:52.028 --> 00:01:53.655
new patterns when you put all this data

57
00:01:53.655 --> 00:01:55.031
into all this compute power,

58
00:01:55.031 --> 00:01:57.325
a lot of new interesting
patterns emerge from it

59
00:01:57.325 --> 00:01:58.660
that are totally new

60
00:01:58.660 --> 00:02:01.538
and for you to be able to pick it up,

61
00:02:01.538 --> 00:02:03.289
or, for me, we would need to have

62
00:02:03.289 --> 00:02:04.874
essentially encyclopedic knowledge

63
00:02:04.874 --> 00:02:06.459
to be able to pick out those patterns

64
00:02:06.459 --> 00:02:07.627
and to even notice them, right?

65
00:02:07.627 --> 00:02:10.420
So now we have these large-scale models

66
00:02:10.420 --> 00:02:11.798
that can pick up on those patterns

67
00:02:11.798 --> 00:02:14.217
and just put new signals
into our decisions.

68
00:02:14.217 --> 00:02:16.344
So I get most excited about things

69
00:02:16.344 --> 00:02:18.221
that are targeting the super users,

70
00:02:18.221 --> 00:02:20.598
people that are deep in their field.

71
00:02:20.598 --> 00:02:21.808
I don't know if you've seen interview

72
00:02:21.808 --> 00:02:24.769
with Alan Alda the other day from "MASH."

73
00:02:24.769 --> 00:02:28.481
He just used AI to help
him co-write a new script

74
00:02:28.481 --> 00:02:29.858
for one of the "MASH" episodes

75
00:02:29.858 --> 00:02:31.526
and was just super playful with it.

76
00:02:31.526 --> 00:02:33.945
And I think that's where when we get gurus

77
00:02:33.945 --> 00:02:35.405
and domain experts to get access

78
00:02:35.405 --> 00:02:37.031
to these tools and compute power,

79
00:02:37.031 --> 00:02:39.367
I think some really magical
things are going to come out.

80
00:02:39.367 --> 00:02:42.745
- How big is the gap between
what the public understands

81
00:02:42.745 --> 00:02:45.999
and where we are actually at with AI,

82
00:02:45.999 --> 00:02:48.668
because there's a lot
of fear, I would say?

83
00:02:48.668 --> 00:02:49.502
- There is fear,

84
00:02:49.502 --> 00:02:51.629
but also when you talk
to the AI companies,

85
00:02:51.629 --> 00:02:53.006
a lot of the leading ones,

86
00:02:53.006 --> 00:02:55.675
I have the privilege of
seeing their roadmaps,

87
00:02:55.675 --> 00:02:57.594
the stuff they have, they
don't even want to share,

88
00:02:57.594 --> 00:02:59.053
because they don't know how to control it,

89
00:02:59.053 --> 00:02:59.888
how to manage it,

90
00:02:59.888 --> 00:03:01.514
how to put in the right guardrails.

91
00:03:01.514 --> 00:03:03.516
So they themselves are a bit skeptical

92
00:03:03.516 --> 00:03:05.560
of putting it out into the public domain,

93
00:03:05.560 --> 00:03:07.437
out of worry that they cannot provide

94
00:03:07.437 --> 00:03:08.897
the right safety measures around it.

95
00:03:08.897 --> 00:03:10.523
So I will leave it at that,

96
00:03:10.523 --> 00:03:12.275
but there is much more
that actually exists,

97
00:03:12.275 --> 00:03:14.402
and it's at our fingertips,

98
00:03:14.402 --> 00:03:15.862
or could be at our fingertips,

99
00:03:15.862 --> 00:03:18.489
than what we have experienced
in our daily lives.

100
00:03:18.489 --> 00:03:19.657
- Vlad, thank you so much.

101
00:03:19.657 --> 00:03:20.491
- Thank you.