1
00:00:03,160 --> 00:00:07,960
Speaker 1: This is Bloombird Law with June Brusso from Bloomberg Radio.

2
00:00:09,680 --> 00:00:13,160
Speaker 1: Starship is capable of doing that. It's capable of of

3
00:00:13,320 --> 00:00:16,040
Speaker 1: getting getting a million tons of surps of Mars and

4
00:00:16,160 --> 00:00:20,200
Speaker 1: creating a self sustaining city. Remember when there was almost

5
00:00:20,239 --> 00:00:24,280
Speaker 1: a myth around Elon Musk, the richest person on the planet,

6
00:00:24,440 --> 00:00:29,960
Speaker 1: a visionary with plans for interplanetary migration. But now prospective

7
00:00:30,080 --> 00:00:34,120
Speaker 1: jurors that his securities fraud trial have called him a delusional,

8
00:00:34,240 --> 00:00:39,560
Speaker 1: narcissist and an idiot. Following a bizarre Musk has been

9
00:00:39,600 --> 00:00:43,519
Speaker 1: dethroned after losing more than two hundred billion dollars, and

10
00:00:43,600 --> 00:00:47,200
Speaker 1: he's facing losing billions more at a trial where Testlas

11
00:00:47,240 --> 00:00:51,639
Speaker 1: shareholders are accusing him of lying in those infamous tweets

12
00:00:51,720 --> 00:00:54,080
Speaker 1: four and a half years ago about a plan to

13
00:00:54,120 --> 00:00:58,800
Speaker 1: take the electric carmaker private with funding secured. Joining me

14
00:00:58,880 --> 00:01:02,480
Speaker 1: is Eric Tallely, a professor at Columbia Law School. What

15
00:01:02,560 --> 00:01:06,560
Speaker 1: do the investors have to prove in this trial? They

16
00:01:06,600 --> 00:01:09,839
Speaker 1: have already proven something from this trial. So the judge

17
00:01:09,880 --> 00:01:14,160
Speaker 1: in this case has already found that the tweets from

18
00:01:15,040 --> 00:01:20,000
Speaker 1: eighteen were false in this leading and either Mr mus

19
00:01:20,040 --> 00:01:22,240
Speaker 1: canoe or he should have known that they were false

20
00:01:22,280 --> 00:01:24,520
Speaker 1: service leading, And that's often a big chunk of a

21
00:01:24,640 --> 00:01:27,959
Speaker 1: securities FROD lawsuits, but it's not the only thing that

22
00:01:28,040 --> 00:01:29,800
Speaker 1: you have to show. And sort of the two or

23
00:01:29,840 --> 00:01:35,600
Speaker 1: three biggest issues remainings are, First, was this tweet material?

24
00:01:36,000 --> 00:01:39,240
Speaker 1: Was the extent of the falsity here material to an investor?

25
00:01:39,640 --> 00:01:43,080
Speaker 1: Would it have dantially changed how they would have acted

26
00:01:43,200 --> 00:01:46,880
Speaker 1: had different more accurate words have than you A second

27
00:01:46,880 --> 00:01:50,120
Speaker 1: issue is going to be whether Mr Musk intended to

28
00:01:50,320 --> 00:01:54,000
Speaker 1: deceive the market. And a third and somewhat related issue

29
00:01:54,120 --> 00:01:56,680
Speaker 1: is whether the false tweet, if it did you know,

30
00:01:56,760 --> 00:01:59,240
Speaker 1: to see people, didn't cause them to suffer a loss.

31
00:01:59,240 --> 00:02:01,600
Speaker 1: And so these are all things that the various class

32
00:02:01,600 --> 00:02:04,000
Speaker 1: members and the class of the whole are going to

33
00:02:04,160 --> 00:02:07,520
Speaker 1: have to prove. Mr Musk's team has, you know, really

34
00:02:07,600 --> 00:02:09,799
Speaker 1: kind of come out of the box pretty strong on

35
00:02:09,840 --> 00:02:14,040
Speaker 1: the materiality point, and their claim is really that hey listened. Uh.

36
00:02:14,600 --> 00:02:20,200
Speaker 1: This particular defendant's personality is such that two things are true. One,

37
00:02:20,480 --> 00:02:22,320
Speaker 1: he kind of shoots from the hip and he does

38
00:02:22,360 --> 00:02:24,600
Speaker 1: it in a bunch of different ways than everyone knows it.

39
00:02:24,840 --> 00:02:28,080
Speaker 1: And too, he has lots of people out there that

40
00:02:28,120 --> 00:02:31,000
Speaker 1: are just so infatuated with him that they will try

41
00:02:31,040 --> 00:02:34,720
Speaker 1: to take cues from, you know, slight facial motions and

42
00:02:34,760 --> 00:02:37,440
Speaker 1: what color shirt he's wearing, and they respond to kind

43
00:02:37,440 --> 00:02:40,120
Speaker 1: of everything. And so the defense in this case, I

44
00:02:40,120 --> 00:02:43,120
Speaker 1: think it's probably going to make a confession that, yes,

45
00:02:43,240 --> 00:02:46,160
Speaker 1: he misspoke in this case that you know, even though

46
00:02:46,200 --> 00:02:48,799
Speaker 1: he had tweeted this idea that he was going to

47
00:02:48,919 --> 00:02:53,160
Speaker 1: take Testla private and he had funding security, didn't really

48
00:02:53,200 --> 00:02:57,080
Speaker 1: have funding secure, but there were conversations going on and

49
00:02:57,720 --> 00:03:01,280
Speaker 1: they were sort of getting towards maybe having secure funding.

50
00:03:01,360 --> 00:03:03,359
Speaker 1: And I think the defense is going to argue, look,

51
00:03:03,600 --> 00:03:05,600
Speaker 1: if he had tweeted that, if you could fit that

52
00:03:05,680 --> 00:03:08,519
Speaker 1: into a tweet, which is another thing, the market still

53
00:03:08,560 --> 00:03:10,639
Speaker 1: would have moved exactly the same way it did. So

54
00:03:10,800 --> 00:03:13,920
Speaker 1: the argument I think that they're going to make is that, yes,

55
00:03:14,080 --> 00:03:17,959
Speaker 1: this was factually a stretch. He didn't have funding secured.

56
00:03:18,360 --> 00:03:20,519
Speaker 1: But even if he had sort of said the same thing,

57
00:03:20,919 --> 00:03:23,200
Speaker 1: so many people sort of believe that you know, Mr

58
00:03:23,280 --> 00:03:26,160
Speaker 1: Musk can move mountains and vent gravity, that the stock

59
00:03:26,200 --> 00:03:28,320
Speaker 1: price would have moved in exactly the same way that

60
00:03:28,400 --> 00:03:31,640
Speaker 1: it did and therefore this wasn't material and therefore it

61
00:03:31,680 --> 00:03:34,880
Speaker 1: didn't cause people to suffer a lot. So my sense,

62
00:03:35,040 --> 00:03:37,480
Speaker 1: just from the way that the current case is laying

63
00:03:37,480 --> 00:03:39,800
Speaker 1: out and from what I picked up from what was

64
00:03:39,840 --> 00:03:41,720
Speaker 1: going on in court yesterday, that's going to be a

65
00:03:41,720 --> 00:03:45,240
Speaker 1: big part of the defense the plaintiffs in this case.

66
00:03:45,240 --> 00:03:48,080
Speaker 1: You're gonna say, no, now, it really matters. His words

67
00:03:48,120 --> 00:03:51,840
Speaker 1: secured send out all kinds of alarm bells for everyone,

68
00:03:51,920 --> 00:03:55,200
Speaker 1: And that was much more definitive than I'm exploring options

69
00:03:55,280 --> 00:03:58,520
Speaker 1: even when uttered or tweeted by Mr Musk. And so

70
00:03:58,520 --> 00:04:00,080
Speaker 1: there's going to be kind of a big issue that

71
00:04:00,200 --> 00:04:03,040
Speaker 1: this jury is going to have to decide about whether

72
00:04:03,080 --> 00:04:06,640
Speaker 1: a reasonable investor it provided with an accurate tweet as

73
00:04:06,640 --> 00:04:09,560
Speaker 1: to what was going on, would have acted any different,

74
00:04:09,640 --> 00:04:12,480
Speaker 1: and whether the market would have acted any different. The

75
00:04:12,600 --> 00:04:16,880
Speaker 1: jurors have seen a different musque in recent months, and

76
00:04:16,960 --> 00:04:20,719
Speaker 1: so is that visionary thing going to still work with

77
00:04:20,800 --> 00:04:24,839
Speaker 1: a jury? Oh, he's special. Yeah, it's an interesting question

78
00:04:24,839 --> 00:04:26,720
Speaker 1: about whether it's going to work. I think you could

79
00:04:26,720 --> 00:04:29,800
Speaker 1: make an argument on either side. On the one hand,

80
00:04:30,440 --> 00:04:33,360
Speaker 1: you know, the last few months, let's just be realistic,

81
00:04:33,400 --> 00:04:36,760
Speaker 1: have not been terribly time to Mr Moskin. I think

82
00:04:37,000 --> 00:04:39,200
Speaker 1: the bloom is a little bit off the rose to

83
00:04:39,600 --> 00:04:43,800
Speaker 1: the extent to which he's this visionary, flawless manager who

84
00:04:43,839 --> 00:04:47,080
Speaker 1: sees into the future and makes the future happen. Today,

85
00:04:47,320 --> 00:04:49,640
Speaker 1: he still has some of those characteristics, but you know,

86
00:04:49,680 --> 00:04:52,479
Speaker 1: I think the much more public and erratic nature of

87
00:04:52,560 --> 00:04:55,440
Speaker 1: his behavior of the last few months has definitely sort

88
00:04:55,440 --> 00:04:58,560
Speaker 1: of taken the veneer off of that impression. But the

89
00:04:58,640 --> 00:05:01,320
Speaker 1: other thing that it has done that's probably worth noting

90
00:05:01,839 --> 00:05:04,560
Speaker 1: is that it is also exposed that, yeah, he does

91
00:05:04,600 --> 00:05:06,919
Speaker 1: shoot from the hip a lot. That he does, and

92
00:05:07,080 --> 00:05:09,800
Speaker 1: people sort of get this that he sort of is

93
00:05:09,839 --> 00:05:13,680
Speaker 1: almost thinking out loud when he is tweeting and speaking

94
00:05:13,839 --> 00:05:18,000
Speaker 1: and writing, and that's how he's always been and people

95
00:05:18,040 --> 00:05:20,640
Speaker 1: have always known that about him as well. So if

96
00:05:20,640 --> 00:05:23,640
Speaker 1: the market had a view of him as being a

97
00:05:23,720 --> 00:05:27,240
Speaker 1: profit for the next ten years, but they also knew

98
00:05:27,240 --> 00:05:29,159
Speaker 1: that he's a guy that engages in sort of a

99
00:05:29,200 --> 00:05:33,600
Speaker 1: modern James Joyce and stream of consciousness on his tweets,

100
00:05:33,720 --> 00:05:36,800
Speaker 1: then you know, the defense could still say, yeah, that's

101
00:05:36,880 --> 00:05:39,719
Speaker 1: the whole idea. People you know, at the time had

102
00:05:39,800 --> 00:05:42,920
Speaker 1: this view and maybe some of them still do, that

103
00:05:43,040 --> 00:05:46,960
Speaker 1: he is a seer and therefore even the casual throw

104
00:05:46,960 --> 00:05:49,760
Speaker 1: off stream of consciousness lines are going to be enough

105
00:05:49,800 --> 00:05:51,880
Speaker 1: to move those people in the market, and those are

106
00:05:51,960 --> 00:05:54,320
Speaker 1: enough people to move the market itself. So I think

107
00:05:54,360 --> 00:05:58,359
Speaker 1: there's a sense in which his now exposed sort of

108
00:05:58,440 --> 00:06:02,840
Speaker 1: short sighted laws and unforced errors as a manager can

109
00:06:03,040 --> 00:06:06,520
Speaker 1: hurt him but could also sort of help him in

110
00:06:06,600 --> 00:06:08,880
Speaker 1: the sense that the defense is going to say, yeah,

111
00:06:08,920 --> 00:06:10,880
Speaker 1: this is a guy that just kind of thinks out loud.

112
00:06:10,920 --> 00:06:13,560
Speaker 1: It goes straight from sinnase to mouth, and everyone has

113
00:06:13,560 --> 00:06:16,200
Speaker 1: known that. And at the time in two thousand eighteen,

114
00:06:16,240 --> 00:06:18,800
Speaker 1: he could do no wrong. And therefore even a light

115
00:06:18,920 --> 00:06:20,880
Speaker 1: hint that this is what he was thinking about, which

116
00:06:20,920 --> 00:06:22,919
Speaker 1: would have been accurate, would have moved the market like

117
00:06:22,960 --> 00:06:26,839
Speaker 1: crazy as well. In picking the jury, it just showed

118
00:06:26,920 --> 00:06:30,920
Speaker 1: how many people are now viewing Musk, particularly from that area.

119
00:06:31,360 --> 00:06:36,599
Speaker 1: One person called him a delusional narcissist, another one an idiot.

120
00:06:36,680 --> 00:06:39,560
Speaker 1: The cars are beautiful, but Mr Musk is an idiot.

121
00:06:40,040 --> 00:06:45,279
Speaker 1: It seems certainly different from prior jurors. That does seem

122
00:06:45,279 --> 00:06:48,280
Speaker 1: different from prior jurors, And there definitely is a is

123
00:06:48,320 --> 00:06:50,640
Speaker 1: a sense, you know, probably not just in the Bay area,

124
00:06:50,960 --> 00:06:53,080
Speaker 1: may even be a little bit more national that the

125
00:06:53,160 --> 00:06:56,320
Speaker 1: Emperor actually hasn't been wearing clothes or at least as

126
00:06:56,320 --> 00:06:59,440
Speaker 1: many clothes this entire time. And so you know, on

127
00:06:59,480 --> 00:07:02,159
Speaker 1: some uble the Must teams, you know, even tried to

128
00:07:02,200 --> 00:07:04,720
Speaker 1: move the Cape out of California claiming he can't get

129
00:07:04,760 --> 00:07:07,520
Speaker 1: can't get a fair trial in the area. That was

130
00:07:07,560 --> 00:07:11,160
Speaker 1: denied by the judge. I think probably appropriately so, but

131
00:07:11,520 --> 00:07:13,760
Speaker 1: the jury is going to be a little bit more

132
00:07:13,840 --> 00:07:16,240
Speaker 1: hostile to Mr Musk, and I think part of what's

133
00:07:16,280 --> 00:07:19,200
Speaker 1: going to be really critical is to make sure that

134
00:07:19,240 --> 00:07:22,640
Speaker 1: the jury stays on track, certainly from the Must team's perspective,

135
00:07:22,720 --> 00:07:25,440
Speaker 1: that the jury stays on track on what the ingredients

136
00:07:25,480 --> 00:07:29,240
Speaker 1: are of a securities Froud lawsuit. And it turns out

137
00:07:29,560 --> 00:07:32,800
Speaker 1: one of the ingredients for a successful securities Froud lawsuit

138
00:07:33,240 --> 00:07:37,360
Speaker 1: is not this person is unlikable or detestable, right, And

139
00:07:37,440 --> 00:07:40,080
Speaker 1: so you know, I think if the judge basically says, okay,

140
00:07:40,160 --> 00:07:42,600
Speaker 1: people on the jury, let's stick to the elements of

141
00:07:42,640 --> 00:07:45,760
Speaker 1: this event, it becomes kind of an interesting question about

142
00:07:45,760 --> 00:07:50,200
Speaker 1: how someone who's iconoclassics as Mr Musk then interfaces with

143
00:07:50,280 --> 00:07:53,480
Speaker 1: the market and what reasonable people might have known at

144
00:07:53,480 --> 00:07:57,240
Speaker 1: the time the idea that funding is secured, it's it's

145
00:07:57,280 --> 00:08:00,640
Speaker 1: a pretty strong form of a statement, and I think

146
00:08:01,000 --> 00:08:03,120
Speaker 1: there was, you know, almost a sense that, look, he

147
00:08:03,200 --> 00:08:05,400
Speaker 1: must have the funding somewhere if if he's saying something

148
00:08:05,480 --> 00:08:08,120
Speaker 1: like that, and so that's gonna be where things sort

149
00:08:08,120 --> 00:08:10,239
Speaker 1: of shake down. And I think part of what maybe

150
00:08:10,400 --> 00:08:12,240
Speaker 1: going to be a challenge for the jury and the

151
00:08:12,360 --> 00:08:14,960
Speaker 1: challenge for the judge is to say, you also need

152
00:08:15,000 --> 00:08:17,840
Speaker 1: to transport yourself back to two thousand eighteen to what

153
00:08:18,040 --> 00:08:21,360
Speaker 1: reasonable market participants would have thought and whether they would

154
00:08:21,360 --> 00:08:25,480
Speaker 1: have changed their response based on the precise wording used

155
00:08:25,480 --> 00:08:28,240
Speaker 1: in the tweet or the existence of the tweets. So

156
00:08:28,360 --> 00:08:31,240
Speaker 1: the defense seemed to be arguing that this was a

157
00:08:31,280 --> 00:08:35,120
Speaker 1: split second decision to tweet, but also that it was

158
00:08:35,160 --> 00:08:39,640
Speaker 1: true that Musk's intention to take Tesla private was true.

159
00:08:40,040 --> 00:08:43,400
Speaker 1: Now is it his intention or his ability to take

160
00:08:43,559 --> 00:08:48,240
Speaker 1: Tesla private when he says funding secured? It certainly seems

161
00:08:48,280 --> 00:08:50,680
Speaker 1: to have been his intention. But I think I think

162
00:08:50,720 --> 00:08:52,959
Speaker 1: what the defense is probably going to do, and it's

163
00:08:53,000 --> 00:08:56,640
Speaker 1: not a crazy theory, is to say, okay, tell you what,

164
00:08:56,760 --> 00:09:00,520
Speaker 1: Let's let's consider an alternative set effects and they're gonna

165
00:09:00,559 --> 00:09:02,880
Speaker 1: be pretty close to these, and let's ask whether you

166
00:09:02,920 --> 00:09:05,600
Speaker 1: can find a securities fraud lawsuit in that. And let's

167
00:09:05,600 --> 00:09:08,040
Speaker 1: suppose this alternative set of fact is there are two

168
00:09:08,080 --> 00:09:12,120
Speaker 1: tweets sent in August twousan eighteen. Let's say I'm thinking

169
00:09:12,160 --> 00:09:15,800
Speaker 1: about taking Tesla private at four and twenty dollars to share,

170
00:09:16,160 --> 00:09:19,400
Speaker 1: and I'm in, you know, pretty serious negotiations with a

171
00:09:19,440 --> 00:09:23,520
Speaker 1: significant provider of capital to help back right now, that

172
00:09:23,679 --> 00:09:29,080
Speaker 1: set of tweets, my sense would be correct, and they

173
00:09:29,080 --> 00:09:33,160
Speaker 1: didn't really omit anything as of the time that they

174
00:09:33,240 --> 00:09:36,400
Speaker 1: are made. Uh, and then you know he comes back

175
00:09:36,400 --> 00:09:39,040
Speaker 1: in and says, no, that everything is off. I think

176
00:09:39,040 --> 00:09:42,920
Speaker 1: in that alternative hypothetical it would be very very hard

177
00:09:43,240 --> 00:09:47,080
Speaker 1: to find securities fraud violation because you know, they're really

178
00:09:47,440 --> 00:09:49,840
Speaker 1: his cards were on the table, he had aspirations, the

179
00:09:49,880 --> 00:09:53,240
Speaker 1: aspirations didn't work out. And I think what the defense

180
00:09:53,320 --> 00:09:55,800
Speaker 1: is gonna probably do is say, look, if you look

181
00:09:55,880 --> 00:09:59,040
Speaker 1: at that type of an alternative hypothetical, how would the

182
00:09:59,080 --> 00:10:01,120
Speaker 1: markets have moved? He said, all would it have been

183
00:10:01,120 --> 00:10:03,200
Speaker 1: pretty much the same thing that we saw it? Like,

184
00:10:03,280 --> 00:10:05,680
Speaker 1: everyone got excited, Oh, Elon is going to be able

185
00:10:05,720 --> 00:10:07,920
Speaker 1: to do this. It's his intent, but of course he can,

186
00:10:08,080 --> 00:10:10,520
Speaker 1: you know, make magic happen, and the market would have

187
00:10:10,520 --> 00:10:12,800
Speaker 1: done exactly what it was about to do. You know,

188
00:10:12,840 --> 00:10:15,440
Speaker 1: members of the plaintiff class would have sold or you know,

189
00:10:15,480 --> 00:10:18,320
Speaker 1: unwound their positions exactly the same the way that they

190
00:10:18,320 --> 00:10:20,760
Speaker 1: would have if that is all true. If the jury

191
00:10:20,800 --> 00:10:23,920
Speaker 1: finds it that that's true, then the defense will have

192
00:10:24,520 --> 00:10:28,520
Speaker 1: drawn a line between the hypothetical that didn't happen but

193
00:10:28,920 --> 00:10:32,600
Speaker 1: pretty clearly wouldn't have big securities, broad implications, and the

194
00:10:32,679 --> 00:10:35,200
Speaker 1: stuff that actually did. And my guess is they're going

195
00:10:35,240 --> 00:10:38,160
Speaker 1: to try to appease that analogy as far as they can.

196
00:10:38,520 --> 00:10:40,680
Speaker 1: From the plantiffs perspective. They're going to have to beat

197
00:10:40,679 --> 00:10:42,480
Speaker 1: back on that. They're going to have to say, Look,

198
00:10:42,800 --> 00:10:46,120
Speaker 1: words matter, and particularly these types of words matter. And

199
00:10:46,160 --> 00:10:48,160
Speaker 1: the fact that you know, this is just something that

200
00:10:48,200 --> 00:10:51,000
Speaker 1: he casually let out, well, that's on him. That's what

201
00:10:51,080 --> 00:10:53,920
Speaker 1: security Throad laws in parts it's supposed to do. It's

202
00:10:53,960 --> 00:10:57,559
Speaker 1: supposed to provide it to turrent for um, a poorly

203
00:10:57,600 --> 00:10:59,800
Speaker 1: thought out half truth that people out there on the

204
00:10:59,800 --> 00:11:02,920
Speaker 1: mark it would reasonably rely on. And so, you know,

205
00:11:02,960 --> 00:11:05,240
Speaker 1: I think there really will be kind of this odd

206
00:11:05,320 --> 00:11:09,319
Speaker 1: fight about just how far away this case was from

207
00:11:09,880 --> 00:11:13,400
Speaker 1: the you know, no liability at all, hypothetical versus the

208
00:11:13,480 --> 00:11:18,760
Speaker 1: sort of a diabolical, full blown, detailed untruth. The other

209
00:11:18,800 --> 00:11:22,120
Speaker 1: thing is interesting about this case is that there's a

210
00:11:22,200 --> 00:11:24,800
Speaker 1: case that it's in litigation. You know, there's so much

211
00:11:24,800 --> 00:11:28,480
Speaker 1: of securities problification, particularly securities cross class actions. They just

212
00:11:28,520 --> 00:11:31,400
Speaker 1: don't go to finish line. They just don't go into litigation.

213
00:11:31,600 --> 00:11:34,400
Speaker 1: And so the handwriting seems to be on the wall

214
00:11:34,440 --> 00:11:37,080
Speaker 1: as to how the case likely to come out, and

215
00:11:37,320 --> 00:11:40,160
Speaker 1: the parties then figure out some sort of a settlement

216
00:11:40,160 --> 00:11:43,120
Speaker 1: and everyone walked away, and that's clearly not happened in

217
00:11:43,120 --> 00:11:45,760
Speaker 1: this case. You know, Mr Musk himself, as we can

218
00:11:45,800 --> 00:11:50,760
Speaker 1: tell from his seemingly continual appearance in various courtrooms states,

219
00:11:50,760 --> 00:11:53,760
Speaker 1: in federal around the country, is willing to take things

220
00:11:53,840 --> 00:11:56,400
Speaker 1: to the finish line often or at least a right

221
00:11:56,520 --> 00:11:58,760
Speaker 1: up to the very edge. And so it's kind of

222
00:11:58,800 --> 00:12:02,160
Speaker 1: an interesting phenomen on because the number of data points

223
00:12:02,240 --> 00:12:05,199
Speaker 1: that we have for what happens in a security sproad

224
00:12:05,240 --> 00:12:08,280
Speaker 1: trial he is surprisingly small. Most of them was, you know,

225
00:12:08,320 --> 00:12:10,200
Speaker 1: we know it awful lot what happened as we're going

226
00:12:10,200 --> 00:12:12,160
Speaker 1: through the process he's leading up to a trial, and

227
00:12:12,160 --> 00:12:13,960
Speaker 1: then of course there's a settlement. I think some new

228
00:12:14,040 --> 00:12:16,440
Speaker 1: lawsuit will be filed tomorrow. So there is kind of

229
00:12:16,440 --> 00:12:19,199
Speaker 1: an interesting aspect of this case that even securities for

230
00:12:19,240 --> 00:12:21,880
Speaker 1: aad litigators are kind of sitting back a thing. I

231
00:12:21,880 --> 00:12:23,719
Speaker 1: guess they are going to trial. You know, this will

232
00:12:23,760 --> 00:12:26,280
Speaker 1: be at least entertaining and probably interesting to see how

233
00:12:26,280 --> 00:12:30,839
Speaker 1: it plays out. How important is Musque's testimony, I think

234
00:12:30,840 --> 00:12:34,239
Speaker 1: it would be important for a couple of different purposes.

235
00:12:34,240 --> 00:12:36,480
Speaker 1: I think one is that he's gonna want to be

236
00:12:36,600 --> 00:12:40,839
Speaker 1: able to tell what was happening from his perspective right,

237
00:12:40,920 --> 00:12:43,640
Speaker 1: And it does seem that there is a story there

238
00:12:43,679 --> 00:12:46,120
Speaker 1: about him, you know, sort of reaching out to the

239
00:12:46,160 --> 00:12:49,240
Speaker 1: Saudiast Sovereign Wealth Fund that had indicated that they were

240
00:12:49,320 --> 00:12:53,040
Speaker 1: interested in backing this sort of transaction, and I think

241
00:12:53,080 --> 00:12:56,160
Speaker 1: there was even an attempt to subtena their appearance and

242
00:12:56,280 --> 00:12:59,200
Speaker 1: court that seems to have gone away, at least momentarily.

243
00:12:59,679 --> 00:13:01,720
Speaker 1: But I think that he's going to have to provide

244
00:13:02,000 --> 00:13:05,640
Speaker 1: some information about what was going through his head and

245
00:13:05,679 --> 00:13:08,640
Speaker 1: where he thought that process was, because his intent is

246
00:13:08,640 --> 00:13:11,960
Speaker 1: gonna matter. So his ability to come off both with

247
00:13:12,040 --> 00:13:16,360
Speaker 1: an accurate recollection of what went on and a credibility that, Yeah,

248
00:13:16,480 --> 00:13:18,160
Speaker 1: I really kind of honestly thought we were going to

249
00:13:18,240 --> 00:13:20,959
Speaker 1: be able to make this happen, so that negotiations far

250
00:13:21,040 --> 00:13:24,040
Speaker 1: enough long. Probably shouldn't used the word secured, but you know,

251
00:13:24,280 --> 00:13:26,480
Speaker 1: from what I knew at the time, the funding seemed

252
00:13:26,520 --> 00:13:29,280
Speaker 1: like it was pretty solidly lined up, and you know,

253
00:13:29,440 --> 00:13:30,960
Speaker 1: we of course had the paper and I had to

254
00:13:30,960 --> 00:13:33,200
Speaker 1: lower it up and study's had to lower it up,

255
00:13:33,240 --> 00:13:35,440
Speaker 1: but it seemed like it was going to happen, and

256
00:13:35,480 --> 00:13:37,240
Speaker 1: then they pulled the rug out from under me, and

257
00:13:37,240 --> 00:13:39,960
Speaker 1: that's when I had to send the corrective tweet. So

258
00:13:40,040 --> 00:13:43,160
Speaker 1: I think his testimony is actually gonna be pretty critical

259
00:13:43,320 --> 00:13:46,840
Speaker 1: for just sort of making people understand what was going

260
00:13:47,000 --> 00:13:50,959
Speaker 1: through his mind and doing so in incredible way. If

261
00:13:51,000 --> 00:13:53,560
Speaker 1: he can't tell a credible story, if he doesn't do

262
00:13:53,640 --> 00:13:56,280
Speaker 1: it in a way that the jury can kind of

263
00:13:56,280 --> 00:13:59,760
Speaker 1: get their head inside, it's gonna be problematic from his

264
00:14:00,000 --> 00:14:02,760
Speaker 1: spective because it's then gonna be much harder for him

265
00:14:02,840 --> 00:14:05,240
Speaker 1: to make the claim, Yeah, this is just an improvident

266
00:14:05,360 --> 00:14:07,960
Speaker 1: choice of words, and I really had no intention to

267
00:14:08,360 --> 00:14:11,920
Speaker 1: manipulate or perpetrate a fraud on the market. So he's

268
00:14:12,000 --> 00:14:16,640
Speaker 1: testified several times at trials, and I believe he's won

269
00:14:16,679 --> 00:14:19,800
Speaker 1: them all. But can lawyers get under his skin if

270
00:14:19,800 --> 00:14:22,840
Speaker 1: they know which buttons to push? Yeah, I think lawyers

271
00:14:22,880 --> 00:14:26,320
Speaker 1: can get under his skin. In fact, lawyers have gotten

272
00:14:26,400 --> 00:14:31,000
Speaker 1: under his skin, including in prior securities litigation involving some

273
00:14:31,040 --> 00:14:33,480
Speaker 1: of these same facts right that you know, he reached

274
00:14:33,480 --> 00:14:38,000
Speaker 1: the settlement with the sec but you know was clearly

275
00:14:38,040 --> 00:14:41,680
Speaker 1: troubled by you know, government lawyers and by some accounts

276
00:14:41,800 --> 00:14:44,120
Speaker 1: even went on and the debt is against some of them.

277
00:14:44,160 --> 00:14:46,520
Speaker 1: So I think it is possible for people to get

278
00:14:46,600 --> 00:14:51,000
Speaker 1: under his skin. It's interesting, kind of weird musk fits

279
00:14:51,120 --> 00:14:53,720
Speaker 1: right now because in some ways the fact that it

280
00:14:53,800 --> 00:14:57,840
Speaker 1: has been a tough couple of months on him may

281
00:14:58,000 --> 00:15:01,160
Speaker 1: cause him to have a little bit more our equanimity

282
00:15:01,360 --> 00:15:04,920
Speaker 1: and how he sort of goes about interacting with with

283
00:15:05,040 --> 00:15:08,560
Speaker 1: lawyers as well. But you know, I think this is

284
00:15:08,600 --> 00:15:11,240
Speaker 1: in some ways, you know, part of what the sort

285
00:15:11,240 --> 00:15:15,200
Speaker 1: of strategic is going on in the back channels here

286
00:15:15,280 --> 00:15:17,360
Speaker 1: with the plain insplayers saying, yeah, I just want to

287
00:15:17,360 --> 00:15:20,040
Speaker 1: get him on the stand and get him ruffled, and

288
00:15:20,400 --> 00:15:23,080
Speaker 1: the defense storney say, now, you've got a very simple story.

289
00:15:23,360 --> 00:15:25,920
Speaker 1: If your story, go back to your story that's the

290
00:15:25,920 --> 00:15:28,400
Speaker 1: story here, and if that's the story that the jury believes,

291
00:15:28,640 --> 00:15:32,520
Speaker 1: then you know, we can defend uh this lawsuit. But absolutely,

292
00:15:32,520 --> 00:15:34,520
Speaker 1: when you get someone on the stand and they're being

293
00:15:34,560 --> 00:15:37,240
Speaker 1: cross examined, part of the whole idea is to try

294
00:15:37,280 --> 00:15:40,520
Speaker 1: to ruffle their feathers and uh get them to start

295
00:15:40,600 --> 00:15:43,680
Speaker 1: monologuing in ways that just weren't planned. How would they

296
00:15:43,720 --> 00:15:47,600
Speaker 1: determine damages here? It's kind of a complex process because

297
00:15:47,600 --> 00:15:49,640
Speaker 1: there's a whole bunch of different people that are part

298
00:15:49,680 --> 00:15:53,440
Speaker 1: of this class action. So typically what's going to happen is,

299
00:15:53,520 --> 00:15:55,520
Speaker 1: you know, when when a case settles things that a

300
00:15:55,760 --> 00:15:59,080
Speaker 1: claim administrator there could be a large claim administrator after

301
00:15:59,160 --> 00:16:04,520
Speaker 1: a damages judgment that then has to figure out, you know,

302
00:16:04,720 --> 00:16:08,160
Speaker 1: who is gonna be part of what you know, chunk

303
00:16:08,200 --> 00:16:11,480
Speaker 1: of the damages. If we get to a damages phase

304
00:16:11,560 --> 00:16:14,840
Speaker 1: of this trial, then the plaintifts are then going to

305
00:16:14,960 --> 00:16:16,800
Speaker 1: have to sort of say, okay, we've got one group

306
00:16:16,840 --> 00:16:21,520
Speaker 1: of plaintiffs that, um, you know, they they you know,

307
00:16:21,800 --> 00:16:24,440
Speaker 1: right after the announcement, they went out and bought a

308
00:16:24,440 --> 00:16:27,320
Speaker 1: bunch of Tesla stock and then it immediately sunk and

309
00:16:27,320 --> 00:16:29,600
Speaker 1: and and they bought and sold during the period of

310
00:16:29,600 --> 00:16:31,640
Speaker 1: the fraud. And here's the difference, and we're going to

311
00:16:31,680 --> 00:16:34,840
Speaker 1: be totaled that up across this class of investors. There's

312
00:16:34,840 --> 00:16:37,600
Speaker 1: another class investors, including the name plaintiff as far as

313
00:16:37,640 --> 00:16:41,040
Speaker 1: I can tell, who has a slightly more complicated situation,

314
00:16:41,120 --> 00:16:45,200
Speaker 1: but because he had actually had a lot of derivatives

315
00:16:45,200 --> 00:16:49,000
Speaker 1: and options in in Tesla, and there were quite a

316
00:16:49,000 --> 00:16:52,000
Speaker 1: few investors that had this that you know, we're basically

317
00:16:52,000 --> 00:16:56,240
Speaker 1: making long term bests on the viability of Teslat, you know,

318
00:16:56,360 --> 00:16:58,640
Speaker 1: five or ten years down the road, and those bets

319
00:16:58,680 --> 00:17:02,000
Speaker 1: paid off to the extent that Tesla kept growing well

320
00:17:02,040 --> 00:17:05,520
Speaker 1: in excess of four dollars. So as soon as like

321
00:17:05,560 --> 00:17:08,560
Speaker 1: a four hundred twenty dollar um you know, go private

322
00:17:08,600 --> 00:17:12,959
Speaker 1: is announced or sort of announced, Uh, those folks just

323
00:17:13,040 --> 00:17:15,800
Speaker 1: their ownership of of of those claims, it kind of

324
00:17:15,800 --> 00:17:18,359
Speaker 1: loses a bunch of its you know, right hand side

325
00:17:18,359 --> 00:17:21,000
Speaker 1: tail and they lose they lose a lot of value.

326
00:17:21,040 --> 00:17:23,119
Speaker 1: And you have a name planet said I just dumped

327
00:17:23,280 --> 00:17:26,119
Speaker 1: you had I had just dumped my my position, And

328
00:17:26,160 --> 00:17:28,800
Speaker 1: so they're gonna have to, you know, figure out the

329
00:17:28,800 --> 00:17:30,720
Speaker 1: planets are gonna have figure out, assuming they can get

330
00:17:30,720 --> 00:17:34,800
Speaker 1: a liability judgment, how to then a portion damages claims

331
00:17:35,000 --> 00:17:37,480
Speaker 1: within the class. How many people are in each group

332
00:17:37,560 --> 00:17:39,760
Speaker 1: and when they thought when they sold, and what that

333
00:17:39,960 --> 00:17:42,800
Speaker 1: gap was, and how much of that gap was due

334
00:17:42,920 --> 00:17:47,560
Speaker 1: to this idiosyncratic movement in testa stock price versus just

335
00:17:47,600 --> 00:17:51,160
Speaker 1: the market going haywire on a given day, and so um,

336
00:17:51,520 --> 00:17:53,359
Speaker 1: if they get to this point of then you know,

337
00:17:53,400 --> 00:17:57,639
Speaker 1: trying to to lay out damages, it could involve a

338
00:17:57,800 --> 00:18:01,240
Speaker 1: fairly detailed taking a part of the of the plaintiff class,

339
00:18:01,720 --> 00:18:04,240
Speaker 1: probably with some financial experts who are sort of coming

340
00:18:04,240 --> 00:18:06,240
Speaker 1: in and trying to say, Okay, here's how I value

341
00:18:06,720 --> 00:18:09,440
Speaker 1: the laws due to the fraud versus just market risk

342
00:18:09,560 --> 00:18:11,760
Speaker 1: that was going on. Which side do you think has

343
00:18:11,800 --> 00:18:14,720
Speaker 1: a more difficult case to make out? Well, you know,

344
00:18:14,800 --> 00:18:17,720
Speaker 1: in theory, the plaintiffs have to prove their case, and

345
00:18:17,760 --> 00:18:21,480
Speaker 1: so all the presumptions go against the plaintiffs, and so

346
00:18:21,640 --> 00:18:25,800
Speaker 1: as a general matter, the deck is stacked against the plaintiffs. Here,

347
00:18:25,800 --> 00:18:29,320
Speaker 1: we're not in quite that position because we've already had

348
00:18:29,359 --> 00:18:32,040
Speaker 1: a finding by the judge that this is a false

349
00:18:32,080 --> 00:18:35,240
Speaker 1: and misleading statement and mustnewers should have known that it

350
00:18:35,320 --> 00:18:38,600
Speaker 1: was false and misleading, which is actually you know, part

351
00:18:38,640 --> 00:18:41,639
Speaker 1: of the way to a successful lawsuit. So where with

352
00:18:41,760 --> 00:18:44,880
Speaker 1: the ordinary case, you know, your plantists realize that it's

353
00:18:44,880 --> 00:18:47,439
Speaker 1: a long shot. They only have to win part the

354
00:18:47,480 --> 00:18:50,160
Speaker 1: battle from here on in because they've already won part

355
00:18:50,200 --> 00:18:52,880
Speaker 1: of it. So this case is a little bit more

356
00:18:52,960 --> 00:18:56,960
Speaker 1: of kind of an evenly matched case, and there is,

357
00:18:57,200 --> 00:18:59,840
Speaker 1: you know, sort of a sense in which the oddity

358
00:19:00,080 --> 00:19:02,520
Speaker 1: this case and regally the oddity of you know, so

359
00:19:02,560 --> 00:19:05,320
Speaker 1: many of these other recent lawsuits that involved Mr Musk

360
00:19:05,320 --> 00:19:08,600
Speaker 1: in particulars that you know, he is such an iconoclass

361
00:19:08,600 --> 00:19:12,720
Speaker 1: in such a non comparable personality, it's kind of hard

362
00:19:12,760 --> 00:19:15,640
Speaker 1: to figure out how to how to fit his square

363
00:19:15,680 --> 00:19:18,560
Speaker 1: peg into the round holes that most of these areas

364
00:19:18,600 --> 00:19:21,920
Speaker 1: of law have. And in this case, as opposed to

365
00:19:21,960 --> 00:19:24,560
Speaker 1: the Delaware case, we're actually at a jury on top

366
00:19:24,600 --> 00:19:26,400
Speaker 1: of that, and so there's there's gonna be a lot

367
00:19:26,480 --> 00:19:31,000
Speaker 1: of how do we understand and socialize the jury? You know, personally,

368
00:19:31,040 --> 00:19:34,159
Speaker 1: I think that the defense still has a reasonable shot

369
00:19:34,320 --> 00:19:37,720
Speaker 1: if they could make a convincing case that Elon Musk,

370
00:19:37,880 --> 00:19:42,160
Speaker 1: you know, sneezing or scratching his elbow would move markets

371
00:19:42,200 --> 00:19:46,000
Speaker 1: in two thousand eighteen, that the gap between his actual

372
00:19:46,040 --> 00:19:50,560
Speaker 1: tweets and these much more non actionable hypothetical tweets was

373
00:19:50,600 --> 00:19:53,040
Speaker 1: pretty small. That's a real argument, and I think it's

374
00:19:53,080 --> 00:19:55,960
Speaker 1: going to get a lot of airtime this week. But

375
00:19:56,240 --> 00:19:58,879
Speaker 1: you know, given the plaintifts have already crossed over some

376
00:19:59,000 --> 00:20:02,560
Speaker 1: of the critical hurdles that usually trips up other plaintiffs,

377
00:20:02,600 --> 00:20:05,320
Speaker 1: you know, I think they're shooting kind of an even

378
00:20:05,400 --> 00:20:08,520
Speaker 1: money gamble at this day, and it'll be interesting plays out.

379
00:20:08,800 --> 00:20:12,399
Speaker 1: Has he won all the lawsuit he testified in. Yeah,

380
00:20:12,480 --> 00:20:14,879
Speaker 1: it depends how you add the mom It is true

381
00:20:14,960 --> 00:20:18,520
Speaker 1: that when he has testified thus far, he has won

382
00:20:18,840 --> 00:20:21,800
Speaker 1: the lawsuits that he's been in. But you know, in

383
00:20:21,840 --> 00:20:24,680
Speaker 1: some cases, you know, the Twitter case, they just settled

384
00:20:24,720 --> 00:20:26,600
Speaker 1: at the eleventh hour, and I think it looks pretty

385
00:20:26,600 --> 00:20:29,840
Speaker 1: clear that things were not looking good for him. Um,

386
00:20:30,080 --> 00:20:33,639
Speaker 1: he ended up reaching a settlement with the SEC regarding

387
00:20:33,680 --> 00:20:36,480
Speaker 1: the same set of actions, and so that kind of

388
00:20:36,520 --> 00:20:39,320
Speaker 1: never went into trial. There's a there's a trial that

389
00:20:39,359 --> 00:20:43,280
Speaker 1: he had involving the acquisition of Solar City in which

390
00:20:43,320 --> 00:20:46,640
Speaker 1: he testified and he won, but that case is currently

391
00:20:46,680 --> 00:20:49,600
Speaker 1: on appeal to the Delaware Supreme Court, and so it's

392
00:20:49,640 --> 00:20:52,560
Speaker 1: conceivable that could be reversed and and then we're right

393
00:20:52,600 --> 00:20:55,480
Speaker 1: back in the same in the same boat. So yeah,

394
00:20:55,520 --> 00:20:57,639
Speaker 1: I think it's probably I think it is fair to

395
00:20:57,640 --> 00:21:00,080
Speaker 1: say that he's at least these sort of corporate to

396
00:21:00,240 --> 00:21:03,359
Speaker 1: these big corporate securities testlo related cases. He's one all

397
00:21:03,359 --> 00:21:06,480
Speaker 1: the big ones that he's that he's testified in. But

398
00:21:06,480 --> 00:21:10,720
Speaker 1: but he settled others um and you know, and some

399
00:21:10,880 --> 00:21:13,720
Speaker 1: of them still haven't played out. We don't have a

400
00:21:13,920 --> 00:21:16,960
Speaker 1: we don't have a judgment from a chance for McCormick

401
00:21:17,040 --> 00:21:20,600
Speaker 1: yet in the in the executive compensation case, Solar City

402
00:21:20,640 --> 00:21:23,840
Speaker 1: could get reversed. We don't know what's going to happen here.

403
00:21:23,920 --> 00:21:26,840
Speaker 1: So the roulette wheel is still sitting on the effect

404
00:21:26,840 --> 00:21:29,680
Speaker 1: of Elon Musk testifying life the court. In my mind,

405
00:21:29,840 --> 00:21:32,520
Speaker 1: you know, Musk does seem to have a lot in

406
00:21:32,600 --> 00:21:37,960
Speaker 1: common with Donald Trump as far as litigating everything to

407
00:21:38,040 --> 00:21:42,280
Speaker 1: the end degree and not settling. You can't not make

408
00:21:42,359 --> 00:21:46,240
Speaker 1: the parallel comparisons, right, because I think they are both

409
00:21:46,280 --> 00:21:50,119
Speaker 1: people who sort of feel like, yeah, that those rules

410
00:21:50,160 --> 00:21:53,000
Speaker 1: don't really quite apply to me, and I'm going to

411
00:21:53,080 --> 00:21:55,840
Speaker 1: fund my nose at them and kind of be stunned

412
00:21:55,920 --> 00:21:58,960
Speaker 1: when I find out in certain circumstances, Oh my gosh,

413
00:21:59,080 --> 00:22:01,359
Speaker 1: they do apply to me because a lot of time,

414
00:22:01,720 --> 00:22:04,919
Speaker 1: you know, it turns out, for some weird reason, I

415
00:22:05,000 --> 00:22:06,720
Speaker 1: got away with that strategy, and then I'm gonna do

416
00:22:06,760 --> 00:22:09,040
Speaker 1: it again and again and again. And that's a personality

417
00:22:09,040 --> 00:22:11,879
Speaker 1: trait that that that both of these guys have in common.

418
00:22:12,160 --> 00:22:15,440
Speaker 1: Thanks so much, Eric, that's Professor Eric Tally of Columbia

419
00:22:15,520 --> 00:22:21,159
Speaker 1: Law School. AI art generators have enjoyed explosive growth in

420
00:22:21,280 --> 00:22:24,960
Speaker 1: recent months, and now a first of its kind copyright

421
00:22:25,000 --> 00:22:29,760
Speaker 1: infringement class action suit over using copyrighted images to train

422
00:22:29,960 --> 00:22:33,520
Speaker 1: artificial intelligence tools. A group of artists has filed a

423
00:22:33,560 --> 00:22:38,760
Speaker 1: potential class action against billion dollar company Stability AI, along

424
00:22:38,800 --> 00:22:42,800
Speaker 1: with two other art generator makers, mid Journey and deviant Art.

425
00:22:43,359 --> 00:22:46,640
Speaker 1: The claim is that the generators downloaded and used billions

426
00:22:46,680 --> 00:22:51,159
Speaker 1: of copyrighted images without obtaining the consent of or compensating

427
00:22:51,200 --> 00:22:54,399
Speaker 1: any of the artists. These lawsuits could limit the number

428
00:22:54,400 --> 00:22:58,480
Speaker 1: of images the tools in ingest for training, ultimately affecting

429
00:22:58,480 --> 00:23:01,920
Speaker 1: the content they produce. Joining me is cal Rowse Tiella,

430
00:23:02,040 --> 00:23:04,640
Speaker 1: a professor at u c l A Law School. Will

431
00:23:04,680 --> 00:23:07,480
Speaker 1: you explain the general issue with the use of AI

432
00:23:07,520 --> 00:23:12,160
Speaker 1: in these cases? Sure? So many artists are concerned by

433
00:23:12,280 --> 00:23:18,520
Speaker 1: the fact that AI technologies train themselves on large data

434
00:23:18,560 --> 00:23:23,439
Speaker 1: sets of photographs, paintings, music, et cetera. And in the

435
00:23:23,480 --> 00:23:27,480
Speaker 1: process of doing that, they consume millions and millions of

436
00:23:27,520 --> 00:23:33,800
Speaker 1: different copyrighted content pieces, artwork, songs, et cetera. And so

437
00:23:33,960 --> 00:23:36,360
Speaker 1: the concern of many artists is in doing that, they

438
00:23:36,400 --> 00:23:39,720
Speaker 1: are one copying their work when they're training on it.

439
00:23:40,320 --> 00:23:42,960
Speaker 1: But too, I think at a deeper level, those ayes

440
00:23:43,000 --> 00:23:48,960
Speaker 1: are than producing incredibly interesting creative work that challenge the

441
00:23:49,000 --> 00:23:52,200
Speaker 1: existing works in the marketplace. There are several of these

442
00:23:52,280 --> 00:23:56,399
Speaker 1: lawsuits going on. Yeah, there's a number of lawsuits, and

443
00:23:56,480 --> 00:24:00,000
Speaker 1: in some ways, some of these lawsuits reflect even earlier lawsuits,

444
00:24:00,080 --> 00:24:02,639
Speaker 1: like there was one quite a while ago around Google Books,

445
00:24:02,720 --> 00:24:06,520
Speaker 1: and so this isn't an entirely new phenomenon, but you know,

446
00:24:06,600 --> 00:24:10,920
Speaker 1: all of the chatbot, GPT, these other more recent AI

447
00:24:11,160 --> 00:24:15,320
Speaker 1: technologies have really led to a lot of alarms on

448
00:24:15,400 --> 00:24:19,200
Speaker 1: the part of content creators. Yeah. One of the lawyers

449
00:24:19,240 --> 00:24:22,080
Speaker 1: in a blog post said that since the November lawsuit,

450
00:24:22,119 --> 00:24:26,879
Speaker 1: they've heard from people all over the world, especially writers, artists, programmers,

451
00:24:26,880 --> 00:24:30,560
Speaker 1: and other creators who are concerned about AI systems being

452
00:24:30,640 --> 00:24:34,320
Speaker 1: trained on vast amounts of copyrighted work with no consent,

453
00:24:34,520 --> 00:24:38,200
Speaker 1: no credit, and no compensation. Is there any other way

454
00:24:38,240 --> 00:24:43,639
Speaker 1: to train these AI systems? You know, I'm not a technologist,

455
00:24:43,680 --> 00:24:46,840
Speaker 1: I'm a lawyer, but my understanding is no, that this

456
00:24:46,920 --> 00:24:51,520
Speaker 1: is how they in essence learn, And you know, it

457
00:24:51,600 --> 00:24:54,840
Speaker 1: raises a complicated issue because of course, you can imagine

458
00:24:55,200 --> 00:24:57,840
Speaker 1: you or I going through a museum. I'm looking at

459
00:24:57,880 --> 00:25:00,600
Speaker 1: paintings and back getting and studying on then going home,

460
00:25:00,640 --> 00:25:04,159
Speaker 1: and paintings totally legal as long as we're not actually

461
00:25:04,160 --> 00:25:08,000
Speaker 1: copying a specific element of a specific painting. We're just

462
00:25:08,000 --> 00:25:11,280
Speaker 1: getting inspiration, we're learning about painting. These are doing that

463
00:25:11,400 --> 00:25:15,040
Speaker 1: same process but on a massive scale, and in doing it,

464
00:25:15,080 --> 00:25:17,879
Speaker 1: they're actually copying the work to create a kind of

465
00:25:17,920 --> 00:25:21,119
Speaker 1: training day set. So it's both the scale and the

466
00:25:21,160 --> 00:25:24,960
Speaker 1: particular way it's done that opens up the possibility of

467
00:25:25,000 --> 00:25:26,720
Speaker 1: these lawsuits. I'm not going to say that they're going

468
00:25:26,760 --> 00:25:30,120
Speaker 1: to prevail, but at least opened the door. UH spokesperson

469
00:25:30,280 --> 00:25:34,880
Speaker 1: for Stability AI said that the company takes these matters

470
00:25:34,920 --> 00:25:37,680
Speaker 1: seriously and that anyone that believes that this is in

471
00:25:37,840 --> 00:25:42,800
Speaker 1: fair use does not understand the technology and misunderstands the law.

472
00:25:43,320 --> 00:25:47,160
Speaker 1: So it seems like their defense will be fair use. Yeah,

473
00:25:47,400 --> 00:25:51,040
Speaker 1: fair use is appears to be the primary defense that

474
00:25:51,119 --> 00:25:53,720
Speaker 1: will be mounted. Of course, we will see if things proceeds.

475
00:25:54,280 --> 00:25:58,000
Speaker 1: The concept of fair use is longstanding and it's sort

476
00:25:58,000 --> 00:26:01,160
Speaker 1: of reflects the idea that I think we all recognize

477
00:26:01,240 --> 00:26:05,280
Speaker 1: that no one creates artwork, songs, whatever, the content is

478
00:26:06,119 --> 00:26:09,439
Speaker 1: out of nothing. You created out of a series of

479
00:26:09,480 --> 00:26:12,440
Speaker 1: things that already exist, that you know of, that you've witnessed,

480
00:26:12,440 --> 00:26:15,679
Speaker 1: that you thought about, that you've consumed, and that you

481
00:26:15,760 --> 00:26:21,480
Speaker 1: often can incorporate aspects into your own work. Now, copyright

482
00:26:21,560 --> 00:26:23,920
Speaker 1: tries to strike a balance, and fair use is meant

483
00:26:23,960 --> 00:26:28,040
Speaker 1: to provide that kind of balance. It wasn't planned or

484
00:26:28,240 --> 00:26:31,000
Speaker 1: intended to deal with things like AI because nobody thought

485
00:26:31,000 --> 00:26:34,879
Speaker 1: about this centuries ago. So this poses some real novel questions.

486
00:26:34,880 --> 00:26:36,720
Speaker 1: But the idea of fair use is not a new one,

487
00:26:36,800 --> 00:26:39,840
Speaker 1: and it is certainly applicable in theory to what the

488
00:26:39,840 --> 00:26:45,080
Speaker 1: AI programs do. This is copying on a broad scale,

489
00:26:45,480 --> 00:26:48,320
Speaker 1: copying the works of others on a broad scale. Is

490
00:26:48,359 --> 00:26:51,399
Speaker 1: that a hard argument for fair use? Yes, and no.

491
00:26:51,720 --> 00:26:55,320
Speaker 1: I think what what's different here is that most of

492
00:26:55,359 --> 00:26:59,760
Speaker 1: these AI training data sets they're copying in a notional

493
00:26:59,760 --> 00:27:02,160
Speaker 1: way into some way. This goes back to the very

494
00:27:02,240 --> 00:27:04,760
Speaker 1: dawn of the kind of computer age where when you

495
00:27:04,760 --> 00:27:08,560
Speaker 1: would download something, even just to look at momentarily, your

496
00:27:08,600 --> 00:27:12,199
Speaker 1: computer would make a little copy for a moment. I'm

497
00:27:12,240 --> 00:27:15,160
Speaker 1: obviously paraphrasing the actual technology, but the point is there's

498
00:27:15,200 --> 00:27:18,720
Speaker 1: there's often kind of notional copying that's taking place just

499
00:27:18,960 --> 00:27:22,920
Speaker 1: with regard to digital item, and so that's what's happening here.

500
00:27:23,000 --> 00:27:25,800
Speaker 1: The AI s are not copying and then selling. It's

501
00:27:25,800 --> 00:27:28,760
Speaker 1: not like Napster. Let's say twenty years ago, when music

502
00:27:28,920 --> 00:27:32,520
Speaker 1: was being downloaded, copied and then distributed out. That was

503
00:27:32,520 --> 00:27:36,359
Speaker 1: a very clear cut kind of situation. This is different.

504
00:27:37,000 --> 00:27:40,240
Speaker 1: The copying is there to train on, and then they're

505
00:27:40,240 --> 00:27:43,080
Speaker 1: producing something totally different. Might be in the style, might

506
00:27:43,160 --> 00:27:46,719
Speaker 1: look vaguely similar, but they're not actually copying specific elements

507
00:27:46,720 --> 00:27:51,359
Speaker 1: into some artwork. So then let's say selling. So it's

508
00:27:51,480 --> 00:27:54,760
Speaker 1: quite different. And so it raises a question about whether

509
00:27:55,600 --> 00:27:58,679
Speaker 1: this is something that gets comfortably within fair use as

510
00:27:58,720 --> 00:28:03,760
Speaker 1: we've understood it, or maybe requires an elaboration of the

511
00:28:03,840 --> 00:28:07,399
Speaker 1: idea of fair use for this new technology. So, for example,

512
00:28:07,440 --> 00:28:09,840
Speaker 1: some people have said, look, we need to kind of

513
00:28:09,920 --> 00:28:13,280
Speaker 1: understand this as fair learning that to learn sometimes you

514
00:28:13,320 --> 00:28:15,959
Speaker 1: need to copy. Again, it's like momentary notional, it's not.

515
00:28:16,400 --> 00:28:18,639
Speaker 1: It's not the traditional copying, like do you write a

516
00:28:18,680 --> 00:28:21,000
Speaker 1: story and I copy your story and tak it off

517
00:28:21,000 --> 00:28:24,600
Speaker 1: his mind? It's not that something different reflects the twenty century.

518
00:28:25,240 --> 00:28:29,200
Speaker 1: That might be a question for Congress. The Supreme Court

519
00:28:29,280 --> 00:28:34,439
Speaker 1: case back in Google the Oracle, which I think was

520
00:28:34,480 --> 00:28:38,520
Speaker 1: a landmark case involving computer code, does that have any

521
00:28:38,600 --> 00:28:43,080
Speaker 1: relevance here? It does because it's one of them. They're

522
00:28:43,080 --> 00:28:45,120
Speaker 1: not a ton of cases from the Spring Court on

523
00:28:45,240 --> 00:28:47,760
Speaker 1: fair use, and so yes, it absolutely does. I'll just

524
00:28:47,800 --> 00:28:50,920
Speaker 1: note that there's another case before the Court right now

525
00:28:51,080 --> 00:28:55,520
Speaker 1: involving fair use and Andy Warhol and a photograph that

526
00:28:55,600 --> 00:28:58,320
Speaker 1: raises a lot of the same kinds of issues, not

527
00:28:58,480 --> 00:29:01,320
Speaker 1: in an AI context, but again, and what is permitted?

528
00:29:01,360 --> 00:29:04,800
Speaker 1: What degree of copying or transformation which is often a

529
00:29:04,840 --> 00:29:07,440
Speaker 1: big part of the understanding of fair use is the

530
00:29:07,600 --> 00:29:10,880
Speaker 1: use of the prior work then transformed in a way

531
00:29:10,920 --> 00:29:15,120
Speaker 1: that changes the meaning and therefore is permissible. The Court

532
00:29:15,200 --> 00:29:16,960
Speaker 1: is gonna all tie on that as well in the

533
00:29:17,000 --> 00:29:20,320
Speaker 1: coming months, and that will also give us some guidance. So, yes,

534
00:29:20,400 --> 00:29:22,600
Speaker 1: this is an evolving area of law, and it's one

535
00:29:22,640 --> 00:29:25,000
Speaker 1: where the Supreme Court has a lot to say. But

536
00:29:25,080 --> 00:29:27,680
Speaker 1: traditionally the Court has not taken on a lot of

537
00:29:28,040 --> 00:29:30,080
Speaker 1: fair use cases. So it's unusual for there to be

538
00:29:30,160 --> 00:29:32,760
Speaker 1: two cases in just a couple of years. As far

539
00:29:32,840 --> 00:29:36,680
Speaker 1: as fair use. Let's say the Court hands down some

540
00:29:36,800 --> 00:29:41,720
Speaker 1: guidelines in the Warhol case. Aren't fair use cases really

541
00:29:41,920 --> 00:29:47,200
Speaker 1: facts specific? Are guidelines going to be definitive? No, that's

542
00:29:47,200 --> 00:29:50,360
Speaker 1: a fair question. They are very fast specific, you know

543
00:29:50,440 --> 00:29:52,760
Speaker 1: that said, the courts have in the past, the Spreme

544
00:29:52,800 --> 00:29:55,760
Speaker 1: Court has talked about, for example, transformation, this idea that

545
00:29:55,880 --> 00:29:59,680
Speaker 1: if you paint something and then I incorporate that into

546
00:29:59,760 --> 00:30:02,880
Speaker 1: a photograph or a painting myself that references it in

547
00:30:02,880 --> 00:30:05,600
Speaker 1: a certain way, that I transform it so that the

548
00:30:05,640 --> 00:30:08,560
Speaker 1: meaning is very different. Maybe it's securical now that that

549
00:30:08,760 --> 00:30:12,400
Speaker 1: is an acceptable form of fair use. And so they've

550
00:30:12,400 --> 00:30:17,040
Speaker 1: given us what lawyers would call standards for evaluating that

551
00:30:17,280 --> 00:30:20,320
Speaker 1: rather than hard and fast rules, because yes, every case

552
00:30:20,480 --> 00:30:25,000
Speaker 1: is distinct. It's a really complicated area, and you know,

553
00:30:25,120 --> 00:30:29,680
Speaker 1: I understand the concerns that artists have. I'm sympathetic to that.

554
00:30:29,760 --> 00:30:33,280
Speaker 1: On the other hand, you can't really imagine in a

555
00:30:33,480 --> 00:30:38,680
Speaker 1: future for AI technologies. It's this form of copying that's

556
00:30:38,720 --> 00:30:43,600
Speaker 1: being employed in training Aiyes, is suddenly subject to copyright

557
00:30:43,840 --> 00:30:46,840
Speaker 1: because there are literally millions and millions of millions of

558
00:30:46,880 --> 00:30:50,360
Speaker 1: play of images. You don't know who actually took those images.

559
00:30:50,520 --> 00:30:53,800
Speaker 1: You couldn't even begin to find the person. So there's

560
00:30:53,840 --> 00:30:57,200
Speaker 1: so many problems with that imposing that on it. So

561
00:30:57,280 --> 00:31:00,000
Speaker 1: these are really hard cases, and I want to emphasize

562
00:31:00,560 --> 00:31:03,720
Speaker 1: there is no easy solution here. Two of the attorneys

563
00:31:03,800 --> 00:31:08,440
Speaker 1: who represent the artist suing stability AI, Joseph Severi and

564
00:31:08,520 --> 00:31:12,280
Speaker 1: Michael Butterick, say it's possible to create a licensing model

565
00:31:12,400 --> 00:31:16,320
Speaker 1: similar to the music worlds transition from file sharing sites

566
00:31:16,440 --> 00:31:20,760
Speaker 1: like Napster two platforms such as Spotify and iTunes that

567
00:31:20,880 --> 00:31:25,320
Speaker 1: licensed content legally and that would allow artists and AI

568
00:31:25,440 --> 00:31:29,760
Speaker 1: to peacefully coexist. Is that a big ask? I think

569
00:31:29,800 --> 00:31:34,640
Speaker 1: it's possible. It works pretty well in music. It's not perfect.

570
00:31:34,920 --> 00:31:36,840
Speaker 1: One thing I should say is we are producing work

571
00:31:36,920 --> 00:31:40,000
Speaker 1: right now, and if it fixed in a tangible medium

572
00:31:40,200 --> 00:31:43,680
Speaker 1: of expression and it have some originality, it's copyrighted. And

573
00:31:43,760 --> 00:31:47,320
Speaker 1: so every day you are producing copyrighted work, whether you

574
00:31:47,360 --> 00:31:49,320
Speaker 1: know it or not, so I think people don't always

575
00:31:49,360 --> 00:31:52,360
Speaker 1: appreciate that. So yes, in theory, if you're like the

576
00:31:52,400 --> 00:31:55,360
Speaker 1: typical artists, he's actually making a song and a studio

577
00:31:55,520 --> 00:31:57,680
Speaker 1: trying to sell it, Yeah, that works well. If you're

578
00:31:57,720 --> 00:32:00,239
Speaker 1: just the ordinary person taking a photograph of we all

579
00:32:00,280 --> 00:32:02,920
Speaker 1: do every day on our phone, maybe uploading them to

580
00:32:03,000 --> 00:32:06,200
Speaker 1: a website or something that others can see, that's really different.

581
00:32:06,360 --> 00:32:09,320
Speaker 1: So I'm not convinced that that solution works on a

582
00:32:09,400 --> 00:32:11,719
Speaker 1: kind of mass scale, but we'll see. You know, creative

583
00:32:11,720 --> 00:32:14,000
Speaker 1: people will obviously work hard to make it viable, and

584
00:32:14,040 --> 00:32:16,160
Speaker 1: if it can work, then yes, that is a solution

585
00:32:16,200 --> 00:32:18,480
Speaker 1: we can we can imagine, and it's it's worked fairly

586
00:32:18,480 --> 00:32:21,840
Speaker 1: well for the musical. I understand that the Copyright and

587
00:32:21,840 --> 00:32:26,120
Speaker 1: Trademark Office won't register our work created solely by AI,

588
00:32:26,200 --> 00:32:31,240
Speaker 1: but it will register works created with assistance from AI.

589
00:32:31,440 --> 00:32:34,160
Speaker 1: Does that have any impact on what we've been talking about?

590
00:32:34,560 --> 00:32:36,640
Speaker 1: You know, that's sort of the other end of the process.

591
00:32:36,720 --> 00:32:39,280
Speaker 1: So can an AI be an author? Who can be

592
00:32:39,320 --> 00:32:42,560
Speaker 1: an author? And yeah, it's generally been the position in

593
00:32:42,560 --> 00:32:46,560
Speaker 1: the United States that only sort of human beings are authors.

594
00:32:46,600 --> 00:32:49,400
Speaker 1: So there's a famous case involving a monkey taking it selfie.

595
00:32:49,600 --> 00:32:52,040
Speaker 1: You know the monkey and author is an AI and author,

596
00:32:52,080 --> 00:32:54,800
Speaker 1: and the answer has generally been no, only a human

597
00:32:54,840 --> 00:32:58,120
Speaker 1: being can be an author. Now is that? The right answer?

598
00:32:58,240 --> 00:33:00,960
Speaker 1: Is one question? Some country take it slightly different for you.

599
00:33:01,280 --> 00:33:04,440
Speaker 1: The second question is what's the line between that and

600
00:33:04,600 --> 00:33:08,800
Speaker 1: using some technology for assistance you're recording our interview, I

601
00:33:08,920 --> 00:33:12,760
Speaker 1: might use photoshops. What's the line that's not so easy

602
00:33:12,800 --> 00:33:15,120
Speaker 1: to discern? So you know, I don't have a great

603
00:33:15,120 --> 00:33:17,400
Speaker 1: answer for you because no one really knows where to

604
00:33:17,440 --> 00:33:20,560
Speaker 1: strike that balance. And it's obviously changing because the technology

605
00:33:20,600 --> 00:33:23,360
Speaker 1: is always changing. Thanks for being on the show. That's

606
00:33:23,360 --> 00:33:26,120
Speaker 1: Professor cal Rols Yella of u c l A Law School,

607
00:33:26,640 --> 00:33:28,960
Speaker 1: And that's it for this edition of The Bloomberg Law Show.

608
00:33:29,320 --> 00:33:31,600
Speaker 1: Remember you can always get the latest legal news on

609
00:33:31,680 --> 00:33:35,959
Speaker 1: our Bloomberg Law Podcast. You can find them on Apple Podcasts, Spotify,

610
00:33:36,160 --> 00:33:41,200
Speaker 1: and at www dot Bloomberg dot com, slash podcast Slash Law,

611
00:33:41,600 --> 00:33:44,240
Speaker 1: And remember to tune into The Bloomberg Law Show every

612
00:33:44,240 --> 00:33:47,360
Speaker 1: week night at ten b m. Wall Street Time. I'm

613
00:33:47,440 --> 00:33:49,880
Speaker 1: June Grosso and you're listening to Bloomberg