1
00:00:00,500 --> 00:00:04,357
WE WANT TO MAKE SURE THAT ALL OF THESE
PEOPLE THAT ARE TOUCHING OUR PRODUCTS AND

2
00:00:04,440 --> 00:00:08,327
LIVING IN THE WORLD GET A SENSE
OF WHERE THE TECHNOLOGY IS AND

3
00:00:08,410 --> 00:00:10,557
WHAT THE KIND OF RISKS
WE HAVE TO MITIGATE ARE.

4
00:00:10,640 --> 00:00:13,467
>> DO YOU FEEL LIKE THERE'S
JUST MORE ATTENTION ON OPENAI?

5
00:00:13,550 --> 00:00:17,197
IT FEELS LIKE OTHERS HAVE NOT HAD
THE SAME LEVEL OF INCIDENTS, OR IT'S.

6
00:00:17,280 --> 00:00:20,507
>> KIND OF WHAT I WAS TRYING
TO GET AT THERE. WE JUST TOUCHED

7
00:00:20,590 --> 00:00:23,237
SO MANY PEOPLE AND WE DO
SO MUCH DIFFERENT STUFF.

8
00:00:23,320 --> 00:00:26,507
>> OKAY, SO JEN-HSUN HUANG, NVIDIA'S CEO,
PUT OUT THIS BASICALLY PIECE OF SOFTWARE

9
00:00:26,590 --> 00:00:29,677
THAT HE SAID WOULD HAVE PREVENTED HUGGING
FACE. AND HE SAID, IT'S AN ENGINEERING

10
00:00:29,760 --> 00:00:34,347
PROBLEM. OPENAI DIDN'T SIGN
ON TO THAT PROJECT. WHY NOT

11
00:00:34,430 --> 00:00:36,887
ANTHROPIC AND OTHERS DID.
DO YOU AGREE WITH JENSEN THAT IT

12
00:00:36,970 --> 00:00:39,857
COULD HAVE BEEN SOLVED?
>> I FROM WHAT I KNOW ABOUT IT

13
00:00:39,940 --> 00:00:42,927
AND I'M NOT SUPER INTO IT,
THE DETAILS, I THINK IT'S A GOOD

14
00:00:43,010 --> 00:00:45,897
THING. I THINK WE ARE DOING
SIMILAR THINGS. I DON'T THINK

15
00:00:45,980 --> 00:00:48,897
IT'S A FULL SOLUTION,
AND I WORRY THAT IF WE TREAT AI

16
00:00:48,980 --> 00:00:50,567
SAFETY AS ONLY AN ENGINEERING

17
00:00:50,650 --> 00:00:53,037
PROBLEM, WE WILL MISS THE VERY

18
00:00:53,120 --> 00:00:58,107
IMPORTANT POINT THAT WE HAVE.
WE HAVE A SCIENCE PROBLEM IN

19
00:00:58,190 --> 00:01:01,947
FRONT OF US. WE STILL HAVE DISCOVERY ABOUT
HOW TO ALIGN THESE MODELS, AND WE HAVE TO

20
00:01:02,030 --> 00:01:03,747
SOLVE THAT SCIENTIFIC PROBLEM TOO.

21
00:01:03,830 --> 00:01:06,817
>> OKAY. AT THE SAME TIME,
AS WE MENTIONED, WE ARE HERE AT

22
00:01:06,900 --> 00:01:09,757
DEVELOPER DAY. YOU HAVE.
I MEAN, IT LOOKS LIKE HUNDREDS OF

23
00:01:09,840 --> 00:01:11,787
DEVELOPERS COMPANIES.
I WOULD THINK INVESTORS HERE TOO. HOW

24
00:01:11,870 --> 00:01:14,787
ARE YOU TOEING THE LINE BETWEEN
TRYING TO GET THESE PEOPLE TO

25
00:01:14,870 --> 00:01:17,827
USE YOUR TECHNOLOGY, SHOW THEM
HOW GREAT IT IS, BUT AT THE

26
00:01:17,910 --> 00:01:20,027
SAME TIME, TELL THEM THAT IT'S SAFE.

27
00:01:20,110 --> 00:01:24,767
>> I MEAN, I THINK IT'S A SIMILAR THRUST,
LIKE PEOPLE WANT TO BE ABLE TO USE THIS

28
00:01:24,850 --> 00:01:29,107
TECHNOLOGY AND KNOW THAT IT'S GOING TO BE
SAFE. SO I THINK PART OF IT BEING GREAT

29
00:01:29,190 --> 00:01:31,637
TECHNOLOGY IS TO HAVE IT BE
THE SAFEST, MOST ALIGNED, MOST

30
00:01:31,720 --> 00:01:33,247
DEPENDABLE AI IN THE INDUSTRY.

31
00:01:33,330 --> 00:01:36,017
>> ONE QUESTION ON LIABILITY.
SO IF SOMETHING DOES GO WRONG,

32
00:01:36,100 --> 00:01:39,917
SO SAY ONE OF THESE STARTUPS HERE BUILDS A
GREAT APP USING OPENAI AND SOMETHING GOES

33
00:01:40,000 --> 00:01:43,687
TERRIBLY WRONG. YEAH. WHOSE FAULT IS THAT?

34
00:01:43,770 --> 00:01:46,687
>> THE LIABILITY FRAMEWORK
IS BEING DEBATED A LOT IN THE

35
00:01:46,770 --> 00:01:50,197
WORLD RIGHT NOW. I SUSPECT LIKE
OTHER INDUSTRIES, THERE WILL BE

36
00:01:50,280 --> 00:01:52,527
A SORT OF MULTI-LEVEL THING.
LIKE IF YOU THINK ABOUT THE

37
00:01:52,610 --> 00:01:55,467
AUTO INDUSTRY, FOR EXAMPLE,
IF THERE'S FAULTY PARTS, YOU KNOW,

38
00:01:55,550 --> 00:01:58,567
YOU CAN BE LIABLE IN SOME WAYS
THERE. IF THE CAR MANUFACTURER

39
00:01:58,650 --> 00:02:01,437
PUTS THEM TOGETHER IN A WAY
THAT'S OUTSIDE OF THE SPEC OF

40
00:02:01,520 --> 00:02:04,307
WHAT THEY'RE SUPPOSED TO DO,
AND THE CAR MANUFACTURER COULD

41
00:02:04,390 --> 00:02:06,577
BE. AND IF SOMEONE'S DRIVING
DRUNK, IT'S THEIR FAULT. SO,

42
00:02:06,660 --> 00:02:09,477
YOU KNOW, DEPENDING ON WHETHER
THE FAULT IS IN THE MODEL AND

43
00:02:09,560 --> 00:02:11,617
HOW IT'S USED OR SOMEONE
MISUSING IT INTENTIONALLY, I

44
00:02:11,700 --> 00:02:14,647
ASSUME WE'LL HAVE LIKE A,
YOU KNOW, A NEW LIABILITY FRAMEWORK.

45
00:02:14,730 --> 00:02:17,687
>> WHAT IS YOUR PITCH TODAY
TO DEVELOPERS? COST SEEMS LIKE

46
00:02:17,770 --> 00:02:21,287
IT'S BEEN A BIG THEME FOR YOU
GUYS, BUT WHAT ARE YOU GOING TO

47
00:02:21,370 --> 00:02:24,327
TELL PEOPLE ON STAGE IN TERMS
OF WHY OPENAI? WHY CHATGPT IN

48
00:02:24,410 --> 00:02:26,497
YOUR MIND IS THE BEST
OPTION FOR THESE GUYS?

49
00:02:26,580 --> 00:02:29,467
>> WE HAVE MORE THAN 20 ANNOUNCEMENTS
TODAY, SO WE'LL BE TELLING.

50
00:02:29,550 --> 00:02:31,437
>> THEM, GIVE US A FLAVOR. >> A LOT.

51
00:02:31,520 --> 00:02:34,437
>> OF WHAT ARE WE WHAT
ARE WE GOING TO HEAR?

52
00:02:34,520 --> 00:02:38,277
>> I MEAN, NEW WAYS TO USE AI,
NEW SORT OF LIKE NEW ABILITY TO

53
00:02:38,360 --> 00:02:41,307
HAVE AI SYSTEMS WORKING FOR YOU
AND HELPING YOU ALL THE TIME.

54
00:02:41,390 --> 00:02:44,417
NEW MODELS. WE'LL TALK ABOUT
THE COST ISSUE. WE REALLY AND

55
00:02:44,500 --> 00:02:47,287
YOU KNOW, A LOT OF NEW TOOLS
FOR DEVELOPERS. IT'S REALLY A

56
00:02:47,370 --> 00:02:49,217
LOT OF, I THINK, EXCITING STUFF.

57
00:02:49,300 --> 00:02:52,327
IT IS COOL TO SEE ALL OF THIS
COME TOGETHER AND THE DEGREE TO

58
00:02:52,410 --> 00:02:56,257
WHICH I THINK WE JUST HAVE THE BEST
OFFERING FOR DEVELOPERS ACROSS THE BOARD.

59
00:02:56,340 --> 00:02:59,227
>> COST IS SUCH A CENTRAL TOPIC HERE.
AND WE'VE ALSO TALKED

60
00:02:59,310 --> 00:03:01,327
BEFORE ABOUT THE ISSUE
OF DISTILLATION. I MEAN, SOME OF

61
00:03:01,410 --> 00:03:04,397
THE OPEN SOURCE MODELS THAT ARE
CHEAPER. YOU AND ANTHROPIC HAVE

62
00:03:04,480 --> 00:03:06,907
ACCUSED OF BASICALLY COPYING
YOUR TECHNOLOGY. HOW ARE YOU

63
00:03:06,990 --> 00:03:08,507
FEELING ABOUT DISTILLATION
AND THOSE MODELS?

64
00:03:08,590 --> 00:03:12,177
>> I THINK WE HAVE THE BEST
MODELS TO OFFER AT EVERY PRICE

65
00:03:12,260 --> 00:03:15,207
POINT. YEAH. AND I DON'T THINK
LIKE, YOU KNOW, PEOPLE DO TALK

66
00:03:15,290 --> 00:03:18,217
ABOUT OPEN SOURCE, BUT I DON'T
THINK THERE'S ANY OPEN SOURCE

67
00:03:18,300 --> 00:03:21,217
MODEL OUT THERE THAT YOU CAN
RUN THAT BEATS US ON A COST

68
00:03:21,300 --> 00:03:23,957
PRICE COMPARISON. AND SPEAKING
OF COST, GENERALLY COME WATCH.

69
00:03:24,040 --> 00:03:27,287
WE GOT WE GOT SOME. >> COOL STUFF.
OKAY. WELL, AGENTS AS WELL. SO MARK

70
00:03:27,370 --> 00:03:30,327
ZUCKERBERG AND MUSE HAVE MADE
A LOT OF PROGRESS. IT SEEMS LIKE

71
00:03:30,410 --> 00:03:34,227
IN AGENTS AND HARDWARE. WHAT DO YOU
MAKE OF MUSE? HAVE YOU TRIED IT AT ALL?

72
00:03:34,310 --> 00:03:37,097
>> I'VE TRIED IT.
IT SEEMS LIKE A NICE PRODUCT.

73
00:03:37,180 --> 00:03:39,107
>> ARE YOU THREATENED
BY ZUCKERBERG AT ALL?

74
00:03:39,190 --> 00:03:43,177
>> NO. >> NOT AT ALL. >> WELL, I MEAN,
HE'S LIKE A VERY IMPRESSIVE PERSON. AND

75
00:03:43,260 --> 00:03:47,047
THAT COMPANY HAS, LIKE GREAT RESOURCES.
BUT I FEEL VERY GOOD ABOUT OUR ROADMAP.

76
00:03:47,130 --> 00:03:49,317
>> HE'S ALSO SHOWED WHAT'S POSSIBLE
WITH HARDWARE. WHERE'S JOHNNY?

77
00:03:49,400 --> 00:03:52,287
>> I'VE I DON'T WANT TO SPOIL
YOUR SURPRISE, BUT I THINK WE

78
00:03:52,370 --> 00:03:54,317
WILL HAVE SOMETHING THAT'S
WORTH WAITING FOR TODAY.

79
00:03:54,400 --> 00:03:57,257
>> WE MAY HAVE SOME HARDWARE.
IS THAT WHAT YOU'RE SAYING?

80
00:03:57,340 --> 00:04:00,257
OKAY, WELL, WHAT IS THE PLAN
IN HARDWARE AND DOES IT PUT MORE

81
00:04:00,340 --> 00:04:03,097
PRESSURE TO HAVE AN EVEN BETTER
PRODUCT? IF MARK ZUCKERBERG

82
00:04:03,180 --> 00:04:06,097
SEEMS TO HAVE KIND OF GOTTEN
OUT OF THE GATES FIRST, EVEN

83
00:04:06,180 --> 00:04:08,967
THOUGH MAYBE SOME OF THEIR
STUFF IS STILL IN THE WORKS.

84
00:04:09,050 --> 00:04:11,397
>> I AM NOT WORRIED ABOUT BEING
FIRST WITH NEW HARDWARE. NEW

85
00:04:11,480 --> 00:04:14,337
HARDWARE IS VERY HARD TO DO,
LIKE NEW COMPUTING FORM FACTORS

86
00:04:14,420 --> 00:04:17,307
COME ALONG VERY RARELY,
AND THE QUALITY BAR THAT WE WANT TO HIT

87
00:04:17,390 --> 00:04:20,307
FOR SOMETHING THAT LIKE YOU
WILL LOVE AND REALLY BE A KEY

88
00:04:20,390 --> 00:04:24,147
PART OF THE WAY YOU INTERFACE
WITH AI, THAT'S GOING TO TAKE A

89
00:04:24,230 --> 00:04:27,287
WHILE, AND WE'RE GOING TO DO
THAT AT A HIGH LEVEL OF QUALITY.

90
00:04:27,370 --> 00:04:30,187
>> I ALSO WANT TO ASK ABOUT
THE PERCEPTION. WE WERE JUST

91
00:04:30,270 --> 00:04:32,387
TALKING ABOUT THE SNL WEEKEND
UPDATE. THERE IS OBVIOUSLY A

92
00:04:32,470 --> 00:04:35,357
MOVIE ABOUT YOU. I DON'T KNOW
IF YOU'VE SEEN THE TRAILER.

93
00:04:35,440 --> 00:04:38,197
THERE'S SOME ISSUES IN TERMS
OF PUBLIC PERCEPTION AND JUST THIS,

94
00:04:38,280 --> 00:04:40,527
THIS THOUGHT THAT MAYBE YOU
MAYBE DARIO AMODEI AS LEADERS

95
00:04:40,610 --> 00:04:44,137
DON'T HAVE THE BEST REPUTATION RIGHT NOW.

96
00:04:44,220 --> 00:04:48,337
>> I MEAN, I THINK EVERYONE
DOING SOMETHING THIS IMPORTANT

97
00:04:48,420 --> 00:04:51,707
TO SOCIETY DESERVES GREAT SCRUTINY.
I DID SEE THE MOVIE

98
00:04:51,790 --> 00:04:53,977
TRAILER. I MEAN, IT LOOKS
LIKE AN AWFUL MOVIE, BUT I

99
00:04:54,060 --> 00:04:56,717
APPRECIATE PEOPLE'S CREATIVITY
AND I REALLY DO SUPPORT THAT.

100
00:04:56,800 --> 00:04:59,770
SO I HOPE, I HOPE THEY
HAVE A FUN TIME WITH IT. THE