Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:03,319 --> 00:00:07,170
Jeff perhaps a similar question if you
2
00:00:06,000 --> 00:00:11,639
may what would you think the most
3
00:00:07,170 --> 00:00:15,088
exciting work to come sorry
4
00:00:11,639 --> 00:00:18,810
then let me start by just saying a few
5
00:00:15,089 --> 00:00:21,330
things that seem obvious I think if you
6
00:00:18,810 --> 00:00:23,430
work as a radiologist you'll like the
7
00:00:21,330 --> 00:00:25,709
coyote that's already over the edge of
8
00:00:23,430 --> 00:00:27,240
the cliff but hasn't yet looked down so
9
00:00:25,710 --> 00:00:30,390
it doesn't realize there's no ground
10
00:00:27,240 --> 00:00:32,399
underneath him people should stop
11
00:00:30,390 --> 00:00:34,950
training radiologists now it's just
12
00:00:32,399 --> 00:00:36,390
completely obvious within five years I'm
13
00:00:34,950 --> 00:00:37,680
deep learning is going to do better than
14
00:00:36,390 --> 00:00:39,719
radiologists because this can be able to
15
00:00:37,680 --> 00:00:41,340
hit a lot more experience and it might
16
00:00:39,719 --> 00:00:44,789
be ten years but we got plenty of
17
00:00:41,340 --> 00:00:46,469
radiologists already well I said this at
18
00:00:44,789 --> 00:00:52,199
the hospital and it didn't go down too
19
00:00:46,469 --> 00:00:53,910
well poppy there's going to be thousands
20
00:00:52,199 --> 00:00:56,550
of applications of the deep learning
21
00:00:53,910 --> 00:01:00,419
technology we currently have especially
22
00:00:56,550 --> 00:01:03,599
stop using and fast chips on to
23
00:01:00,420 --> 00:01:04,769
thousands of applications and hundreds
24
00:01:03,600 --> 00:01:07,320
of them are going to work and those are
25
00:01:04,769 --> 00:01:08,820
the ones you're going to hear about so
26
00:01:07,320 --> 00:01:10,889
there'll be lots of that and I think of
27
00:01:08,820 --> 00:01:13,380
that now is business as usual take any
28
00:01:10,890 --> 00:01:15,570
old problem where you have to predict
29
00:01:13,380 --> 00:01:17,220
something and you have a lot of data and
30
00:01:15,570 --> 00:01:18,330
deep learning is probably going to make
31
00:01:17,220 --> 00:01:19,980
it work better than the existing
32
00:01:18,330 --> 00:01:21,360
techniques
33
00:01:19,980 --> 00:01:26,010
you
34
00:01:21,360 --> 00:01:26,010
[Music]
2485
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.