1
00:00:00,875 --> 00:00:02,809
Every single one of us will lose

2
00:00:02,833 --> 00:00:06,684
or has already lost something
we rely on every single day.

3
00:00:06,708 --> 00:00:09,518
I am of course talking about our keys.

4
00:00:09,542 --> 00:00:11,184
(Laughter)

5
00:00:11,208 --> 00:00:12,476
Just kidding.

6
00:00:12,500 --> 00:00:16,309
What I actually want to talk about is one
of our most important senses: vision.

7
00:00:16,333 --> 00:00:19,143
Every single day we each lose
a little bit of our ability

8
00:00:19,167 --> 00:00:20,476
to refocus our eyes

9
00:00:20,500 --> 00:00:22,809
until we can't refocus at all.

10
00:00:22,833 --> 00:00:24,809
We call this condition presbyopia,

11
00:00:24,833 --> 00:00:27,351
and it affects two billion
people worldwide.

12
00:00:27,375 --> 00:00:29,518
That's right, I said billion.

13
00:00:29,542 --> 00:00:31,268
If you haven't heard of presbyopia,

14
00:00:31,292 --> 00:00:34,101
and you're wondering,
"Where are these two billion people?"

15
00:00:34,125 --> 00:00:36,226
here's a hint before
I get into the details.

16
00:00:36,250 --> 00:00:39,518
It's the reason why people wear
reading glasses or bifocal lenses.

17
00:00:39,542 --> 00:00:42,400
I'll get started by describing
the loss in refocusing ability

18
00:00:42,424 --> 00:00:43,934
leading up to presbyopia.

19
00:00:43,958 --> 00:00:46,184
As a newborn, you would have
been able to focus

20
00:00:46,208 --> 00:00:48,351
as close as six and a half centimeters,

21
00:00:48,375 --> 00:00:49,643
if you wished to.

22
00:00:49,667 --> 00:00:52,809
By your mid-20s, you have about
half of that focusing power left.

23
00:00:52,833 --> 00:00:54,101
10 centimeters or so,

24
00:00:54,125 --> 00:00:56,684
but close enough that you never
notice the difference.

25
00:00:56,708 --> 00:00:57,976
By your late 40s though,

26
00:00:58,000 --> 00:01:00,351
the closest you can focus
is about 25 centimeters,

27
00:01:00,375 --> 00:01:01,643
maybe even farther.

28
00:01:01,667 --> 00:01:03,768
Losses in focusing ability
beyond this point

29
00:01:03,792 --> 00:01:06,018
start affecting near-vision
tasks like reading,

30
00:01:06,042 --> 00:01:07,643
and by the time you reach age 60,

31
00:01:07,667 --> 00:01:10,018
nothing within a meter
radius of you is clear.

32
00:01:10,042 --> 00:01:12,309
Right now some of you
are probably thinking,

33
00:01:12,333 --> 00:01:15,643
that sounds bad but he means
you in a figurative sense,

34
00:01:15,667 --> 00:01:19,101
only for the people that actually
end up with presbyopia.

35
00:01:19,125 --> 00:01:23,559
But no, when I say you, I literally mean
that every single one of you

36
00:01:23,583 --> 00:01:26,809
will someday be presbyopic
if you aren't already.

37
00:01:26,833 --> 00:01:28,226
That sounds a bit troubling.

38
00:01:28,250 --> 00:01:31,934
I want to remind you that presbyopia
has been with us for all of human history

39
00:01:31,958 --> 00:01:34,768
and we've done a lot
of different things to try and fix it.

40
00:01:34,792 --> 00:01:38,684
So to start, let's imagine
that you're sitting at a desk, reading.

41
00:01:38,708 --> 00:01:40,059
If you were presbyopic,

42
00:01:40,083 --> 00:01:42,143
it might look a little
something like this.

43
00:01:42,167 --> 00:01:45,059
Anything close by,
like the magazine, will be blurry.

44
00:01:45,083 --> 00:01:46,434
Moving on to solutions.

45
00:01:46,458 --> 00:01:48,101
First, reading glasses.

46
00:01:48,125 --> 00:01:50,184
These have lenses
with a single focal power

47
00:01:50,208 --> 00:01:52,601
tuned so that near objects
come into focus.

48
00:01:52,625 --> 00:01:55,226
But far objects
necessarily go out of focus,

49
00:01:55,250 --> 00:01:57,726
meaning you have to constantly
switch back and forth

50
00:01:57,750 --> 00:01:59,518
between wearing and not wearing them.

51
00:01:59,542 --> 00:02:00,809
To solve this problem

52
00:02:00,833 --> 00:02:04,101
Benjamin Franklin invented
what he called "double spectacles."

53
00:02:04,125 --> 00:02:06,226
Today we call those bifocals,

54
00:02:06,250 --> 00:02:09,684
and what they let him do
was see far when he looked up

55
00:02:09,708 --> 00:02:11,434
and see near when he looked down.

56
00:02:11,458 --> 00:02:14,434
Today we also have progressive lenses
which get rid of the line

57
00:02:14,458 --> 00:02:17,059
by smoothly varying the focal power
from top to bottom.

58
00:02:17,083 --> 00:02:18,518
The downside to both of these

59
00:02:18,542 --> 00:02:21,143
is that you lose field of vision
at any given distance,

60
00:02:21,167 --> 00:02:23,768
because it gets split up
from top to bottom like this.

61
00:02:23,792 --> 00:02:25,184
To see why that's a problem,

62
00:02:25,208 --> 00:02:28,059
imagine that you're climbing
down a ladder or stairs.

63
00:02:28,083 --> 00:02:31,684
You look down to get
your footing but it's blurry.

64
00:02:31,708 --> 00:02:33,101
Why would it be blurry?

65
00:02:33,125 --> 00:02:36,559
Well, you look down
and that's the near part of the lens,

66
00:02:36,583 --> 00:02:39,143
but the next step was past arm's reach,

67
00:02:39,167 --> 00:02:41,268
which for your eyes counts as far.

68
00:02:41,292 --> 00:02:44,184
The next solution I want to point out
is a little less common

69
00:02:44,208 --> 00:02:46,601
but comes up in contact lenses
or LASIK surgeries,

70
00:02:46,625 --> 00:02:47,976
and it's called monovision.

71
00:02:48,000 --> 00:02:50,518
It works by setting up
the dominant eye to focus far

72
00:02:50,542 --> 00:02:52,143
and the other eye to focus near.

73
00:02:52,167 --> 00:02:54,976
Your brain does the work
of intelligently putting together

74
00:02:55,000 --> 00:02:56,976
the sharpest parts from each eye's view,

75
00:02:57,000 --> 00:02:59,268
but the two eyes see
slightly different things,

76
00:02:59,292 --> 00:03:01,934
and that makes it harder
to judge distances binocularly.

77
00:03:01,958 --> 00:03:03,351
So where does that leave us?

78
00:03:03,375 --> 00:03:05,226
We've come up with a lot of solutions

79
00:03:05,250 --> 00:03:07,851
but none of them quite restore
natural refocusing.

80
00:03:07,875 --> 00:03:09,934
None of them let you
just look at something

81
00:03:09,958 --> 00:03:11,434
and expect it to be in focus.

82
00:03:11,458 --> 00:03:12,809
But why?

83
00:03:12,833 --> 00:03:14,101
Well, to explain that

84
00:03:14,125 --> 00:03:16,851
we'll want to take a look
at the anatomy of the human eye.

85
00:03:16,875 --> 00:03:20,101
The part of the eye that allows us
to refocus to different distances

86
00:03:20,125 --> 00:03:21,684
is called the crystalline lens.

87
00:03:21,708 --> 00:03:25,559
There are muscles surrounding the lens
that can deform it into different shapes,

88
00:03:25,583 --> 00:03:27,684
which in turn changes its focusing power.

89
00:03:27,708 --> 00:03:30,184
What happens when someone
becomes presbyopic?

90
00:03:30,208 --> 00:03:32,434
It turns out that
the crystalline lens stiffens

91
00:03:32,458 --> 00:03:35,143
to the point that it doesn't
really change shape anymore.

92
00:03:35,167 --> 00:03:38,893
Now, thinking back
on all the solutions I listed earlier,

93
00:03:38,917 --> 00:03:42,643
we can see that they all have
something in common with the others

94
00:03:42,667 --> 00:03:44,143
but not with our eyes,

95
00:03:44,167 --> 00:03:46,184
and that is that they're all static.

96
00:03:46,208 --> 00:03:49,101
It's like the optical equivalent
of a pirate with a peg leg.

97
00:03:49,125 --> 00:03:52,268
What is the optical equivalent
of a modern prosthetic leg?

98
00:03:52,292 --> 00:03:55,559
The last several decades have seen
the creation and rapid development

99
00:03:55,583 --> 00:03:58,268
of what are called "focus-tunable lenses."

100
00:03:58,292 --> 00:03:59,976
There are several different types.

101
00:04:00,000 --> 00:04:01,809
Mechanically-shifted Alvarez lenses,

102
00:04:01,833 --> 00:04:03,268
deformable liquid lenses

103
00:04:03,292 --> 00:04:05,851
and electronically-switched,
liquid crystal lenses.

104
00:04:05,875 --> 00:04:07,684
Now these have their own trade-offs,

105
00:04:07,708 --> 00:04:10,309
but what they don't skimp on
is the visual experience.

106
00:04:10,333 --> 00:04:13,601
Full-field-of-view vision that can be
sharp at any desired distance.

107
00:04:13,625 --> 00:04:15,768
OK, great. The lenses we need
already exist.

108
00:04:15,792 --> 00:04:17,684
Problem solved, right?

109
00:04:17,708 --> 00:04:19,143
Not so fast.

110
00:04:19,167 --> 00:04:22,143
Focus-tunable lenses add a bit
of complexity to the equation.

111
00:04:22,167 --> 00:04:26,018
The lenses don't have any way of knowing
what distance they should be focused to.

112
00:04:26,042 --> 00:04:27,351
What we need are glasses

113
00:04:27,375 --> 00:04:29,934
that, when you're looking far,
far objects are sharp,

114
00:04:29,958 --> 00:04:31,268
and when you look near,

115
00:04:31,292 --> 00:04:33,726
near objects come into focus
in your field of view,

116
00:04:33,750 --> 00:04:35,601
without you having to think about it.

117
00:04:35,625 --> 00:04:38,143
What I've worked on
these last few years at Stanford

118
00:04:38,167 --> 00:04:40,768
is building that exact intelligence
around the lenses.

119
00:04:40,792 --> 00:04:44,393
Our prototype borrows technology
from virtual and augmented reality systems

120
00:04:44,417 --> 00:04:45,934
to estimate focusing distance.

121
00:04:45,958 --> 00:04:49,601
We have an eye tracker that can tell
what direction our eyes are focused in.

122
00:04:49,625 --> 00:04:52,434
Using two of these, we can
triangulate your gaze direction

123
00:04:52,458 --> 00:04:53,768
to get a focus estimate.

124
00:04:53,792 --> 00:04:55,976
Just in case though,
to increase reliability,

125
00:04:56,000 --> 00:04:57,559
we also added a distance sensor.

126
00:04:57,583 --> 00:04:59,976
The sensor is a camera
that looks out at the world

127
00:05:00,000 --> 00:05:01,601
and reports distances to objects.

128
00:05:01,625 --> 00:05:04,601
We can again use your gaze direction
to get a distance estimate

129
00:05:04,625 --> 00:05:05,893
for a second time.

130
00:05:05,917 --> 00:05:07,893
We then fuse those two distance estimates

131
00:05:07,917 --> 00:05:10,393
and update the focus-tunable
lens power accordingly.

132
00:05:10,417 --> 00:05:13,351
The next step for us was
to test our device on actual people.

133
00:05:13,375 --> 00:05:16,518
So we recruited about 100 presbyopes
and had them test our device

134
00:05:16,542 --> 00:05:18,351
while we measured their performance.

135
00:05:18,375 --> 00:05:21,643
What we saw convinced us right then
that autofocals were the future.

136
00:05:21,667 --> 00:05:25,018
Our participants could see more clearly,
they could focus more quickly

137
00:05:25,042 --> 00:05:28,059
and they thought it was an easier
and better focusing experience

138
00:05:28,083 --> 00:05:29,559
than their current correction.

139
00:05:29,583 --> 00:05:31,643
To put it simply, when it comes to vision,

140
00:05:31,667 --> 00:05:34,893
autofocals don't compromise
like static corrections in use today do.

141
00:05:34,917 --> 00:05:36,851
But I don't want to get ahead of myself.

142
00:05:36,875 --> 00:05:39,601
There's a lot of work
for my colleagues and me left to do.

143
00:05:39,625 --> 00:05:41,893
For example, our glasses are a bit --

144
00:05:41,917 --> 00:05:42,934
(Laughter)

145
00:05:42,958 --> 00:05:44,309
bulky, maybe?

146
00:05:44,333 --> 00:05:47,684
And one reason for this
is that we used bulkier components

147
00:05:47,708 --> 00:05:50,518
that are often intended
for research use or industrial use.

148
00:05:50,542 --> 00:05:52,809
Another is that we need
to strap everything down

149
00:05:52,833 --> 00:05:56,643
because current eye-tracking algorithms
don't have the robustness that we need.

150
00:05:56,667 --> 00:05:57,976
So moving forward,

151
00:05:58,000 --> 00:06:00,476
as we move from a research
setting into a start-up,

152
00:06:00,500 --> 00:06:02,434
we plan to make future autofocals

153
00:06:02,458 --> 00:06:05,018
eventually look a little bit more
like normal glasses.

154
00:06:05,042 --> 00:06:08,393
For this to happen,
we'll need to significantly improve

155
00:06:08,417 --> 00:06:10,559
the robustness
of our eye-tracking solution.

156
00:06:10,583 --> 00:06:14,768
We'll also need to incorporate smaller
and more efficient electronics and lenses.

157
00:06:14,792 --> 00:06:16,976
That said, even with
our current prototype,

158
00:06:17,000 --> 00:06:19,726
we've shown that today's
focus-tunable lens technology

159
00:06:19,750 --> 00:06:23,309
is capable of outperforming
traditional forms of static correction.

160
00:06:23,333 --> 00:06:24,976
So it's only a matter of time.

161
00:06:25,000 --> 00:06:27,059
It's pretty clear that in the near future,

162
00:06:27,083 --> 00:06:30,143
instead of worrying about which pair
of glasses to use and when,

163
00:06:30,167 --> 00:06:32,708
we'll be able to just focus
on the important things.

164
00:06:33,667 --> 00:06:34,934
Thank you.

165
00:06:34,958 --> 00:06:36,583
(Applause)