【ARK Invest】SpaceX 秘密 IPO？2 万亿美金到底贵不贵？

1
00:00:00,140 --> 00:00:06,139
人类已经成功将探测器送到了离地球最远的地方

2
00:00:06,419 --> 00:00:09,820
这是否是因为这里的情况变得越来越糟

3
00:00:09,820 --> 00:00:10,400
布莱特

4
00:00:10,400 --> 00:00:14,570
或者我们应该为此感到自豪

5
00:00:14,570 --> 00:00:15,529
我换个角度说

6
00:00:15,529 --> 00:00:16,010
这太棒了

7
00:00:16,010 --> 00:00:19,100
我们获得了SpaceX秘密提交的IPO申请

8
00:00:19,100 --> 00:00:22,339
同时还有阿尔忒弥斯计划和NASA的项目

9
00:00:22,339 --> 00:00:24,579
这正在实现令人惊叹的成就

10
00:00:24,579 --> 00:00:27,320
但代价巨大

11
00:00:27,320 --> 00:00:27,920
呃

12
00:00:27,920 --> 00:00:31,379
你怎么看待这两种动态的发展

13
00:00:31,379 --> 00:00:32,140
呃

14
00:00:32,140 --> 00:00:39,298
随着私营与公营航天领域日益交织

15
00:00:39,298 --> 00:00:44,139
这是一系列私营发射服务商提供运载工具

16
00:00:44,139 --> 00:00:47,899
恰好执行阿尔忒弥斯任务的月球远侧探测器

17
00:00:47,899 --> 00:00:51,789
使用的运载火箭

18
00:00:51,789 --> 00:00:52,310
我的意思是

19
00:00:52,310 --> 00:00:54,850
表面上是航天飞机计划的备用部件

20
00:00:54,850 --> 00:00:55,950
但其实

21
00:00:55,950 --> 00:00:58,409
这就像设计现代汽车

22
00:00:58,409 --> 00:01:00,270
但前提是

23
00:01:00,270 --> 00:01:02,268
发动机设计

24
00:01:02,268 --> 00:01:06,168
要基于福特普利茅斯发动机

25
00:01:06,168 --> 00:01:12,709
我认为高昂的费用源于成本结构

26
00:01:12,709 --> 00:01:17,310
以及国会采取的这种仓促方式

27
00:01:17,310 --> 00:01:24,069
这证明政府资助确实能实现雄心勃勃的目标

28
00:01:24,069 --> 00:01:26,870
我相信未来我们将能完成更宏伟的项目

29
00:01:26,870 --> 00:01:32,180
一旦我们采用现代工程体系配合政府资金

30
00:01:32,180 --> 00:01:33,239
你怎么看

31
00:01:33,239 --> 00:01:33,719
是的

32
00:01:33,719 --> 00:01:37,780
我认为常被忽略的一个关键点

33
00:01:37,780 --> 00:01:39,200
这没问题

34
00:01:39,200 --> 00:01:42,280
私营部门不会浪费资金

35
00:01:42,280 --> 00:01:45,200
而NASA却挥霍无度

36
00:01:45,200 --> 00:01:49,010
这很糟糕 所以解决方案是私营公司

37
00:01:49,010 --> 00:01:51,159
但这并不正确

38
00:01:51,159 --> 00:01:51,498
没错

39
00:01:51,498 --> 00:01:55,569
如果政府仅选择资助私营企业

40
00:01:55,569 --> 00:02:00,198
且缺乏淘汰低效项目的机制

41
00:02:00,198 --> 00:02:01,918
依然会产生大量浪费

42
00:02:01,918 --> 00:02:04,469
只是通过不同渠道

43
00:02:04,469 --> 00:02:09,278
我认为关键在于竞争

44
00:02:09,278 --> 00:02:10,419
别谈公营

45
00:02:10,419 --> 00:02:11,899
私营政府

46
00:02:11,899 --> 00:02:19,500
无论哪种形式 健康的竞争总能带来更好结果

47
00:02:19,500 --> 00:02:21,460
我认为这就是

48
00:02:21,460 --> 00:02:24,259
我们现在开始看到的 当只有SLS时

49
00:02:25,259 --> 00:02:32,819
以及各种航天计划遗留的包袱

50
00:02:32,819 --> 00:02:35,289
这就是航天计划

51
00:02:35,289 --> 00:02:36,810
当我们没有去月球的时候

52
00:02:36,810 --> 00:02:38,770
其实更像是一个就业计划

53
00:02:38,770 --> 00:02:41,039
因为你知道

54
00:02:41,039 --> 00:02:41,620
你知道的

55
00:02:41,620 --> 00:02:43,900
南方各州拥有大量工厂

56
00:02:43,900 --> 00:02:46,060
还有游说集团

57
00:02:46,060 --> 00:02:47,680
必须确保所有人都保持就业

58
00:02:47,680 --> 00:02:50,020
然后就有了这些项目

59
00:02:50,020 --> 00:02:56,159
但现在我觉得太空领域正在多线展开真正的竞赛

60
00:02:56,159 --> 00:02:58,979
中国正在尝试在月球建立基地

61
00:02:58,979 --> 00:03:02,580
美国则在尝试建立人类在月球的存在

62
00:03:02,580 --> 00:03:03,588
在月球上

63
00:03:03,588 --> 00:03:05,329
还有轨道数据中心

64
00:03:05,329 --> 00:03:06,669
这是另一个发展方向

65
00:03:06,669 --> 00:03:10,868
还有几家主要参与者都在争夺这个目标

66
00:03:10,868 --> 00:03:15,560
因此我认为这是太空领域加速发展的时刻

67
00:03:15,560 --> 00:03:16,400
你知道的

68
00:03:16,400 --> 00:03:18,099
太空领域正在加速发展

69
00:03:18,099 --> 00:03:20,800
并且更加资本高效

70
00:03:20,800 --> 00:03:22,520
因为目标明确

71
00:03:22,520 --> 00:03:25,939
如果人们未能达标

72
00:03:25,939 --> 00:03:28,020
你就不再受

73
00:03:28,020 --> 00:03:29,099
只是

74
00:03:29,340 --> 00:03:30,659
这里只有一个路径

75
00:03:30,659 --> 00:03:34,219
你知道你需要扩展并进一步

76
00:03:34,219 --> 00:03:40,969
这个系统显然是为满足NASA设定的要求而设计的

77
00:03:40,969 --> 00:03:43,229
与国会合作

78
00:03:43,229 --> 00:03:46,189
或许是为了完成特定任务

79
00:03:46,189 --> 00:03:50,379
而SpaceX则设定了宏大的目标

80
00:03:50,379 --> 00:03:51,439
那就是我们要着陆

81
00:03:51,439 --> 00:03:54,900
我们知道要建立火星殖民地并持续推进

82
00:03:54,900 --> 00:03:55,649
北方

83
00:03:55,649 --> 00:03:57,049
不是星球的北极

84
00:03:57,049 --> 00:03:57,848
或许吧

85
00:03:57,848 --> 00:03:58,288
嗯

86
00:03:59,329 --> 00:04:03,389
在构建前往火星的技术体系过程中

87
00:04:09,400 --> 00:04:15,020
嗯，因此星链的出现源于他们降低每次发射成本

88
00:04:15,020 --> 00:04:15,620
你知道的

89
00:04:15,620 --> 00:04:17,720
试图降低成本以实现火星任务

90
00:04:17,720 --> 00:04:23,259
如果存在脱离政府的商业化路径

91
00:04:23,259 --> 00:04:28,339
这能为公司提供更好的激励机制

92
00:04:28,339 --> 00:04:30,819
而非政府直接支付

93
00:04:30,819 --> 00:04:34,300
政府会设定特定任务的价格

94
00:04:34,300 --> 00:04:38,189
嗯，所以现在正如你所说

95
00:04:38,189 --> 00:04:42,310
在近地轨道卫星通信领域

96
00:04:42,310 --> 00:04:43,939
显然存在巨大的

97
00:04:43,939 --> 00:04:44,519
你知道的

98
00:04:44,519 --> 00:04:48,860
超过百亿美元的商业市场，SpaceX正在争夺

99
00:04:48,860 --> 00:04:50,040
其他公司也在竞逐

100
00:04:50,040 --> 00:04:52,139
然后通过轨道计算

101
00:04:52,139 --> 00:04:53,230
这里有一个

102
00:04:53,230 --> 00:04:53,970
你知道的

103
00:04:53,970 --> 00:04:55,470
至少数量级

104
00:04:55,470 --> 00:05:01,870
也许还有更多市场，SpaceX和其他公司正在争夺的运载工具领域

105
00:05:01,870 --> 00:05:04,410
他们为抓住这些机会所做的所有工程工作

106
00:05:08,769 --> 00:05:10,410
但NASA可以指出，比如

107
00:05:10,410 --> 00:05:12,810
我们想要建立月球基地

108
00:05:12,810 --> 00:05:15,920
或者我们想要建立火星殖民地

109
00:05:15,920 --> 00:05:18,079
这就是我认为我们所处的位置

110
00:05:18,079 --> 00:05:18,800
就这样结束了

111
00:05:18,800 --> 00:05:23,029
这是一个迹象，至少有正面报道

112
00:05:23,029 --> 00:05:24,389
美国人仍然被

113
00:05:24,389 --> 00:05:26,769
我们正在向月球背面派送人员的事实所激励

114
00:05:26,769 --> 00:05:31,610
这意味着政府可能为这类行动提供资金

115
00:05:31,610 --> 00:05:35,220
他们距离我们如此遥远真是令人惊叹

116
00:05:36,779 --> 00:05:41,908
据报道SpaceX正在秘密提交IPO申请

117
00:05:41,908 --> 00:05:46,500
你认为文件中最具突破性的是什么

118
00:05:46,500 --> 00:05:47,300
你知道的

119
00:05:47,300 --> 00:05:49,759
人们会在文件中关注什么

120
00:05:49,759 --> 00:05:53,589
然后在文件公布前

121
00:05:53,589 --> 00:05:57,310
我看到最流行的推文是

122
00:05:57,310 --> 00:05:57,870
你知道的

123
00:05:57,870 --> 00:05:58,509
人们只是说

124
00:05:58,509 --> 00:05:59,050
哦看

125
00:05:59,050 --> 00:06:00,470
亚马逊的营收是这个数字

126
00:06:00,470 --> 00:06:02,439
他们的市值是这个

127
00:06:02,439 --> 00:06:04,220
SpaceX的营收是这个

128
00:06:04,220 --> 00:06:08,430
他们将以相似估值上市

129
00:06:08,430 --> 00:06:08,889
嗯

130
00:06:08,889 --> 00:06:10,750
我觉得这种方式

131
00:06:10,750 --> 00:06:11,629
我这样想

132
00:06:11,629 --> 00:06:12,410
就像你知道的

133
00:06:12,410 --> 00:06:13,370
在私募市场

134
00:06:13,370 --> 00:06:13,790
就像

135
00:06:13,790 --> 00:06:14,509
你的估值是多少

136
00:06:14,509 --> 00:06:15,949
一旦上市后

137
00:06:15,949 --> 00:06:16,730
或许没错

138
00:06:16,730 --> 00:06:18,600
我最近看到

139
00:06:18,600 --> 00:06:20,819
我想从你的视角理解

140
00:06:20,819 --> 00:06:23,879
如何可能合理

141
00:06:23,879 --> 00:06:27,399
预计年营收约二十亿美元

142
00:06:27,399 --> 00:06:31,379
或在2025财年已公布

143
00:06:31,379 --> 00:06:33,680
他们以25%的速度增长

144
00:06:33,680 --> 00:06:36,860
显然他们收购了xAI

145
00:06:36,860 --> 00:06:38,149
这

146
00:06:38,350 --> 00:06:39,069
我不确定

147
00:06:39,069 --> 00:06:43,850
SpaceX股东将承受多大程度的稀释

148
00:06:43,850 --> 00:06:48,579
他们相对于其他聊天机器人已大幅落后

149
00:06:48,579 --> 00:06:52,579
嗯，然后它们的年增长率只有20%到25%

150
00:06:52,579 --> 00:06:54,139
或者至少有报道称如此

151
00:06:54,139 --> 00:06:56,259
因此估值可能达到一点

152
00:06:56,259 --> 00:06:56,620
七

153
00:06:56,620 --> 00:06:59,949
五万亿或一到两万亿

154
00:06:59,949 --> 00:07:01,110
无论你了解的

155
00:07:01,110 --> 00:07:05,468
我认为估值由当天决定

156
00:07:05,468 --> 00:07:05,968
我不

157
00:07:05,968 --> 00:07:06,749
我不知道如何

158
00:07:06,749 --> 00:07:08,689
我不理解

159
00:07:09,329 --> 00:07:12,250
我喜欢一个解释说明如何得出这个结论

160
00:07:12,250 --> 00:07:14,250
数学上当然没错

161
00:07:15,250 --> 00:07:15,449
是的

162
00:07:15,449 --> 00:07:15,610
是的

163
00:07:15,610 --> 00:07:16,029
是的

164
00:07:16,029 --> 00:07:23,670
现在SpaceX的增长率就像

165
00:07:23,670 --> 00:07:25,529
不能只看收入增长就说

166
00:07:25,529 --> 00:07:25,709
哦

167
00:07:25,709 --> 00:07:27,850
这就是其增长速度

168
00:07:27,850 --> 00:07:31,009
这取决于有效载荷量

169
00:07:31,009 --> 00:07:33,279
他们可以有效发射

170
00:07:33,279 --> 00:07:35,759
即忽略其他因素

171
00:07:35,759 --> 00:07:37,139
暂且计算一下

172
00:07:37,139 --> 00:07:46,129
目前他们能以极高的投资回报率部署星链卫星

173
00:07:47,129 --> 00:07:52,100
因为他们在上传带宽，人们会以多种方式使用

174
00:07:52,100 --> 00:07:53,740
你在飞机上使用

175
00:07:53,740 --> 00:07:55,509
你在房屋中使用

176
00:07:56,509 --> 00:07:59,589
你将用它实现直接手机覆盖

177
00:07:59,589 --> 00:08:03,480
这是一个极佳的投资回报机会

178
00:08:03,480 --> 00:08:05,160
投资资本回报机会

179
00:08:06,279 --> 00:08:11,240
但存在限制，他们能

180
00:08:11,240 --> 00:08:17,730
在利润范围内发射直到触及弹性曲线

181
00:08:17,730 --> 00:08:19,290
嗯，所以在我们的模型中

182
00:08:19,290 --> 00:08:20,250
萨姆，模型显示收入在百至二百亿美元之间

183
00:08:20,250 --> 00:08:23,990
他们可以从中获利

184
00:08:23,990 --> 00:08:26,029
这将极其有利可图

185
00:08:26,029 --> 00:08:28,788
对他们来说

186
00:08:28,788 --> 00:08:29,228
嗯

187
00:08:29,228 --> 00:08:32,729
我们认为模型显示极其有利可图

188
00:08:32,729 --> 00:08:34,129
因为他们

189
00:08:34,129 --> 00:08:34,469
嗯

190
00:08:35,589 --> 00:08:36,028
他们

191
00:08:36,028 --> 00:08:36,849
呃

192
00:08:36,849 --> 00:08:41,558
他们是唯一能处理有效载荷的公司

193
00:08:41,558 --> 00:08:44,019
一旦转向星舰

194
00:08:44,019 --> 00:08:45,219
呃

195
00:08:45,219 --> 00:08:47,678
假设星舰可重复使用且为助推器

196
00:08:47,678 --> 00:08:51,360
发射成本下降一个数量级

197
00:08:51,360 --> 00:08:54,100
因此更加有利可图

198
00:08:54,100 --> 00:08:55,879
他们可以采取的一系列行动

199
00:08:55,879 --> 00:08:58,009
而你的收入会加速增长

200
00:08:58,009 --> 00:08:58,529
就像

201
00:08:58,529 --> 00:08:58,970
是的

202
00:08:58,970 --> 00:08:59,629
你的收入推动增长

203
00:08:59,629 --> 00:09:00,710
你的增长正在加速

204
00:09:00,710 --> 00:09:02,029
大幅加速增长

205
00:09:02,029 --> 00:09:05,649
因为一枚火箭可以将百吨载荷送入轨道

206
00:09:05,649 --> 00:09:10,379
你知道他们现在只有约二十枚猎鹰九号

207
00:09:10,379 --> 00:09:12,259
他们正在设计数百枚

208
00:09:12,259 --> 00:09:15,769
未来甚至可能数千枚星舰堆栈

209
00:09:15,769 --> 00:09:19,059
从资本市场的角度来看

210
00:09:19,059 --> 00:09:21,100
我们要向某人提供资本

211
00:09:21,100 --> 00:09:25,078
然后他们能部署这笔资本并获得巨大回报

212
00:09:25,078 --> 00:09:27,818
这就是你应该看待SpaceX的方式

213
00:09:27,818 --> 00:09:29,639
他们选择上市的原因

214
00:09:29,639 --> 00:09:31,889
是因为可以自筹资金建设星链

215
00:09:31,889 --> 00:09:34,809
嗯，这是为了应对AI算力需求

216
00:09:34,809 --> 00:09:38,090
这其实是一个更大的机会

217
00:09:38,090 --> 00:09:41,789
你知道可能需要六十倍的更多卫星

218
00:09:41,789 --> 00:09:42,629
AI算力

219
00:09:42,629 --> 00:09:45,309
有多种方式可以计算这个需求

220
00:09:45,309 --> 00:09:46,570
他们已公开宣布

221
00:09:46,570 --> 00:09:46,830
嘿

222
00:09:46,830 --> 00:09:48,009
我们需要一百万

223
00:09:48,009 --> 00:09:51,750
这是为AI算力申请的一百万卫星

224
00:09:51,750 --> 00:09:56,590
而他们只为星链申请了约四万颗

225
00:09:56,590 --> 00:10:02,149
这意味着卫星数量是二十倍之多

226
00:10:02,149 --> 00:10:03,234
嗯

227
00:10:04,009 --> 00:10:07,250
因此他们会筹集资本

228
00:10:07,250 --> 00:10:09,879
并将其投入这个机会中

229
00:10:09,879 --> 00:10:10,919
作为股东

230
00:10:10,919 --> 00:10:12,240
你应该考虑

231
00:10:12,240 --> 00:10:14,799
鉴于已投入的资本

232
00:10:14,799 --> 00:10:16,389
我能获得什么回报

233
00:10:16,389 --> 00:10:20,970
这并不直接关联当前资本的收入

234
00:10:20,970 --> 00:10:22,269
他们目前拥有的资本

235
00:10:22,269 --> 00:10:24,730
事实上两者可能完全独立

236
00:10:24,730 --> 00:10:26,090
你知道的

237
00:10:26,090 --> 00:10:28,809
我拥有发行货币的许可

238
00:10:28,809 --> 00:10:31,220
我只需要向你索取资金

239
00:10:31,220 --> 00:10:32,500
然后它会产生

240
00:10:32,500 --> 00:10:32,919
你知道的

241
00:10:32,919 --> 00:10:34,360
无论回报如何

242
00:10:34,360 --> 00:10:36,820
我认为这就是看待这家公司的方式

243
00:10:36,820 --> 00:10:38,328
所以说

244
00:10:38,328 --> 00:10:39,068
嘿

245
00:10:39,068 --> 00:10:41,610
现在就像

246
00:10:41,610 --> 00:10:43,330
我不反对你的观点

247
00:10:43,330 --> 00:10:43,529
尼克

248
00:10:43,529 --> 00:10:45,830
我认为市场参与者会看到这并想

249
00:10:45,830 --> 00:10:46,769
这太疯狂了

250
00:10:46,769 --> 00:10:47,990
是x倍

251
00:10:47,990 --> 00:10:49,779
无论当前收入是多少

252
00:10:49,779 --> 00:10:55,230
而这实际上是一个基于熟人之间叙事价值的范围

253
00:10:55,230 --> 00:10:58,029
相信企业应仅基于现金流运营

254
00:10:58,029 --> 00:10:59,370
他们今天产生的现金流

255
00:10:59,370 --> 00:11:01,600
与应基于未来现金流的企业形成对比

256
00:11:01,600 --> 00:11:07,809
应基于将要交付给股东的现金流进行估值

257
00:11:07,809 --> 00:11:09,029
已有效率地投入资本

258
00:11:09,029 --> 00:11:09,350
我想

259
00:11:09,350 --> 00:11:09,789
是的

260
00:11:09,789 --> 00:11:10,190
对我来说

261
00:11:10,190 --> 00:11:13,529
更多是相对估值基础

262
00:11:13,529 --> 00:11:16,029
当你开始分析市销率倍数时

263
00:11:16,029 --> 00:11:18,470
如果说这个估值达到两万亿美元

264
00:11:18,470 --> 00:11:19,750
他已公开否认

265
00:11:19,750 --> 00:11:23,389
但对我们来说这是一个更易说的四舍五入数字

266
00:11:23,389 --> 00:11:23,669
好的

267
00:11:23,669 --> 00:11:25,578
市盈率高达一百倍

268
00:11:25,578 --> 00:11:28,318
销售额增长2.5%

269
00:11:28,318 --> 00:11:31,418
相比之下看meta的估值为一

270
00:11:31,418 --> 00:11:33,379
或估值为一点

271
00:11:34,409 --> 00:11:36,129
实现两百亿美元营收

272
00:11:36,129 --> 00:11:37,830
增长3.3%

273
00:11:37,830 --> 00:11:40,750
这些市值在六千亿级别

274
00:11:40,750 --> 00:11:44,090
七千亿级别的公司拥有印钞权

275
00:11:44,090 --> 00:11:47,759
并展现出显著的ROIC

276
00:11:47,759 --> 00:11:51,950
尤其在AI机遇面前

277
00:11:51,950 --> 00:11:52,870
不不

278
00:11:52,870 --> 00:11:53,409
但它们

279
00:11:53,409 --> 00:11:53,870
它们确实

280
00:11:53,870 --> 00:11:55,730
它们拥有印钞权

281
00:11:55,730 --> 00:11:55,970
是的

282
00:11:55,970 --> 00:11:56,210
它们

283
00:11:56,210 --> 00:11:57,850
正在印钞

284
00:11:57,850 --> 00:12:01,100
但边际上在回购股份

285
00:12:01,100 --> 00:12:02,279
发放股息

286
00:12:02,279 --> 00:12:04,679
它们没有闲置资金

287
00:12:04,679 --> 00:12:05,179
哦不

288
00:12:05,179 --> 00:12:08,399
或许四五年前提不出这个观点

289
00:12:08,399 --> 00:12:10,200
但现在借助AI确实成立

290
00:12:10,200 --> 00:12:11,129
就像它们已经

291
00:12:11,129 --> 00:12:15,549
将数据整合并转化为巨大AI叙事

292
00:12:15,549 --> 00:12:18,129
至少谷歌和大多数公司

293
00:12:18,129 --> 00:12:21,559
亚马逊也开始从数据中心布局

294
00:12:21,559 --> 00:12:24,490
显然Meta需要内部化这一点

295
00:12:24,490 --> 00:12:26,149
转化为广告利润

296
00:12:26,149 --> 00:12:29,089
或许构建自身AI能力

297
00:12:29,089 --> 00:12:35,269
但它们都具备获取巨大机遇的相似结构

298
00:12:35,269 --> 00:12:36,208
所以对我来说

299
00:12:36,208 --> 00:12:37,869
这只是相对而言

300
00:12:37,869 --> 00:12:39,489
比如你说嘿

301
00:12:39,489 --> 00:12:41,950
你可以投入一百美元到市场

302
00:12:41,950 --> 00:12:45,250
我不知道为什么要这样砸进去

303
00:12:45,250 --> 00:12:47,370
交易一百次像

304
00:12:47,690 --> 00:12:48,049
是啊

305
00:12:48,049 --> 00:12:48,370
我知道

306
00:12:48,370 --> 00:12:49,809
我不确定

307
00:12:50,340 --> 00:12:51,559
我认为这是个好公司

308
00:12:51,559 --> 00:12:52,059
顺便说一下

309
00:12:52,059 --> 00:12:53,340
这并不是一个

310
00:12:53,340 --> 00:12:54,019
这不是一个

311
00:12:54,019 --> 00:12:54,179
哦

312
00:12:54,179 --> 00:12:55,080
我不喜欢所有东西

313
00:12:55,080 --> 00:12:55,500
你在说

314
00:12:55,500 --> 00:12:56,539
我完全同意

315
00:12:56,539 --> 00:12:57,700
我只是不明白

316
00:12:57,700 --> 00:12:58,620
为什么要理解那个

317
00:12:58,620 --> 00:12:59,299
你做吗

318
00:12:59,299 --> 00:12:59,820
你明白

319
00:12:59,820 --> 00:13:00,539
显著的

320
00:13:00,539 --> 00:13:01,879
我所强调的区别

321
00:13:01,879 --> 00:13:02,580
比如

322
00:13:02,580 --> 00:13:05,480
想象我卖给你一个嗯

323
00:13:05,480 --> 00:13:06,419
一家公司

324
00:13:06,419 --> 00:13:09,320
它像洗车店并产生

325
00:13:09,320 --> 00:13:10,860
每年一百美元左右

326
00:13:10,860 --> 00:13:11,659
你需要支付

327
00:13:11,659 --> 00:13:12,220
你就像

328
00:13:12,220 --> 00:13:14,120
它会持续每年产生一百美元

329
00:13:14,120 --> 00:13:14,820
这没问题

330
00:13:14,820 --> 00:13:17,409
相比之下我说嘿

331
00:13:17,409 --> 00:13:19,909
我有这种新的洗车技术

332
00:13:19,909 --> 00:13:22,470
嗯我只需要建造洗车店

333
00:13:22,470 --> 00:13:24,960
然后它们每年能赚四百美元

334
00:13:24,960 --> 00:13:26,879
但需要前期资本投入

335
00:13:26,879 --> 00:13:27,559
但就

336
00:13:27,559 --> 00:13:29,860
但最终我需要支付市场

337
00:13:29,860 --> 00:13:30,860
在这里支付

338
00:13:30,860 --> 00:13:34,159
为每年四百美元支付一千六百美元

339
00:13:34,159 --> 00:13:35,919
那情况就不同了

340
00:13:35,919 --> 00:13:36,279
对吧

341
00:13:36,279 --> 00:13:37,059
就像这样

342
00:13:37,059 --> 00:13:37,500
你知道的

343
00:13:37,500 --> 00:13:40,578
这里的关键在于最终结果

344
00:13:40,578 --> 00:13:41,619
发生巨大变化

345
00:13:41,619 --> 00:13:42,859
长期机会

346
00:13:42,859 --> 00:13:45,099
因为你提前定价

347
00:13:45,099 --> 00:13:48,278
在IPA或未来几年大部分

348
00:13:48,278 --> 00:13:48,739
如果你是

349
00:13:48,739 --> 00:13:49,549
如果你是

350
00:13:49,549 --> 00:13:50,230
你必须

351
00:13:50,230 --> 00:13:52,230
你必须真正成长并进行评估

352
00:13:52,230 --> 00:13:53,830
你并不

353
00:13:55,899 --> 00:13:56,259
呃

354
00:13:56,259 --> 00:14:00,360
我认为企业能够增长的速度将会

355
00:14:00,360 --> 00:14:04,610
我认为这将是人们的大惊喜

356
00:14:04,610 --> 00:14:06,850
选择站在场外的人

357
00:14:06,850 --> 00:14:10,529
就像看英伟达三年的营收增长

358
00:14:10,529 --> 00:14:11,970
你会觉得这不可能

359
00:14:13,090 --> 00:14:13,350
没错

360
00:14:13,350 --> 00:14:13,929
就像

361
00:14:13,929 --> 00:14:15,230
这是因为需求存在

362
00:14:15,230 --> 00:14:17,909
人们甚至不知道这些芯片有需求

363
00:14:17,909 --> 00:14:22,500
然后他们开始建设并扩展了这项业务

364
00:14:22,500 --> 00:14:22,940
就像

365
00:14:22,940 --> 00:14:25,450
我们知道这方面有需求

366
00:14:25,450 --> 00:14:26,769
或者我们相当有信心

367
00:14:26,769 --> 00:14:31,500
这似乎很明显，因为人们正通过其他方式尝试获取

368
00:14:31,500 --> 00:14:32,580
但效率较低

369
00:14:32,580 --> 00:14:34,059
这就是连接性

370
00:14:34,059 --> 00:14:38,679
因此你可以抓住已存在的机会

371
00:14:39,059 --> 00:14:40,740
视频交易估值约20倍

372
00:14:40,740 --> 00:14:41,899
销售额

373
00:14:45,120 --> 00:14:45,679
不

374
00:14:45,679 --> 00:14:46,399
是的

375
00:14:46,399 --> 00:14:49,340
而我认为再次

376
00:14:49,340 --> 00:14:50,839
这是一个问题

377
00:14:50,839 --> 00:14:51,479
嗯

378
00:14:51,479 --> 00:14:51,979
你知道的

379
00:14:51,979 --> 00:14:56,240
市场仿佛倒转时钟回到

380
00:14:56,240 --> 00:14:57,059
某种程度上

381
00:14:57,059 --> 00:14:57,799
呃

382
00:14:57,799 --> 00:14:59,639
或许二十世纪初

383
00:14:59,639 --> 00:15:04,120
股票曾经仅按分红定价

384
00:15:04,120 --> 00:15:05,289
就是如此

385
00:15:05,289 --> 00:15:06,590
基本上就像

386
00:15:06,590 --> 00:15:07,549
股息是多少

387
00:15:07,549 --> 00:15:09,210
这就是我要支付的

388
00:15:09,210 --> 00:15:14,490
为什么要去投资无法产生现金流回报的业务

389
00:15:14,490 --> 00:15:16,009
然后他们意识到

390
00:15:16,009 --> 00:15:20,029
其实有些企业不应发放股息

391
00:15:20,029 --> 00:15:22,990
因为应该投资未来基础设施

392
00:15:22,990 --> 00:15:26,710
以提升整体业务的净现值

393
00:15:26,710 --> 00:15:29,750
企业估值方式

394
00:15:29,750 --> 00:15:31,269
随着时间变化

395
00:15:31,269 --> 00:15:32,308
过去

396
00:15:32,308 --> 00:15:35,109
过去是按账面价值定价

397
00:15:35,109 --> 00:15:36,489
仅账面价值

398
00:15:36,489 --> 00:15:41,068
而类似于账面价值可能会产生一定的股息收益率

399
00:15:41,068 --> 00:15:43,349
长期以来价值投资者就像

400
00:15:43,349 --> 00:15:45,109
为什么要支付高于账面价值的价格

401
00:15:45,109 --> 00:15:46,168
这简直疯狂

402
00:15:46,168 --> 00:15:48,559
为什么要支付高于账面价值的价格

403
00:15:48,559 --> 00:15:49,960
结果发现

404
00:15:49,960 --> 00:15:50,500
你知道的

405
00:15:50,500 --> 00:15:53,120
账面价值忽略了大量无形资产

406
00:15:53,120 --> 00:15:54,610
这存在各种问题

407
00:15:54,610 --> 00:15:57,250
以SpaceX为例

408
00:15:57,250 --> 00:16:00,538
这将取决于嗯

409
00:16:00,538 --> 00:16:02,318
他们能否制造可重复使用的火箭

410
00:16:02,318 --> 00:16:06,519
如果成功回报率是多少

411
00:16:06,519 --> 00:16:10,379
他们将在该业务中获得的投资资本回报

412
00:16:10,379 --> 00:16:12,840
他们能持续多久

413
00:16:12,840 --> 00:16:18,200
基本上在这个循环中不断用现金投资更多卫星

414
00:16:18,200 --> 00:16:20,480
不断用现金投资更多卫星

415
00:16:20,480 --> 00:16:22,090
在有人赶上之前

416
00:16:22,090 --> 00:16:26,769
SpaceX与其他所有机会的区别在于

417
00:16:26,769 --> 00:16:31,100
在某个范围内没有竞争对手

418
00:16:31,100 --> 00:16:32,379
你应该一开始就强调这一点

419
00:16:32,379 --> 00:16:35,490
在他们爱物能力方面

420
00:16:35,490 --> 00:16:37,389
能够成功发射

421
00:16:37,389 --> 00:16:41,029
我认为这是最重要的论点之一

422
00:16:41,029 --> 00:16:43,100
这非常具体

423
00:16:43,100 --> 00:16:48,990
世界上没有任何公司能在竞争中拥有像SpaceX这样的领先优势

424
00:16:48,990 --> 00:16:51,490
可以非常简洁地说

425
00:16:51,490 --> 00:16:53,830
他们拥有一枚火箭

426
00:16:53,830 --> 00:16:56,519
能够高效重复使用

427
00:16:56,519 --> 00:16:58,200
他们已经持续了

428
00:16:58,200 --> 00:16:59,460
过去十一年

429
00:16:59,460 --> 00:17:06,799
蓝色起源是首个成功回收轨道火箭的公司

430
00:17:06,799 --> 00:17:08,700
但他们尚未重复使用

431
00:17:08,700 --> 00:17:11,670
问题是

432
00:17:12,069 --> 00:17:15,450
改进的速度如何

433
00:17:15,450 --> 00:17:17,450
我认为可以这样说

434
00:17:17,450 --> 00:17:20,690
SpaceX仍在加速发展

435
00:17:20,690 --> 00:17:23,710
他们的领先优势正在扩大

436
00:17:23,710 --> 00:17:25,589
我认为这就是答案

437
00:17:25,589 --> 00:17:28,089
因为很多人会说和我一样的观点

438
00:17:28,089 --> 00:17:28,910
哇

439
00:17:28,910 --> 00:17:31,459
估值倍数非常疯狂

440
00:17:31,459 --> 00:17:33,419
我不知道如何理解

441
00:17:33,419 --> 00:17:34,829
答案是

442
00:17:34,829 --> 00:17:36,869
没有人能做他们正在做的事

443
00:17:36,869 --> 00:17:37,869
或者可以

444
00:17:37,869 --> 00:17:40,349
但需要十年才能达到规模

445
00:17:40,349 --> 00:17:42,190
这给了他们巨大发展空间

446
00:17:42,190 --> 00:17:44,430
太空领域的机遇无限

447
00:17:44,430 --> 00:17:45,839
因为太空是无限的

448
00:17:45,839 --> 00:17:46,319
嗯

449
00:17:46,319 --> 00:17:48,920
所以我们可以去采矿小行星和黄金

450
00:17:48,920 --> 00:17:52,319
然后前往钻石星球做各种有趣的事情

451
00:17:52,319 --> 00:17:54,309
这才是真正的机会

452
00:17:54,309 --> 00:17:55,470
这样说当然没问题

453
00:17:55,470 --> 00:17:58,730
因为我认为你可以获得比他人更高的估值倍数

454
00:17:58,730 --> 00:18:02,230
如果你拥有显著的竞争护城河

455
00:18:02,230 --> 00:18:05,138
你知道你展现了高ROIC

456
00:18:05,138 --> 00:18:07,739
或能持续在未来证明这一点

457
00:18:07,739 --> 00:18:10,849
我觉得这完全是一个合理的观点

458
00:18:10,849 --> 00:18:12,769
这是对的反驳观点

459
00:18:12,769 --> 00:18:16,339
这个估值简直离谱

460
00:18:16,339 --> 00:18:17,920
因为很多人会这么说

461
00:18:17,920 --> 00:18:20,460
一旦玩家也参与进来

462
00:18:20,460 --> 00:18:21,940
一旦S1价格下跌

463
00:18:21,940 --> 00:18:23,279
人们会看到

464
00:18:23,279 --> 00:18:23,660
你知道的

465
00:18:23,660 --> 00:18:24,059
现在

466
00:18:24,059 --> 00:18:25,250
一切都是传言

467
00:18:25,250 --> 00:18:27,049
但一旦人们看到

468
00:18:27,049 --> 00:18:29,089
这将是首要论点

469
00:18:31,180 --> 00:18:35,849
我也认为资本成本的问题

470
00:18:35,849 --> 00:18:38,349
这一点被低估了

471
00:18:38,349 --> 00:18:39,130
也就是说

472
00:18:39,130 --> 00:18:39,670
假设

473
00:18:39,670 --> 00:18:47,609
有一家公司能像SpaceX一样高效执行

474
00:18:47,609 --> 00:18:49,630
但他们做不到

475
00:18:49,630 --> 00:18:50,638
呃

476
00:18:50,638 --> 00:18:57,400
SpaceX的资本成本及其投资周期内的成本

477
00:18:57,400 --> 00:18:57,920
要达到

478
00:18:57,920 --> 00:19:03,528
如今的水平，其成本远低于其他公司

479
00:19:03,528 --> 00:19:08,690
即使假设有一家公司能像SpaceX一样优秀

480
00:19:08,690 --> 00:19:10,509
并设定其发展路径

481
00:19:10,509 --> 00:19:13,159
如果他们需要融资

482
00:19:13,358 --> 00:19:14,798
将不具备竞争力

483
00:19:14,798 --> 00:19:17,670
你对此怎么看

484
00:19:18,670 --> 00:19:19,349
你不这么认为

485
00:19:19,349 --> 00:19:20,309
布雷特

486
00:19:20,750 --> 00:19:21,109
嗯

487
00:19:21,109 --> 00:19:21,829
我我我

488
00:19:21,829 --> 00:19:23,750
这只是个令人不满的论点

489
00:19:23,750 --> 00:19:24,430
因为毕竟

490
00:19:24,430 --> 00:19:24,730
这确实

491
00:19:24,730 --> 00:19:25,309
但这是事实

492
00:19:25,309 --> 00:19:25,630
就像

493
00:19:25,630 --> 00:19:26,880
就像这里这里

494
00:19:26,954 --> 00:19:27,700
是的

495
00:19:27,700 --> 00:19:30,900
能够筹集到计划中的估值所需资金

496
00:19:30,900 --> 00:19:32,440
他们计划以该估值筹集资金

497
00:19:32,440 --> 00:19:33,000
你知道的

498
00:19:33,000 --> 00:19:34,779
股东权益相关措施

499
00:19:34,779 --> 00:19:38,380
你们将获得大量可部署的资本

500
00:19:38,380 --> 00:19:39,170
这并非

501
00:19:39,170 --> 00:19:42,289
从相对角度来看成本很高

502
00:19:42,289 --> 00:19:45,690
嗯，我认为那个

503
00:19:45,690 --> 00:19:46,509
尼克的

504
00:19:46,509 --> 00:19:47,230
你知道的

505
00:19:47,230 --> 00:19:47,890
尼克

506
00:19:47,890 --> 00:19:49,130
我们希望你能加入我们这边

507
00:19:49,130 --> 00:19:50,650
我暂时换另一边说

508
00:19:50,650 --> 00:19:51,990
这属于

509
00:19:51,990 --> 00:19:54,839
还有其他类型的

510
00:19:54,839 --> 00:19:55,680
不仅仅是

511
00:19:55,680 --> 00:19:58,200
公司每营收的价值

512
00:19:58,200 --> 00:19:59,220
随着上市流通

513
00:19:59,220 --> 00:20:00,420
但你知道的

514
00:20:00,420 --> 00:20:06,400
他们能否将计算资源优势转化为实际变现优势

515
00:20:06,400 --> 00:20:08,599
在AI领域仍是开放性问题

516
00:20:08,599 --> 00:20:10,929
尚未得到验证

517
00:20:11,048 --> 00:20:14,848
那么需求曲线将持续多久

518
00:20:14,848 --> 00:20:16,048
其实并非无限

519
00:20:16,048 --> 00:20:17,189
或许确实是

520
00:20:17,189 --> 00:20:20,220
但我们认为AI计算需求

521
00:20:20,220 --> 00:20:24,849
其增长将持续超出合理预期

522
00:20:24,849 --> 00:20:27,329
但尚未证实

523
00:20:27,329 --> 00:20:30,829
因此这些问题即使由

524
00:20:30,829 --> 00:20:33,449
擅长资本部署的人

525
00:20:33,449 --> 00:20:34,568
也会面临挑战

526
00:20:34,568 --> 00:20:36,409
关键在于回报率如何

527
00:20:36,409 --> 00:20:38,519
看起来像一条退出路径

528
00:20:38,519 --> 00:20:38,640
是

529
00:20:38,640 --> 00:20:41,940
我认为他们可以将AI计算能力出售给他人

530
00:20:41,940 --> 00:20:44,099
而非全部自用

531
00:20:44,099 --> 00:20:47,089
如果AI变现效果不佳

532
00:20:47,089 --> 00:20:49,589
则形成基础设施即服务业务

533
00:20:49,589 --> 00:20:50,829
这是优质商业模式

534
00:20:50,829 --> 00:20:53,369
且具备持续扩展能力

535
00:20:53,369 --> 00:20:57,779
另一个待解问题

536
00:20:57,779 --> 00:20:59,579
仍需进一步验证

537
00:20:59,579 --> 00:21:02,750
当前计算需求有多大

538
00:21:02,750 --> 00:21:05,470
目前我们明显面临算力短缺

539
00:21:05,470 --> 00:21:06,769
人们总会说

540
00:21:06,769 --> 00:21:09,670
我们即将进入过剩阶段

541
00:21:09,670 --> 00:21:10,690
这就是他们所做

542
00:21:10,690 --> 00:21:13,809
参照英伟达第三、四年增长数据

543
00:21:13,809 --> 00:21:14,650
这就是他们

544
00:21:14,650 --> 00:21:15,710
他们正在预测

545
00:21:15,710 --> 00:21:19,269
超出管理层预期的周期终点

546
00:21:19,269 --> 00:21:23,990
我认为这种状态将持续到实际周期结束

547
00:21:23,990 --> 00:21:28,509
但我觉得我们在头脑风暴中提到的

548
00:21:29,549 --> 00:21:32,670
你可以拥有无限的类无限需求

549
00:21:32,670 --> 00:21:38,789
但这并不意味着周期中没有凸性仍然存在对吧

550
00:21:38,789 --> 00:21:40,009
就像所有你

551
00:21:40,009 --> 00:21:45,519
可能有千倍甚至百万倍的需求增长

552
00:21:45,519 --> 00:21:53,160
但你需要的只是一个资金或疑虑的波动来减缓进程

553
00:21:53,160 --> 00:21:56,549
然后你会看到凸性向相反方向发展

554
00:21:56,549 --> 00:21:58,009
时间线被推迟

555
00:21:58,009 --> 00:21:59,759
因为人们不愿投资

556
00:21:59,759 --> 00:22:02,039
然后就像下一个模型不会发布了

557
00:22:02,039 --> 00:22:02,839
这将释放

558
00:22:02,839 --> 00:22:03,880
更多需求

559
00:22:03,880 --> 00:22:09,450
所以人们以为一切都很糟糕直到拐点出现

560
00:22:09,450 --> 00:22:10,630
然后你知道

561
00:22:10,630 --> 00:22:12,679
十五年后回头看

562
00:22:12,679 --> 00:22:13,078
你会说

563
00:22:13,078 --> 00:22:13,298
是的

564
00:22:13,298 --> 00:22:15,480
你看这就是无限的

565
00:22:16,079 --> 00:22:16,619
是的

566
00:22:16,619 --> 00:22:18,180
我认为会有快速的波动

567
00:22:18,180 --> 00:22:20,960
在我们剩余的几分钟里

568
00:22:20,960 --> 00:22:22,339
又一次大幅融资

569
00:22:22,339 --> 00:22:27,119
另一家大型公司面临诸多

570
00:22:28,400 --> 00:22:30,160
待解问题

571
00:22:30,160 --> 00:22:32,799
人工智能融资一百二十亿美元

572
00:22:32,799 --> 00:22:35,259
远超七十亿七十五亿美元

573
00:22:35,259 --> 00:22:38,769
为什么还要去公开市场融资七十亿

574
00:22:38,769 --> 00:22:41,430
当私募市场能拿到一百二十亿

575
00:22:41,430 --> 00:22:43,329
我敢说马斯克在后悔说

576
00:22:43,329 --> 00:22:44,829
我为什么要启动这个流程

577
00:22:44,829 --> 00:22:48,509
本可以直接在私募市场融同等金额

578
00:22:48,509 --> 00:22:50,289
不用处理所有这些麻烦

579
00:22:50,289 --> 00:22:52,630
布雷特你对公开

580
00:22:52,630 --> 00:22:54,569
人工智能融资一百二十亿

581
00:22:54,569 --> 00:22:55,470
拆解分析

582
00:22:55,470 --> 00:23:03,299
然后我们需要讨论所有巧合出现的新闻

583
00:23:03,299 --> 00:23:08,359
恰好在融资公告后围绕OpenAI

584
00:23:08,359 --> 00:23:10,440
背后的戏剧性事件

585
00:23:10,440 --> 00:23:12,200
管理层重组

586
00:23:12,200 --> 00:23:15,359
有很多内容要讨论

587
00:23:17,480 --> 00:23:21,940
我认为高管团队的噪音越大

588
00:23:21,940 --> 00:23:22,859
嗯

589
00:23:22,859 --> 00:23:28,170
我认为可以将这视为AI、OpenAI、Gemini和Anthropic

590
00:23:28,170 --> 00:23:31,059
都在竞争一个

591
00:23:31,059 --> 00:23:37,089
我认为这是十五到二十万亿美元规模的基础模型市场

592
00:23:37,089 --> 00:23:43,390
以企业价值为基础争夺的企业级AI软件市场

593
00:23:43,390 --> 00:23:47,440
基于到2030年

594
00:23:47,440 --> 00:23:48,200
嗯

595
00:23:48,200 --> 00:23:49,160
我们的中期预测

596
00:23:52,559 --> 00:23:56,119
企业为AI生产力价值付费

597
00:23:56,119 --> 00:23:58,759
以实际生产力大幅折扣

598
00:23:58,759 --> 00:24:02,839
这七万亿美元不会全部流向OpenAI

599
00:24:02,839 --> 00:24:04,480
或xAI

600
00:24:04,480 --> 00:24:08,359
或那些流入相关公司的

601
00:24:08,359 --> 00:24:08,839
你知道的

602
00:24:08,839 --> 00:24:13,539
仍会有特定的AI软件产品位于技术栈顶端

603
00:24:13,539 --> 00:24:17,140
并支付部分成本基数给基础模型提供商

604
00:24:17,140 --> 00:24:20,819
如Palantir等平台即服务提供商

605
00:24:20,819 --> 00:24:23,619
但在收入瀑布模型中

606
00:24:23,619 --> 00:24:27,019
显示约两万亿美元归集到这些

607
00:24:27,019 --> 00:24:30,589
基础模型公司如OpenAI

608
00:24:30,589 --> 00:24:31,150
嗯

609
00:24:31,150 --> 00:24:34,809
在我们预估的利润率结构下

610
00:24:34,809 --> 00:24:35,930
我们认为它们将实现

611
00:24:35,930 --> 00:24:39,049
合理的市盈率倍数

612
00:24:39,049 --> 00:24:45,219
预示基础模型提供商的企业价值约十五万亿美元

613
00:24:45,219 --> 00:24:48,700
但取决于市场给予的倍数

614
00:24:48,700 --> 00:24:51,200
所以十五到二十万亿美元，以Open

615
00:24:51,200 --> 00:24:54,440
AI市值八百亿美元，瞄准

616
00:24:54,720 --> 00:24:57,299
二十万亿美元的机会

617
00:24:57,299 --> 00:24:58,680
但它们并未拥有

618
00:24:58,680 --> 00:25:00,440
未必能完全获得

619
00:25:00,440 --> 00:25:02,089
这是一个竞争领域

620
00:25:02,089 --> 00:25:04,269
即使被分割为三部分

621
00:25:04,269 --> 00:25:06,569
那也是七万亿美元

622
00:25:06,569 --> 00:25:11,400
因此从宏观角度看回报合理

623
00:25:11,400 --> 00:25:13,700
还需审视其具体

624
00:25:13,700 --> 00:25:19,299
公开报告的管理层营收预期

625
00:25:19,299 --> 00:25:23,750
我们认为这些目标可达成

626
00:25:23,750 --> 00:25:25,769
鉴于其增长速度

627
00:25:25,769 --> 00:25:29,730
营收规模约两百多亿美元

628
00:25:29,730 --> 00:25:36,460
查看期刊等公开报告的预期营收

629
00:25:36,460 --> 00:25:38,920
据期刊报道达两百五十亿美元

630
00:25:38,920 --> 00:25:43,529
营收两百五十亿美元且快速增长

631
00:25:44,210 --> 00:25:46,690
认为这并不过分

632
00:25:46,690 --> 00:25:51,170
可能成为两万五千亿公司

633
00:25:51,170 --> 00:25:54,309
取决于利润率及市场预期

634
00:25:54,309 --> 00:25:55,930
在企业领域

635
00:25:55,930 --> 00:25:56,750
我认为这是

636
00:25:56,750 --> 00:25:57,789
你知道的

637
00:25:57,789 --> 00:26:03,210
定位极佳的企业，增长迅速并改变企业工作者的工作方式

638
00:26:03,210 --> 00:26:03,769
嗯

639
00:26:03,769 --> 00:26:06,250
我认为Anthropic是真实竞争对手

640
00:26:06,250 --> 00:26:07,109
如xAI

641
00:26:07,109 --> 00:26:08,289
如Gemini

642
00:26:08,289 --> 00:26:09,849
我认为并非全胜

643
00:26:09,849 --> 00:26:10,900
全部纳入

644
00:26:10,900 --> 00:26:12,500
如同基础设施即服务

645
00:26:12,500 --> 00:26:13,519
将出现

646
00:26:13,519 --> 00:26:16,279
少数竞争对手均能产生现金流

647
00:26:16,279 --> 00:26:17,500
这些都可能非常有价值

648
00:26:17,500 --> 00:26:19,599
这就是基础设施即服务

649
00:26:19,599 --> 00:26:24,680
你知道基础设施即服务现在已是一个数百亿美元的市场

650
00:26:24,680 --> 00:26:27,160
我们认为这将是一个万亿美元级别的市场

651
00:26:27,160 --> 00:26:29,109
为企业提供AI模型

652
00:26:29,109 --> 00:26:30,450
然后在消费者端

653
00:26:30,450 --> 00:26:31,990
你知道尼克

654
00:26:31,990 --> 00:26:33,130
这方面我得请教你

655
00:26:33,130 --> 00:26:40,230
但嗯OpenAI每周活跃用户约9000万，增长迅速

656
00:26:41,630 --> 00:26:49,220
整个领域可能在2030年前后达到45亿周活跃用户

657
00:26:49,220 --> 00:26:50,779
所以你有

658
00:26:50,779 --> 00:26:55,809
我们估计整个领域目前周活跃用户已超1亿

659
00:26:56,170 --> 00:26:57,529
实际可能达1.1亿

660
00:26:57,529 --> 00:26:58,130
1.2亿

661
00:26:58,130 --> 00:26:59,089
我猜啊

662
00:26:59,089 --> 00:27:01,289
基于此有四倍的增长动力

663
00:27:01,289 --> 00:27:03,049
人们会更频繁使用这些服务

664
00:27:03,049 --> 00:27:04,470
所以你知道

665
00:27:04,470 --> 00:27:04,890
我是说

666
00:27:06,049 --> 00:27:09,589
存在大量未开发的变现机会，可通过广告实现

667
00:27:09,589 --> 00:27:11,029
嗯

668
00:27:11,029 --> 00:27:13,190
类似采购代理的行为模式

669
00:27:13,190 --> 00:27:13,710
尼克

670
00:27:13,710 --> 00:27:18,079
你认为广告端还是代理端会率先突破

671
00:27:18,079 --> 00:27:18,640
是的

672
00:27:18,640 --> 00:27:19,420
我觉得

673
00:27:19,420 --> 00:27:20,359
嗯

674
00:27:20,359 --> 00:27:20,759
对啊

675
00:27:22,039 --> 00:27:26,119
两者似乎都在同步推进，挺有意思的

676
00:27:26,119 --> 00:27:27,599
我是说他们正在并行发展

677
00:27:27,599 --> 00:27:27,759
没错

678
00:27:27,759 --> 00:27:30,240
广告业务正在增长

679
00:27:30,240 --> 00:27:32,670
ARR已达1亿美元

680
00:27:32,670 --> 00:27:36,829
这个术语在广告界并不常用

681
00:27:36,829 --> 00:27:41,980
因为它反映的并非近期实际投放，自项目上线后

682
00:27:41,980 --> 00:27:44,079
实际可能只有800-1000万美元

683
00:27:44,079 --> 00:27:44,339
不过

684
00:27:44,339 --> 00:27:45,420
需要细分来看

685
00:27:45,420 --> 00:27:47,599
但我认为这是真实机会

686
00:27:47,599 --> 00:27:53,279
我们预计到2030年可能有6700亿美元的机会

687
00:27:53,279 --> 00:27:57,069
如果观察全球广告市场

688
00:27:57,069 --> 00:28:00,490
并假设AI在广告中的占比

689
00:28:00,490 --> 00:28:04,329
而我们定义的AI广告是工具内使用

690
00:28:04,329 --> 00:28:07,470
所以我们已部分计入现有流量

691
00:28:07,470 --> 00:28:11,180
包括Meta和OpenAI的所有操作

692
00:28:11,180 --> 00:28:13,400
在电商领域

693
00:28:13,400 --> 00:28:15,900
这一方向似乎更具挑战性

694
00:28:15,900 --> 00:28:18,279
尤其是观察OpenAI的发展历程

695
00:28:18,279 --> 00:28:22,180
不是说他们错了，也不是认为会失败

696
00:28:22,180 --> 00:28:25,269
我只是觉得他们正在证明

697
00:28:25,269 --> 00:28:28,599
你或人们正在使用代理来发现

698
00:28:28,599 --> 00:28:33,400
但人们尚未准备好将OpenAI作为完整的结账解决方案

699
00:28:33,400 --> 00:28:36,460
这在一定程度上可能是因为产品设计有误

700
00:28:36,460 --> 00:28:38,180
嗯，我认为他们能达到吗

701
00:28:38,180 --> 00:28:38,539
是的

702
00:28:38,539 --> 00:28:38,920
绝对正确

703
00:28:38,920 --> 00:28:41,388
如果你有九亿每周活跃用户

704
00:28:41,388 --> 00:28:41,828
嗯

705
00:28:41,828 --> 00:28:44,788
你可以获得巨额总收入

706
00:28:44,788 --> 00:28:52,099
我们预计到2030年全球将有约9万亿美元通过代理流动

707
00:28:52,099 --> 00:28:56,049
在线销售的促成者

708
00:28:56,049 --> 00:29:01,619
我们预计这将占全球商业和在线销售的2.5%

709
00:29:01,619 --> 00:29:04,240
我们确实认为这些机会非常真实

710
00:29:04,240 --> 00:29:06,799
但正如你指出的企业领域

711
00:29:06,799 --> 00:29:08,069
这并非

712
00:29:08,630 --> 00:29:09,509
你知道的

713
00:29:09,509 --> 00:29:10,750
开放领域

714
00:29:10,750 --> 00:29:11,470
绝非如此

715
00:29:11,470 --> 00:29:12,730
确实存在竞争对手

716
00:29:12,730 --> 00:29:15,869
我认为Meta可能在这里暗中布局

717
00:29:15,869 --> 00:29:17,630
还不清楚苹果在做什么

718
00:29:17,630 --> 00:29:18,789
我知道你会笑我

719
00:29:18,789 --> 00:29:20,769
但我仍相信

720
00:29:20,769 --> 00:29:22,769
一旦他们使用谷歌的模型

721
00:29:22,769 --> 00:29:24,888
会推出极具价值的产品

722
00:29:24,888 --> 00:29:26,068
我懂了，果然如此

723
00:29:26,068 --> 00:29:26,568
果然如此

724
00:29:26,568 --> 00:29:27,469
我能预知未来

725
00:29:27,469 --> 00:29:29,729
我知道Brett何时会发笑并否定我

726
00:29:29,729 --> 00:29:30,529
嗯

727
00:29:30,529 --> 00:29:34,309
而我认为Gemini可能是真正的竞争对手

728
00:29:34,309 --> 00:29:36,019
他们在消费者端

729
00:29:36,019 --> 00:29:38,519
正如我们在节目中讨论过

730
00:29:38,519 --> 00:29:42,459
因为他们可以长期补贴这一业务

731
00:29:42,459 --> 00:29:46,679
对任何参与者施加真实的价格压力

732
00:29:46,679 --> 00:29:49,439
我对补贴方面不太担心

733
00:29:49,439 --> 00:29:53,119
更关注谷歌是否有能力

734
00:29:53,119 --> 00:29:58,660
将所有操作与消费者产品更直接整合

735
00:29:58,660 --> 00:29:59,460
他们确实有

736
00:29:59,460 --> 00:30:03,720
我认为苹果的优势在于拥有大量消费者数据

737
00:30:03,720 --> 00:30:04,299
因此

738
00:30:04,299 --> 00:30:07,920
因此能够无缝匹配这些数据

739
00:30:07,920 --> 00:30:12,440
但可以看看微软在企业Copilot上的失误

740
00:30:12,440 --> 00:30:14,878
在企业Copilot方面

741
00:30:14,878 --> 00:30:15,598
可以说

742
00:30:15,598 --> 00:30:17,519
苹果显然在效仿

743
00:30:17,519 --> 00:30:19,680
实际上比微软更差的策略

744
00:30:19,680 --> 00:30:21,819
因此无需相信苹果

745
00:30:21,819 --> 00:30:25,220
能以吸引消费者的方式整合这些资源

746
00:30:25,220 --> 00:30:26,140
但我认为谷歌

747
00:30:27,740 --> 00:30:27,940
是的

748
00:30:27,940 --> 00:30:28,799
他们现金流充足

749
00:30:28,799 --> 00:30:29,888
但他们也有

750
00:30:29,888 --> 00:30:31,608
你知道他们可以做广告

751
00:30:31,608 --> 00:30:32,729
比如在Gmail里

752
00:30:32,729 --> 00:30:34,949
他们正试图把我推向使用Gemini

753
00:30:34,949 --> 00:30:37,608
以及Google Drive和Google Sheets

754
00:30:37,608 --> 00:30:41,189
还有他们所有面向消费者的产品

755
00:30:41,189 --> 00:30:44,190
他们不仅能推动人们使用

756
00:30:44,190 --> 00:30:47,430
还能无缝整合用户体验

757
00:30:47,430 --> 00:30:52,029
这似乎是苹果做不到的

758
00:30:52,029 --> 00:30:54,839
或者似乎还没意识到需要这样做

759
00:30:54,839 --> 00:30:58,240
嗯，我知道技术都具备

760
00:30:58,240 --> 00:31:00,359
只是存在各种限制

761
00:31:00,359 --> 00:31:00,960
请

762
00:31:00,960 --> 00:31:03,160
时间只过了几年

763
00:31:04,019 --> 00:31:08,680
请给我带摄像头的AirPods，能记住所有内容

764
00:31:08,680 --> 00:31:09,740
你一定会拥有

765
00:31:09,740 --> 00:31:11,980
嗯，好吧

766
00:31:11,980 --> 00:31:16,150
时间已经过去

767
00:31:16,150 --> 00:31:23,539
但有人提出了一系列关于Uber及其最新公告的问题

768
00:31:24,019 --> 00:31:26,339
关于自动驾驶方面

769
00:31:26,339 --> 00:31:32,289
并定位自己为自动驾驶平台

770
00:31:32,529 --> 00:31:33,369
布莱特

771
00:31:33,369 --> 00:31:37,049
你认为Uber会成为自动驾驶平台吗

772
00:31:37,049 --> 00:31:41,240
还是会被淘汰呢

773
00:31:41,240 --> 00:31:41,740
他们已经签约

774
00:31:41,740 --> 00:31:43,960
签署了五家不同合作伙伴

775
00:31:43,960 --> 00:31:45,059
显然他们在推进

776
00:31:45,059 --> 00:31:46,000
墙角涂墙

777
00:31:46,000 --> 00:31:49,579
策略是大约五家左右

778
00:31:49,579 --> 00:31:51,180
我不知道具体数量

779
00:31:51,180 --> 00:31:52,140
但数量很多

780
00:31:52,140 --> 00:31:54,380
如果你无法将自动驾驶车辆

781
00:31:54,380 --> 00:32:00,480
交付给消费者使用

782
00:32:00,480 --> 00:32:01,799
使得它们能够

783
00:32:01,799 --> 00:32:04,279
低利用率

784
00:32:04,279 --> 00:32:06,279
灵活融入网络

785
00:32:06,279 --> 00:32:10,309
Uber在其中扮演重要角色

786
00:32:10,309 --> 00:32:11,930
就像Waymo一样

787
00:32:11,930 --> 00:32:13,829
Waymo的问题在于

788
00:32:13,829 --> 00:32:15,650
要么必须大幅

789
00:32:15,650 --> 00:32:17,950
超前部署网络以应对高峰需求

790
00:32:17,950 --> 00:32:22,650
否则无法依赖Waymo

791
00:32:22,650 --> 00:32:26,159
高峰期可能两小时都叫不到

792
00:32:26,159 --> 00:32:26,759
嗯

793
00:32:26,759 --> 00:32:30,159
无法通过价格信号激励更多供给

794
00:32:30,159 --> 00:32:35,029
这意味着服务注定不完整

795
00:32:35,029 --> 00:32:36,609
我推测

796
00:32:36,609 --> 00:32:41,440
虽然未经证实，但人们可能默认选择完整服务

797
00:32:41,440 --> 00:32:43,579
然后用另一个作为备用

798
00:32:43,579 --> 00:32:43,740
哦

799
00:32:43,740 --> 00:32:45,200
我要检查这个是否可用

800
00:32:45,200 --> 00:32:49,349
这就是为什么Waymo必须与Uber签约

801
00:32:49,349 --> 00:32:53,329
我认为亚特兰大是因为这样消费者可以去说

802
00:32:53,329 --> 00:32:54,529
我马上叫车

803
00:32:54,529 --> 00:32:59,000
然后Uber会根据方便性调用Waymo车辆

804
00:32:59,000 --> 00:33:01,329
但无论如何都会有车可用

805
00:33:01,329 --> 00:33:04,529
唯一可信的替代方案是特斯拉的策略

806
00:33:04,529 --> 00:33:08,069
即拥有消费者自有的车辆在路面上

807
00:33:08,069 --> 00:33:12,759
我在SD车上驾驶并让它载我前行

808
00:33:12,759 --> 00:33:15,140
这样我就享受专车服务

809
00:33:15,140 --> 00:33:19,630
未来某个时候会在我手机上弹出提示

810
00:33:19,630 --> 00:33:19,950
嘿

811
00:33:19,950 --> 00:33:23,869
你是否愿意让车辆现在为你完成价值三百美元的工作

812
00:33:23,869 --> 00:33:24,710
你会答应

813
00:33:24,710 --> 00:33:25,730
我接受三百美元

814
00:33:25,730 --> 00:33:26,480
谢谢

815
00:33:26,480 --> 00:33:29,839
这就能解决网络不完整的问题

816
00:33:29,839 --> 00:33:31,319
我同意

817
00:33:32,400 --> 00:33:35,460
你刚才描述的难道不可行吗

818
00:33:35,460 --> 00:33:36,210
不过

819
00:33:36,210 --> 00:33:37,769
所以好吧

820
00:33:37,769 --> 00:33:40,369
假设你拥有Waymo车队

821
00:33:40,369 --> 00:33:41,809
为什么不呢

822
00:33:41,809 --> 00:33:48,089
让个人付费使用这些车进行私人行程

823
00:33:48,089 --> 00:33:49,089
你想要

824
00:33:49,089 --> 00:33:49,349
是的

825
00:33:49,349 --> 00:33:50,049
就是这个意思

826
00:33:50,049 --> 00:33:51,839
比如销售车辆

827
00:33:51,839 --> 00:33:52,480
是的

828
00:33:52,480 --> 00:33:57,140
或许这就是Waymo需要与制造伙伴或Wave签约的合作

829
00:33:57,140 --> 00:34:01,049
或与这些公司合作实际销售车辆

830
00:34:01,049 --> 00:34:02,490
打造更完整的网络

831
00:34:02,490 --> 00:34:05,990
但他们尚未这样做

832
00:34:05,990 --> 00:34:09,849
他们供应量如此稀缺

833
00:34:09,849 --> 00:34:13,829
难以有效填补任何区域

834
00:34:13,829 --> 00:34:14,670
仅针对乘车需求

835
00:34:14,670 --> 00:34:15,528
叫车需求

836
00:34:15,528 --> 00:34:17,009
嗯不

837
00:34:17,009 --> 00:34:17,489
叫车需求

838
00:34:17,489 --> 00:34:17,829
我是说

839
00:34:17,829 --> 00:34:20,748
像每英里三美元的价格点

840
00:34:20,748 --> 00:34:22,489
这大致是人们接受的价位

841
00:34:22,489 --> 00:34:24,429
我们认为市场规模

842
00:34:24,429 --> 00:34:25,329
你知道的

843
00:34:25,329 --> 00:34:30,150
会增长数十倍

844
00:34:30,150 --> 00:34:33,050
从三美元到一美元每英里

845
00:34:33,050 --> 00:34:37,208
因此你需要实际的供应

846
00:34:37,208 --> 00:34:37,989
嗯

847
00:34:37,989 --> 00:34:42,009
所以我认为优步的战略是一个很好的选择

848
00:34:42,009 --> 00:34:45,068
只要市场保持供应受限状态

849
00:34:45,068 --> 00:34:47,260
在那个三美元的价格点位

850
00:34:47,260 --> 00:34:50,099
然后就会遭遇灾难性失败

851
00:34:50,099 --> 00:34:51,780
当价格跌至一美元时

852
00:34:52,739 --> 00:34:54,079
太喜欢了 好吧

853
00:34:54,079 --> 00:34:56,389
我们下周再见

854
00:34:56,389 --> 00:35:00,530
希望宇航员能在那之前安全回家

855
00:35:00,530 --> 00:35:03,030
Arc投资管理有限责任公司是SEC

856
00:35:03,030 --> 00:35:04,880
注册投资顾问

857
00:35:04,880 --> 00:35:07,880
金融与公共部门是独立无关实体

858
00:35:07,880 --> 00:35:10,519
彼此之间不涉及任何公司关系

859
00:35:10,519 --> 00:35:13,840
不会进行相互营销或销售产品服务

860
00:35:13,840 --> 00:35:19,800
Arc金融与公共部门不对任何明确声明

861
00:35:19,800 --> 00:35:22,599
任何责任或损失承担

862
00:35:22,599 --> 00:35:23,239
遗漏

863
00:35:23,239 --> 00:35:26,648
其他实体或其关联方的损失或损害

864
00:35:26,648 --> 00:35:27,289
高管

865
00:35:27,289 --> 00:35:28,750
员工或代理人

866
00:35:28,750 --> 00:35:31,230
任何公司不得对间接

867
00:35:31,230 --> 00:35:31,690
附带

868
00:35:31,690 --> 00:35:34,710
后果性或特殊损害承担责任

869
00:35:34,710 --> 00:35:37,670
源于或关联于此播客或其内容

870
00:35:37,670 --> 00:35:41,429
本节目提供的信息仅作参考用途

871
00:35:41,429 --> 00:35:47,239
不应作为投资决策依据且可能随时变更

872
00:35:47,239 --> 00:35:53,528
不构成Arc公司任何形式的服务或产品提供

873
00:35:53,528 --> 00:36:00,489
投资者需自行判断特定投资服务是否符合需求

874
00:36:00,489 --> 00:36:06,480
关于公司或证券的所有陈述均为Arc的个人观点

875
00:36:06,480 --> 00:36:07,260
或节目嘉宾

876
00:36:07,260 --> 00:36:10,880
不代表Arc对任何公司或证券的认可

877
00:36:10,880 --> 00:36:12,880
或建议买入

878
00:36:12,880 --> 00:36:15,230
卖出或持有任何证券

879
00:36:15,230 --> 00:36:18,679
历史结果不代表未来表现

880
00:36:18,679 --> 00:36:23,019
本节目部分陈述可能为未来展望

881
00:36:23,019 --> 00:36:25,920
及其他前瞻性声明

882
00:36:25,920 --> 00:36:27,409
基于Arc当前观点与假设

883
00:36:27,409 --> 00:36:32,230
涉及已知或未知风险可能导致实际结果

884
00:36:32,230 --> 00:36:37,550
与预期存在重大差异

885
00:36:37,550 --> 00:36:42,130
Arc不承担更新前瞻性信息的义务

886
00:36:42,130 --> 00:36:45,869
Arc及其客户及相关人员可能

887
00:36:45,869 --> 00:36:51,719
但未必持有讨论中的证券或发行人权益

888
00:36:51,719 --> 00:36:55,639
部分信息源自Arc认为可靠的第三方

889
00:36:55,639 --> 00:36:56,059
然而

890
00:36:56,059 --> 00:37:01,659
Arc不对第三方信息的准确性完整性负责

共 1 条，按点赞数排序。

采集完成，该视频无弹幕