2022 年 4 月 25 日,马斯克和 Twitter 签下 440 亿美元的收购协议。签字之前,他没有向 Twitter 问过一个关于虚假账号的问题。
这句话出自他自己的庭外取证笔录。律师问他:2022 年 4 月 25 日签约之前,你有没有就虚假账号在平台上的普遍程度问过哪怕一个问题?马斯克说,他跟 Jack Dorsey 聊过很多次,问过 Dorsey 打算怎么处理那些扰乱系统的假号和垃圾号。律师追问,除了跟 Dorsey 的这些对话,还有别的吗。他说,想不起来了。
这份笔录来自 Pampena 诉马斯克案,原告是 Twitter 股东,被告是马斯克。争的是他在 2022 年那场收购拉锯里说过的话、发过的推,以及这些话对市场造成的实际影响。对面律师的提问方式很统一:把他在公开场合的每一句断言,拿去和他私下的实际动作对照。
对照出来的结果,是一份相当安静的笔录。马斯克在这份文件里回答得最多的三个字是「我不记得」。
他要求了尽职调查的权利,签约前一次都没用
收购协议里,马斯克给自己争到了一项权利:确认性商业尽职调查(confirmatory business due diligence)。律师问他,为什么要这项权利。
「我们只是需要确认他们的公开声明是准确的。」(马斯克)
律师问,还有更具体的理由吗。他说没有。
接下来是一串很扎实的追问。签约前,你或者你的代表有没有要求过支撑那个估算的底层数据?「我想没有。」签约前,你有没有要求和参与估算过程的 Twitter 员工谈过?「不记得有过。」合并协议里,有没有人阻止你加入条款,要求 Twitter 描述它的计算方法、并提供所有备份信息?「不记得有。」有没有人阻止你加入条款,要求 Twitter 提供用户的原始数据?「不记得有。」
律师问到这里,笔录上留下了一个注记:一声嗤笑。
这四问连在一起,指向同一件事。马斯克后来指控 Twitter 隐瞒信息,但他在拥有谈判筹码的那一刻,没有把这些要求写进合同,也没有在签字前行使过一次提问权。他要的是一个可以随时启动的权利,但他一次都没启动。签字那天,他手上关于虚假账号的信息,和他刷推特时看到的一样多。
这不是事后诸葛亮。合并协议是可以写条款的,他自己也承认没有任何人拦着他写。
推文说「交易搁置」,可没人停下手里的活
2022 年 5 月 15 日,马斯克发推,大意是:Twitter 收购交易暂时搁置,等待支撑「虚假账号占比低于 5%」这一计算结果的细节。
律师逐字拆这条推文。你提到的这个计算,指的是 Twitter 在 SEC 申报文件里对虚假账号的计算,对吗?对。你说的「支撑这个计算的细节」,指的是你在等 Twitter 提供细节,对吗?对。你说交易搁置、等这些信息,你的意思是,在收到这些信息之前交易不会完成,对吗?
「只是说会有一个暂时的延迟。交易完成的延迟。非常暂时的延迟,直到我收到这些信息。」(马斯克)
然后是三个几乎一模一样的问句:你有没有指示任何顾问暂停收购?没有。你有没有告诉律师停止处理这笔交易?没有。你有没有告诉银行家停止处理这笔交易?没有。
「我们就是在等文件。」(马斯克)
这三问三答,是整份笔录的骨架。公开市场上,一条推文让所有人以为交易停摆;公司内部,律师在干活,银行家在干活,融资在推进,没有人收到停工指令。他对外发布的是一个状态,对内执行的是另一个状态。两个状态之间没有任何衔接文件,只有一条推文。
这种做法在法律上有个直接后果。如果一方真的认为对方的陈述存在重大不实,正常路径是发违约通知、暂停履约、保留证据。马斯克什么都没做。他的团队继续按原计划推进,一直推进到七月初。
「可能超过 90%」,他承认没有任何计算基础
律师问,截至 2022 年 5 月 15 日,你有没有从 Twitter 收到任何分析,显示虚假账号超过 5%?「我想没有。」
那你在推文里说「可能超过 90%」,为什么用这个数字?
「我只是在说,当时我没有任何理性计算的基础。所以,一个非常高的数字并非不可能。」(马斯克)
这个回答的诚实程度值得停下来看一眼。他没有辩称 90% 有依据,他直接承认没有依据,并且给出了自己的推理:因为没有依据,所以极端值不能被排除。
这个推理在逻辑上成立,在信息价值上等于零。一个没有依据的数字,既不能被证实,也不能被证伪,它对判断的帮助是负的。而这条推文发出去的时候,市场把它当成了一条有信息量的声明。特斯拉和 Twitter 的股价都动了。
后面还有一段对话更能说明问题。律师问他,你判断虚假账号的方法是什么。马斯克说,他的疑问分两步:第一,怎么挑出那 100 个被抽样的账号;第二,在这 100 个里,怎么判断哪个是真的、哪个是假的。
「我从来没有得到过他们怎么挑出这 100 个账号的答案。我也从来没有得到过他们怎么判断这 100 个里哪个是真的的答案。」(马斯克)
律师问:这听起来像一个方法吗?
「这不像一个方法。」(马斯克)
「这不像方法」,但他从没要求见做方法的人
这里出现了一个断层。
2022 年 5 月 6 日,在旧金山,Twitter 时任 CEO Parag Agrawal 和时任 CFO Ned Segal 当面给马斯克讲过 Twitter 的抽样方法。马斯克在笔录里承认,他大概当场表达过不满。律师问他具体说了什么,他说不记得。但他记得自己的心态:
「我不会特别高兴,因为他们没法证明系统里最基础的指标之一是怎么算出来的,而这个指标对公司价值至关重要。」(马斯克)
律师顺着这个心态往下问:既然你觉得他们的描述不能让你信服,那你有没有要求见 Twitter 的管理团队、工程师,或者抽样团队的成员,把计算的基础搞清楚?
「我不记得。」5 月 6 日当天没有,5 月 6 日之后也没有?「我不记得。」
这个组合很奇怪。一个人如果真认为对方的核心指标算不出来,正常反应是找做计算的人当面对质,看代码、看样本、看判定规则。马斯克的选择是发推。他把一个技术争议搬到了公开市场,而没有在技术层面推进过一次。
更要紧的是时间点。他第一次认真看 MDAU(monetizable daily active users,可变现日活跃用户)这个指标,按他自己的说法,是在和 Parag、Ned 见面的时候。律师问,那是 5 月 6 日在旧金山吗?他说记不清确切日期,但应该是在交易签署之后。
签下 440 亿美元的那一天,他还没有细看过这家公司最核心的用户指标。这句话他自己说出来了,说得很平静。
7 月 1 日到手的备份数据,7 月 8 日终止交易时他还不知道
这一段是整份笔录里最有戏剧性的部分,而且戏剧性来自流程,不来自任何人的表态。
2022 年 7 月 1 日,马斯克这边的律所 Quinn Emanuel 收到了 Twitter 抽样过程的备份数据。7 月 8 日,马斯克方发出终止交易的通知。
律师问他:有没有人告诉过你,Quinn Emanuel 在 2022 年 7 月 1 日收到了关于 Twitter 抽样过程的备份数据?「我不记得。」如果在你一周后终止交易之前知道这件事,会不会是你想知道的?「大概会。」
接着是另一条线。Twitter 曾经邀请马斯克的数据科学家和 Twitter 的数据科学家见面,邀请函抄送给了 Alex Spiro、Andrew Rossman,以及 Skadden Arps 的律师。律师问他,有人告诉过你这个邀请吗?「我想没有。」Quinn Emanuel 有人把这封信转给你吗?「我不记得。」
律师把话收紧:也就是说,你不记得任何一位被抄送在这些信件上的顾问,Alex Spiro、Andrew Rossman,或者 Skadden Arps 的律师,向你转达过 Twitter 关于数据科学家会面的邀请,对吗?
「对,我不记得。」(马斯克)
笔录里还引了一封邮件。一位名叫 Kate 的人在第四段写道,作为跟进,Ned 提出和马斯克、Parag 做一次小范围的讨论,他认为这比等原始数据输出更有建设性,也更有帮助,因为部分数据受留存期限限制,用「活跃度」和「MDAU」的差别来解释,能更好地把问题框清楚。
律师问,你从任何团队成员那里收到过这条信息吗?「我不记得。」之后有没有和 Ned 或 Parag 做过跟进的小组讨论?「我想没有。」他们的方法和 Twitter 原来的做法有什么不同?「我不记得。」
信息到了他自己的律师手里,邀请发到了他自己的顾问邮箱里,对方主动提出降级沟通、用更省事的方式解释方法,而这三条线没有一条通到他这里。一周后他签了终止通知。
这段的杀伤力在于它同时打中两边。对马斯克来说,他主张「我要的信息 Twitter 不给」,可 Twitter 给了,只是没给到他手上,而他自己的团队也没上报。对 Twitter 来说,它确实做了配合动作,但沟通链条断在对方内部,它也没法证明马斯克本人知情。
一个 440 亿美元的交易,关键信息的传递靠的是抄送名单和「我不记得」。
股价一天涨 27%,他自己都不敢信
笔录中间有个短促的插曲,读起来像换了口气。
律师说,这份文件提交之后,股市一天涨了大约 27%。
「哇。一天?你是说一天涨了 27%?」(马斯克)
律师确认,是的,文件提交后一天涨了 27%。马斯克说,这听起来很高。律师又问,那这份文件提交后股价涨了这件事,会让你意外吗。马斯克开始绕:股市有涨有跌,但如果你说它涨了,我其实不知道。然后他又问了一遍:你是说一天涨了 27%?这是对的吗?
从时间线看,这大概率指的是 2022 年 4 月他披露持有 Twitter 9.2% 股份的 13D 文件。市场对「马斯克入股」这件事的反应,比他本人预想的猛得多。
这段插曲值得留意,因为它暴露了一个反差。他对市场情绪是有感知的,他知道自己一条推文能让股价动,他也知道 27% 这个幅度不正常。但在 5 月和 7 月那几条关键推文发出去之前,他没有把这个感知折算成成本。
一个能影响 27% 股价的人,说话的成本和普通人不一样。这一点他在笔录里承认得很自然,好像从没想过它意味着什么。
收购完成后那份分析,他连是谁做的都说不出
交易在 2022 年 10 月完成。之后不久,马斯克这边做了一份关于虚假账号的内部计算。
律师开始拆这份计算。你用了外部数据科学家吗?「没有,我们内部就能做分析。」Twitter 内部是谁做的这份分析?「我不记得。」是一个人,十个人,还是一个团队?「我想是一个团队。」你能指认其中任何一个人吗?「不能。」这份计算在什么地方有文档记录吗?「我不记得。」
四个问题,四个空白。
这份分析是他公开指控 Twitter 的核心依据。他在多个场合说过,收购后自己人一算,虚假账号的比例远高于 Twitter 披露的 5%。可到了取证现场,这份分析没有可指认的执行人,没有可定位的文档,没有可复现的方法。一份用来支撑数百亿美元索赔的计算,在法庭上找不到作者。
这里要留一点余地。大公司内部做数据分析,参与者记不清具体是谁、文档存在哪个系统,并不罕见,尤其当事人是 CEO。这个回答本身不构成撒谎的证据。
但它构成一个事实:他公开引用的那个数字,在他自己这边的证据链上,起点是空的。
广告商、投资人、审计师,都只听过他的推文
接下来是一组关于「告知」的问题,答案的结构惊人地一致。
你有没有告知广告商,虚假账号在平台账号中的占比更高?「我不记得。」你有没有告知股权投资者或债务投资者,虚假账号占比更高?「除了我的公开声明之外,没有。」也就是说,除了公开声明,你和任何股权或债务投资者都没有就这个判断做过直接沟通,对吗?「对,除了公开声明没有。」
接管公司之后,你有没有指示 Twitter 或 X 的任何人去告知广告商,Twitter 此前对虚假账号的估算不准确?「我不记得。」
X 的审计师是谁?「普华永道(PricewaterhouseCoopers)。」你有没有告知普华永道,你已经认定 Twitter 此前对虚假账号的估算不准确?「除了我的公开声明之外,没有。」
这四组回答拼出一条清晰的路径。关于虚假账号的所有对外表达,只有一条通道:他自己的推特账号。广告商没有收到正式通知,股权和债务投资人没有收到正式通知,审计师没有收到正式通知,公司内部也没有人被指示去通知任何人。
这条路径在法律上有两层含义。如果这个判断真的重大,那它应该走正式披露渠道,走推文的方式本身就有问题。如果这个判断没有走任何正式渠道,那它就更像一种公开谈判手段,而马斯克事后主张「我基于这个信息才终止交易」,就少了一大块支撑。
还有一层更实际的含义。收购完成后,X 的主要收入来源是广告。如果 CEO 真的相信平台上一半账号是假的,广告商是第一批需要知道的人,因为他们的钱按曝光量结算。他没有告诉广告商。这个选择本身说明了他对这个数字的信心程度。
出价之前他就认定 Twitter 在夸大用户数
笔录后段,律师把时间线往前推,推到了出价之前。
你告诉过摩根士丹利,在你的掌控下 Twitter 可能值 2000 亿到 3000 亿美元。有这回事吗?「我可能说过。」你当时是这么认为的吗?「我觉得有大量未开发潜力。」
还有一句更关键的。在你提出报价之前的初期阶段,你有没有告诉摩根士丹利,你认为 Twitter 的真实用户数远低于它披露的数字?
「是的,我认为 Twitter 在夸大他们的数字。」(马斯克)
这两句话放在一起,构成一个很难绕过去的组合。他在出价前就认定这家公司的核心用户数据注水,他仍然出了 440 亿美元,他同时预期这笔资产在自己手里能翻到 2000 亿至 3000 亿。
那么「我被虚假账号数据误导才出价」这个主张,就要面对一个反问:你出价的时候就已经不信这些数据了,你怎么会被它误导。
他自己的解释是「未开发潜力」。这个解释在商业上完全说得通。一个认为标的被低估、并且相信自己能改造它的人,本来就会在数据不完美的时候出手。问题是,这个逻辑和他后来终止交易的逻辑互相抵消。
律师还问了一句旁证。截至 2022 年 4 月 11 日,你看过 MDAU 的数字吗?他说不确定日期。第一次关注这些数字是什么时候?可能是我和 Parag、Ned 见面的时候。那是在交易签署之后?「我想是的。」
签约在 4 月 25 日,见面在 5 月 6 日。他先买,后看指标,再质疑指标,最后用指标当理由退出。这个顺序在商业上不常见,在法律上很难讲通。
最后让他掏钱的,是他不信任那位法官
尽管有上面这一切,马斯克最终还是撤销了自己的主张,完成了对 Twitter 的收购。
律师问,你这么做,是因为你认为特拉华州的法官会命令你买下它?
「我被告知,我们从 Kathaleen McCormick 那里不会得到公平的听证。她对我有强烈的偏见。她后来关于特斯拉股东投票的裁决证明了这一点。」(马斯克)
律师问,这是你在 Twitter 案上让步的原因吗?「是的,我不相信我会得到公平的听证。」还有别的原因吗?「没有。」你为什么认为她有偏见?「就因为她的裁决。」
然后是两个收口的问题。即使没有你声称 Twitter 未提供的信息,你还是拿到了债务融资?「债务融资确实到位了。」你还是拿到了股权融资?「是的。」
这两个回答把「信息缺失导致交易无法完成」这条线彻底切断。钱到位了,融资没停,律师没停,银行家没停。他缺的从来不是数据,是一个他愿意接受的裁决结果。
这一段是整份笔录里最锋利的地方,而且锋利得毫不费力。他没有回避,没有修饰,直接给出了真实动机:他不是被说服了才买,他是判断自己会输才买。
一个在公开场合把「虚假账号」讲了几个月的人,在法庭上承认,最后决定这笔 440 亿美元交易的那件事,是法官的倾向,不是数据的真假。
这份笔录真正记录了什么
把整份证词读完,容易得出一个草率的结论:他在撒谎。这个结论太快了,也不准确。
笔录里更多的是空白,不是矛盾。他记不清谁做的分析、记不清收到过什么邮件、记不清会上说了什么、记不清有没有转达邀请。这些「不记得」在法律上是有意义的,它们不能被当成否认,只能被当成无法证实。原告律师拿到的是一堆洞,不是一个谎。
真正被记录下来的是一个结构。公开叙事和内部执行,在这笔交易里是两条平行轨道,几乎不交汇。推文负责施压,团队负责推进,两条轨道各自运行,中间没有文件、没有会议纪要、没有正式的决策记录。
这个结构在短期内极其高效。一条推文就能让对手的股价抖动,一个「搁置」就能换来谈判空间,而所有实际工作照常进行。它在长期内极其脆弱。当对方把证据递到你律师手上,当你的数据科学家被邀请去开会,当法官问你「你当时到底想要什么」,这个结构给不出任何答案。
最有意思的是,这个结构最后伤到了他自己。7 月 1 日备份数据到了 Quinn Emanuel,7 月 8 日他签了终止通知,中间七天,信息卡在某个他记不清的环节。他的团队为了保护他,或者只是为了流程效率,把可能改变决策的信息留在了半路。
443 亿美元买一家公司,最后决定这件事的,是一位他不信任的法官,和一堆他记不清的邮件。
对做技术、做产品、做交易的人来说,这份笔录有不少可以直接拿走的东西。
- 把公开表态和合同条款分开记。你在推上、在会上、在饭桌上说的话,不构成任何一方的义务。真正有约束力的只有写进合同的那几行。每次你准备公开发表一个强硬立场之前,先确认对应的条款是否存在。不存在,那句话就只是情绪。
- 尽调权利要在签字前用掉。合同里争到的确认性尽调权、信息索取权、会面权,价值全部集中在签字之前。签字之后再用,对方有一百种方式拖延。把「我要看底层数据」这个要求往前挪,挪到还没有承诺任何东西的时候。
- 给关键信息设一条必须上报的路由。这个案子里最贵的失误是流程失误:备份数据到了律所,邀请函发到了顾问邮箱,两条线都没通到决策人。无论团队多大,都该有一份清单:哪些类型的信息到了谁手里,必须在多少小时内上报到谁。清单要写下来,不能靠自觉。
- 判断一个人是不是真想退出,看他的团队有没有停工。公开说「暂停」很便宜,让律师停下、让银行家停下、让融资条款失效,这些才贵。对方只发声明不停工,说明他在谈判,不在退出。反过来,你自己要退出的时候,也要清楚停工这个动作会被对方当成证据。
- 引用一个数字之前,先确认它在你的证据链上有起点。谁算的、怎么算的、文档在哪、能不能复现。这四问答不上来,这个数字在公开场合可以煽动情绪,在法庭和董事会上等于零。你自己内部做的分析尤其要留痕,因为那是你唯一能拿出来的东西。
- 读原始文件,别只读标题。这场取证的所有新闻标题都指向「马斯克承认虚假账号可能超过 90%」,但真正有信息量的是他紧接着说的那句「我当时没有任何理性计算的基础」,以及他在融资、停工、告知义务上的那些回答。原始笔录很长、很重复、很枯燥,价值全在那些重复里。
0:00The calculation that you were referring to in this tweet, were you referring to Twitter's calculation of fake accounts underlying its SEC filings, right?
0:08Yes. And the details supporting the calculation refers to the fact that you were waiting on details from Twitter supporting its calculation, right? Yes. And when you said the deal was on hold pending this information from Twitter, you meant that until you received this information, the deal would not close, right?
0:28Just that it would that is there's a temporary delay. Delay on the deal closing. Yeah, it's a very temporary a temporary delay. Until you receive this information.
0:39Yes. Did you instruct any of your advisors to suspend the purchase? No. Did you tell your attorneys to stop working on the deal? No. Did you tell your bankers to stop working on the deal? No, this was we're literally just waiting for the documents. Had you received at this point in time, May 15th, any analysis from Twitter that there was more than 5%?
1:07I don't think so. When you said might be over 90%, why did you use that number? I'm simply saying that I had no basis at this point for a rational number calculation. So, it was not impossible that it it could be some very high number. And just so I'm following So, the the first question is how you identify those 100 and then of those 100, how do you determine whether real or fake? Yes. And you're saying Twitter didn't give you that answer.
1:39I I never got an answer for how they picked the 100. And I never got an answer for how they determined of the 100, which one were which one's real. Does that sound like a method to you? It sound like a method to me. And your advisor
1:56Sounds like Did anyone ever tell you that Quinn Emanuel had received backup data on their on Twitter's sampling process on July 1st, 2022? I don't recall. Would that have been something that you would have wanted to know before you terminated the deal a week later? Probably. Did you find it useful for your data scientists to meet with Twitter's data scientists?
2:20I would have that would make sense. Yeah. Do you know if that was done? I don't know. Did anyone advise you that Twitter had made that invitation? I don't think so.
2:41It's copied to Alex Spiro and Andrew Rossman. Did anyone at Quinn Emanuel pass along this letter to you? I don't recall. And is it fair to say that you don't recall any of your advisors that are copied on this letter, Alex Spiro and Rossman, or the attorneys from Skadden Arps relaying the invitation that Twitter had made for your data scientists to meet?
3:09Correct? I don't recall. Kate also writes on the fourth paragraph, quote, "As follow-up, he offered to have a smaller group discussion with Elon to for him and Parag. He said he thought that would be constructive and more helpful than waiting for raw data output, some of which is limited due to data retention, and instead explaining activity versus MDAU will better frame the question." Close quote.
3:37Did you ever receive that message from anyone of your team members? I don't recall. Did you have any follow-up group discussions with Ned or Parag? I don't think so. How did their methodologies differ from what Twitter was doing? I don't recall. And the stock market went up about 27%.
3:59Wow. In one day? Does that surprise you? It went up went up 20%. Are you saying it went up 27% in one day? It went Yes, 27% after this filing was made. In one day? Would that surprise you? That sounds high. Would it surprise you that it went up at all after this filing? Well, the stock market stocks go up and down, but but if you're saying it went up I actually don't know You're saying it it went up 27% in one day?
4:31Is is that correct? When did you conduct the analysis? shortly after the acquisition. Did you use outside data scientists to perform this calculation? No, we just we could do internal analysis. Okay. And who at Twitter performed this analysis? I don't recall. And what was it one person, 10 people, was it a team? I think it was a team.
5:00And you can't identify any of those people? Not a fan. was this calculation documented somewhere? I don't recall. did you inform advertisers that fake accounts represented a higher percentage of the platform's account holders? I don't recall. Do you Did you inform your equity or debt investors that fake accounts represented a higher percentage of the platform's account holders? Not beyond my public statements.
5:37So, you did not have any direct communications with any equity or debt investor about your determination that spam and fake accounts represented a higher percentage of the platform's holders, correct? Other than your public statements.
5:58Not other than the public statements. After you took control of the company, did you instruct anybody at Twitter or X to inform advertisers that Twitter's prior estimates of fake accounts were inaccurate? I don't recall. Do you have a public accountant at X? PricewaterhouseCoopers.
6:29Is PricewaterhouseCoopers your accountant? Our accountants. Did you inform PricewaterhouseCoopers that you had determined that Twitter's prior estimates of fake accounts were inaccurate? not beyond my public statements. Yet after all that, you agreed eventually to dismiss your claims and to buy Twitter, correct?
6:45Yes. And you did that because you thought the judge in Delaware would order you to buy it. I was told that we would not receive a fair hearing from Kathleen McCormick. that she has she had a strong bias against me. And her you know, subsequent ruling regarding the Tesla shareholder vote showed that to be true.
7:11And is that the reason you agreed to capitulate on the Twitter case? Yes. I did not believe I would receive a fair hearing. Any other reason? No. Why do you think she's biased against you? Just because of her rulings?
7:24Yes. Were you able to obtain the debt financing even without the information that you claim Twitter had not provided you? The debt financing came through, yes. Were you able to obtain equity financing even without the information that you claim Twitter had not provided you? Yes.
7:41You told Morgan Stanley that Twitter could be worth 200 to 300 billion dollars under your ownership. Yeah. Do you remember that? I may have said that. Did you feel that at the time? I thought there was a lot of untapped potential. Did you tell Morgan Stanley during these initial stages before you made an offer that you believed Twitter's real user numbers were much lower than the numbers Twitter was disclosing?
8:09I Yeah, I thought I thought that Twitter was exaggerating their numbers. Have you ever heard of the term monetizable daily active users? Yes. And that was the metric that Twitter was using? Yes. Had you as of April 11th, 2022, looked at its MDAU numbers? I I'm not sure of the date. Do you remember when the first time is that you focused on those numbers?
8:37I think it may have been when I met with Parag and Ned. In San Jose? I think it was in San Francisco.
8:45Oh, in San Francisco on May 6th? I don't know the exact date, but
8:49Okay. But it was after the deal closed? Or sorry, it was after the deal was signed? I believe so, yes. Okay. You wanted the right to take confirmatory business due diligence as part of this offer. Yes. Okay. And why did you want to do that? we just need to make sure that the their public statements are accurate.
9:19Right. Anything more in particular? No. Prior to signing, did you or anyone on acting on your behalf ever ask for the underlying data used to make that estimate? I don't think so. Prior to signing, did you ever ask to speak to any Twitter employees involved in their estimation process?
9:50Not that I can recall. Prior to signing on April 25th, 2022, had you asked a single question about the prevalence of fake accounts on Twitter's platform? I mean I raised many many times with Jack with the with the you know, what was he doing about the fake and spam accounts that were playing with the system. Other than these conversations with Mr. Dorsey that you were described before, any other conversation?
10:20not that I can recall. Did anything or anyone prevent you from including language in the merger agreement requiring Twitter to describe its methodology and to provide you with all backup information it used to estimate the percentage of fake accounts? Not that I can recall. Anything prevent you from including specific language in the agreement requiring Twitter to provide you with Twitter's raw data regarding its users? [snorts]
10:53Not that I can recall. At this meeting was anybody upset? With you to the other side like perceptively upset? I don't recall. So at the meeting itself, did you express your displeasure to them? I probably did. Do you remember one way or the other? I don't recall exactly what I said, but I wouldn't have been particularly pleased that they could not just justify that the how they calculated one of the most fundamental metrics in the system that was fundamental to the value of the company. I understand that's how you felt internally, but did you express that to anyone at Twitter at the meeting?
11:36I think I did. Do you remember specifically what you said? No. Is it fair to say that Ned and Parag's description of Twitter's methodology was not so concerning that it caused you to request to meet with Twitter's management team or its engineers or its sampling team team members to better understand the basis for the calculations?
12:00I don't recall. You never did that on May 6th and you never did that after May 6th, right? I don't recall. The calculation that you were referring to in this tweet, were you referring to Twitter's calculation of fake accounts underlying its SEC filings, right? Yes. And the details supporting the calculation refers to the fact that you were waiting on details from Twitter supporting its calculation, right?
12:23Yes. And when you said the deal was on hold pending this information from Twitter, you meant that until you received this information, the deal would not close, right? Just that it would there that is there's a temporary delay. Delay on the deal closing. Yeah, it's a very temporarily a temporary delay.
12:43Until you receive this information. Yes. Did you instruct any of your advisors to suspend the purchase? No. Did you tell your attorneys to stop working on the deal? No. Did you tell your bankers to stop working on the deal? No, this is we're literally just waiting for the documents. Had you received at this point in time, May 15th, any analysis from Twitter that there was more than 5%?
13:13I don't think so. When you said might be over 90%, why did you use that number? I'm simply saying that I had no basis at this point for a rational number calculation. So it was not impossible that it it could be some very high number. And just so I'm following, so the first question is how you identify those 100 and then of those 100, how do you determine whether real or fake? Yes.
13:42And you're saying Twitter didn't give you that answer. I never got an answer for how they picked the 100. And I never got an answer for how they determined of the 100 which one were which one's real. Does that sound like a method to you? It doesn't sound like a method to me. And your advisor
14:02Sounds like Did anyone ever tell you that Quinn Emanuel had received backup data on on Twitter's sampling process on July 1st, 2022? I don't recall. Would that have been something that you would have wanted to know before you terminated the deal a week later? Probably. Did you find it useful for your data scientist to meet with Twitter's data scientists?
14:26I would have that would make sense. Yeah. Do you know if that was done? I don't know. Did anyone advise you that Twitter had made that invitation? I don't think so.
14:46It's copied to Alex Spiro and Andrew Rossman. Did anyone at Quinn Emanuel pass along this letter to you? I don't recall. And is it fair to say that you don't recall any of your advisors that are copied on this letter, Alex Spiro and Andrew Rossman or the attorneys from Skadden Arps relaying the invitation that Twitter had made for your data scientists to meet?
15:14Correct? I don't recall. Kate also writes in the fourth paragraph, quote, "As follow-up, he offered to have a smaller group discussion with Elon to for him and Parag. He said he thought that would be constructive and more helpful than waiting for raw data output, some of which is limited due to data retention, and instead explaining activity versus MDAU will better frame the question." Close quote.
15:43Did you ever receive that message from any one of your team members? I don't recall. Did you have any follow-up group discussions with Ned or Parag? I don't think so. How did their methodologies differ from what Twitter was doing? I don't recall. And the stock market went up about 27%.
16:05Wow. In one day? Does that surprise you? It went went up 20%. Are you saying it went up 27% in one day? It went Yes, 27% after this filing was made. In one day? Would that surprise you? That sounds high. Yeah. Would it surprise you that it went up at all after this filing? Well, the stock market stocks go up and down, but but if you say it went up went up actually don't do You're saying it went It went up 27% in one day?
16:37Is is that correct? When did you conduct the analysis? shortly after the acquisition. Did you use outside data scientists to perform this calculation? No, we just we could do internal analysis. And who at Twitter performed this analysis? I don't recall. You know, what was it one person, 10 people, was it a team? I think it was a team.
17:06And you can't identify any of those people? Not that I can. Was this calculation documented somewhere? I don't recall. Did you inform advertisers that fake accounts represented a higher percentage of the platform's account holders? I don't recall. Did you Did you inform your equity or debt investors that fake accounts represented a higher percentage of the platform's account holders? Not beyond my public statements.
17:43So you did not have any direct communications with any equity or debt investor about your determination that spam and fake accounts represented a higher percentage of the platform's holders, correct? Other than your public statements.
18:03Not other than the public statements. After you took control of the company, did you instruct anybody at Twitter or X to inform advertisers that Twitter's prior estimates of fake accounts were inaccurate? I don't recall. Do you have a public accountant at X? PricewaterhouseCoopers.
18:34Is Pricewaterhouse your accountant? I believe so. Did you inform Pricewaterhouse that you had determined that Twitter's prior estimates of fake accounts were inaccurate? not beyond my public statements. Yet after all that, you agreed eventually to dismiss your claims and to buy Twitter, correct?
18:51Yes. And you did that because you thought the judge in Delaware would order for you to buy it. I was told that we would not receive a fair hearing from Kathleen McCormick. that she has she had a strong bias against me. And her you know, subsequent ruling regarding the Tesla shareholder vote showed that to be true.
19:17And is that the reason you agreed to capitulate on the Twitter case? Yes. I did not believe I would receive a fair hearing. Any other reason? No. Why do you think she's biased against you? Just because of her rulings? Yes. Were you able to obtain the debt financing even without the information that you claim Twitter had not provided you?
19:38The debt financing came through, yes. Were you able to obtain equity financing even without the information that you claim Twitter had not provided you? Yes. You told Morgan Stanley that Twitter could be worth 200 to 300 billion dollars under your ownership. Yeah. Do you remember that?
19:55I may have said that. Did you feel that at the time? I thought there was a lot of untapped potential. Did you tell Morgan Stanley during these initial stages before you made an offer that you believe Twitter's real user numbers were much lower than the numbers Twitter was disclosing?
20:15I Yeah, I thought I thought Twitter was exaggerating the numbers. Have you ever heard of the term monetizable daily active users? Yes. And that was the metric that Twitter was using? Yes. Had you as of April 11th looked at its MDAU numbers? I'm not sure of the date. Do you remember when the first time is that you focused on those numbers?
20:43I think it may have been when I met with Parag and Ned. In San Jose? I think it was in San Francisco. Oh, in San Francisco on May 6th? I don't know the exact date, but But it was after the deal closed? Or sorry, it was after the deal was signed? I believe so, yes. Okay. You wanted the right to take confirmatory business due diligence as part of this offer.
21:08Yes. And why did you want to do that? we just need to make sure that the the public statements are accurate. Right. Anything more in particular? No. Prior to signing, did you or anyone on acting on your behalf ever ask for the underlying data used to make that estimate? I don't think so.
21:49Prior to signing, did you ever ask to speak to any Twitter employees involved in their estimation process? Not that I can recall. Prior to signing on April 25th, 2022 had you asked a single question about the prevalence of fake accounts on Twitter's platform? I mean I raised many times with Jack but the but the you know, what was he doing about the fake and spam accounts that were plaguing the system. Other than these conversations with Mr. Dorsey that you were described before, any other conversations?
22:26not that I can recall. Did anything or anyone prevent you from including language in the merger agreement requiring Twitter to describe its methodology and to provide you with all backup information it used to estimate the percentage of fake accounts? Not that I can recall. Anything prevent you from including specific language in the agreement requiring Twitter to provide you with Twitter's raw data regarding its users?
22:58Not that I can recall. This meeting, was anybody upset? If you know, to the other side like perceptively upset. I don't recall. So, at the meeting itself, did you express your displeasure to them? I probably did. Do you remember one way or the other? I don't recall exactly what I said, but I wouldn't have been particularly pleased that they could not just justify that the how they calculated one of the most fundamental metrics in the system that was fundamental to the value of the company. I understand that's how you felt internally, but did you express that to anyone at Twitter at the meeting?
23:42I think I did. Do you remember specifically what you said? No. Is it fair to say that Ned and Parag's description of Twitter's methodology was not so concerning that it caused you to request to meet with Twitter's management team or its engineers or its sampling team team members to better understand the basis for the calculations?
24:05I don't recall. You never did that on May 6th and you never did that after May 6th, right? I don't recall. The calculation that you were referring to in this tweet, were you referring to Twitter's calculation of fake accounts underlying its SEC filings, right? Yes. And the details supporting the calculation refers to the fact that you were waiting on details from Twitter supporting its calculation, right?
24:29Yes. And when you said the deal was on hold pending this information from Twitter you meant that until you received this information, the deal would not close, right? Just that it would there there is there's a temporary delay. Delay on the deal closing. Yeah, it's temporary temporarily temporary delay.
24:49Until you receive this information. Yes. Did you instruct any of your advisors to suspend the purchase? No. Did you tell your attorneys to stop working on the deal? No. Did you tell your bankers to stop working on the deal? No, this is we're literally just waiting for the documents. Had you received at this point in time, May 15th, any analysis from Twitter that there was more than 5%?
25:18I don't think so. When you said might be over 90%, why did you use that number? I'm simply saying that I had no basis at this point for a rational number calculation. So, it was not impossible that it it could be some very high number. And just so I'm following, so the first question is how you identify those 100 and then of those 100, how do you determine whether they're real or fake? Yes.
25:48And you're saying Twitter didn't give you that answer. I never got an answer for how they picked the 100. And I never got an answer for how they determined of the 100 which one were which one's real. Does that sound like a method to you? It doesn't sound like a method to me. And your advisor
26:08Sounds like Did anyone ever tell you that Quinn Emanuel had received backup data on their on Twitter's sampling process on July 1st, 2022? I don't recall. Would that have been something that you would have wanted to know before you terminated the deal a week later? Probably. Did you find it useful for your data scientists to meet with Twitter's data scientists?
26:32I it would have That would make sense. Do you know if that was done? I don't know. Did anyone advise you that Twitter had made that invitation? I don't think so.
26:52It's copied to Alex Spiro and Andrew Rossman. Did anyone at Quinn Emanuel pass along this letter to you? I don't recall. And is it fair to say that you don't recall any of your advisors that are copied on this letter, Alex Spiro, Andrew Rossman or the attorneys from Skadden Arps relaying the invitation that Twitter had made for your data scientists to meet?
27:20Correct? I don't recall. Kate also writes on the fourth paragraph, quote, "As follow-up, he offered to have a smaller group discussion with Elon two for him and Parag. He said he thought that would be constructive and more helpful than waiting for raw data output. Some of which is limited due to data retention and instead explaining activity versus MDAU will better frame the question." Close quote.
27:48Did you ever receive that message from any one of your team members? I don't recall. Did you have any follow-up group discussions with Ned or Parag? I don't think so. How did their methodologies differ from what Twitter was doing? I don't recall. Okay. The stock market went up about 27%
28:10Wow. In one day? Does that surprise you? It went went up 20%. Are you saying it went up 27% in one day? It went Yes, 27% after this filing was made. In one day? Would that surprise you? That sounds high. Would it surprise you that it went up at all after this filing? Well, the stock market stocks go up and down, but but if you say it went up actually don't do you're saying it would it went up 27% in one day?
28:42Is is that correct? When did you conduct the analysis? shortly after the acquisition. Did you use outside data scientists to perform this calculation? No, we just we could do internal analysis. And who at Twitter performed this analysis? I don't recall. And what was it one person, 10 people, was it a team?
29:10I think it was a team. And you can't identify any of those people? Not off hand. was this calculation documented somewhere? I don't recall. did you inform advertisers that fake accounts represented a higher percentage of the platform's account holders? I don't recall. Do you Did you inform your equity or debt investors that fake accounts represented a higher percentage of the platform's account holders? Not beyond my public statements.
29:49So, you did not have any direct communications with any equity or debt investor about your determination that spam and fake accounts represent a higher percentage of the platforms folders, correct? Other than your public statements.
30:09Not other than the public statements. After you took control of the company, did you instruct anybody at Twitter or X to inform advertisers that Twitter's prior estimates of fake accounts were inaccurate? I don't recall. Do you have a public accountant at X? PricewaterhouseCoopers.
30:40Is PricewaterhouseCoopers your accountant? I believe so. Did you inform PricewaterhouseCoopers that you had determined that Twitter's prior estimates of fake accounts were inaccurate? not beyond my public statements. Yet after all that, you agreed eventually to dismiss your claims and to buy Twitter, correct?
30:57Yes. And you did that because you thought the judge in Delaware would order you to buy it. I was told that we would not receive a fair hearing from Kathleen McCormick. that she has she had a strong bias against me. And you know, subsequent ruling regarding the Tesla shareholder vote showed that to be true.
31:23And is that the reason you agreed to capitulate on the Twitter case? Yes. I did not believe I would receive a fair hearing. Any other reason? No. Why do you think she's biased against you? Just because of her rulings?
31:36Yes. Were you able to obtain the debt financing even without the information that you claim Twitter had not provided you? The debt financing came through, yes. Were you able to obtain equity financing even without the information that you claim Twitter had not provided you? Yes. You told Morgan Stanley that Twitter could be worth 200 to 300 billion dollars under your ownership.
31:58Yeah. Do you remember that? I may have said that. Did you feel that at the time? I thought there was a lot of untapped potential. Did you tell Morgan Stanley during these initial stages before you made an offer that you believed Twitter's real user numbers were much lower than the numbers Twitter was disclosing?
32:21I Yeah, I thought I thought Twitter was exaggerating the numbers. Have you ever heard of the term monetizable daily active users? Yes. That was the metric that Twitter was using? Yes. Had you as of April 11th looked at its MDAU numbers? I'm not sure of the date. Do you remember when the first time is that you focused on those numbers?
32:48I think it may have been when I met with Parag and Ned in San Jose? I think it was in San Francisco. Oh, in San Francisco on May 6th? I don't know the exact date, but But it was after the deal closed? Or sorry, it was after the deal was signed? I believe so, yes. Okay. You wanted the right to take confirmatory business due diligence as part of this offer.
33:14Yes. Okay. And why did you want to do that? we just need to make sure that the the public statements are accurate. Right. Anything more in particular? No. Prior to signing, did you or anyone on acting on your behalf ever ask for the underlying data used to make that estimate? I don't think so.
33:54Prior to signing, did you ever ask to speak to any Twitter employees involved in their estimation process? Not that I can recall. Prior to signing on April 25th, 2022 had you asked a single question about the prevalence of fake accounts on Twitter's platform? I mean I raised many many times with Jack with the with the you know, what was he doing about the fake and spam accounts that were plaguing the system. Other than these conversations with Mr. Dorsey that you were described before, any other conversations?
34:32not that I can recall. Did anything or anyone prevent you from including language in the merger agreement requiring Twitter to describe its methodology and to provide you with all backup information it used to estimate the percentage of fake accounts? Not that I can recall. Anything prevent you from including specific language in the agreement requiring Twitter to provide you with Twitter's raw data regarding its users?
35:04Not that I can recall. At this meeting, was anybody upset? With you to the other side like perceptively upset? I don't recall. So, at the meeting itself, did you express your displeasure to them? I probably did. Do you remember one way or the other? I don't recall exactly what I said, but I wouldn't have been particularly pleased that they could not just justify that the how they calculated one of the most fundamental metrics in the system that was fundamental to the value of the company.
35:43I understand that's how you felt internally, but did you express that to anyone at Twitter at the meeting? I think I did. Do you remember specifically what you said? No. Is it fair to say that Ned and Parag's description of Twitter's methodology was not so concerning that it caused you to request to meet with Twitter's management team or its engineers or its sampling team team members to better understand the basis for the calculations?
36:11I don't recall. You never did that on May 6th and you never did that after May 6th, right? I don't recall.
资源与出处
- 视频在 YouTube 收看
- 档案silicon-valley.tech 档案页本页视频与文稿数据来自 Elon Musk Archive
- 精读本篇精读 Markdown 版