Craig Wright Admits to Editing Bitcoin White Paper Presented in COPA Trial

CoinDeskPolicy發佈於 2024-02-22更新於 2024-02-23

文章摘要

COPA's trial to find out whether or not Craig Wright is the pseudonymous creator of bitcoin Satoshi Nakamoto has completed its third week.

  • The bitcoin developers' lawyer, Alexander Gunning, presented evidence that Craig Wright had made new edits to his whitepaper, which Wright acknowledged.
  • The Crypto Open Patent Alliance wants to prove that Wright's claim to be the founder of bitcoin is a lie backed forgeries.

Craig Wright admitted to making changes to the version of the bitcoin whitepaper he presented in the Crypto Open Patent Alliance's (COPA) trial while testifying Friday.

The trial to prove whether or not Wright is the anonymous creator of the bitcoin white paper completed its third week. COPA wants to prove that Wright's claim to be Satoshi Nakamoto is a lie afforded by "industrial style forgeries," and the bitcoin developers lawyer Alexander Gunning is helping them.

On Friday, Gunning demonstrated that Wright made edits to the bitcoin whitepaper in his "LaTeX files," which Wright agreed were accurate. Wright said the edits were simply a demonstration for his representatives at Shoosmiths (his law firm).

Advertisement
Advertisement

"You were not showing this to anyone, we know the times you were showing this to Shoosmiths, you were doing it for yourself," Gunning said.

"What you are doing is tweaking parameters.. to get them to fit " the layout of the bitcoin whitepaper, Gunning added. The file was uploaded as recently as November 2023, Gunning said.

Gunning ended his questioning by asking: "Your claim to be Satoshi Nakomoto is a fraudulent claim isn't it?" which Wright disputed.

Week three

Wright's testimony capped off the third week of the trial, which saw some of COPA's witnesses took to the stand to face questioning from Wright's lawyers.

Zooko Wilcox-O'Hearn, a computer scientist and the founder of Zcash, testified on Thursday where he was questioned on how well he knew Nakamoto. Wilcox said he wouldn't call himself "pals" with the pseudonymous Bitcoin creator. In court documents, he said that he was not sure if he had any private conversations with Nakamoto.

Other witnesses were more confident about their interactions with Nakamoto.

Computer scientist Marti Malmi spoke on Wednesday, disputing the dates that Wright put forward about Malmi's interaction with Nakamoto. Malmi later released his emails with Nakamoto on X (formerly Twitter).

Wright said in his witness statement that Malmi approached Nakamoto in Feb. 2009 but Malmi said in his statement this was "incorrect," and that the date was actually May 1, 2009.

Adam Back, the CEO of bitcoin technology company Blockstream, said in his first statement he corresponded with someone purporting to be Nakamoto via email.

Advertisement
Advertisement

In his statement he showed an email he received from Nakamoto on Aug. 20, 2008, where he said he planned to cite Back's paper about a proof-of-work system and Back responded by sending more resources.

Wright described Back's interactions with Nakamoto as "dismissive," which Back said was inaccurate.

Next week will see expert witnesses will be questioned.

Edited by Nikhilesh De.

你可能也喜歡

特朗普媒体公布2.38亿美元亏损,并重新审视其比特币策略

特朗普媒体科技集团在二季度净亏损2.38亿美元后,正重新审视其数字货币储备策略。亏损中约1.904亿美元源于数字资产及相关证券的未实现损失。该公司计划调整储备结构,以保持长期数字资产敞口的同时,降低价格波动影响并提高资产负债表效率。 该公司运营Truth Social等平台,与特朗普关系密切。报告显示,公司已使用期权管理比特币波动并赚取溢价,并将部分比特币投入借贷等收益方案。同时,公司将分配更多资源给核心媒体业务。 截至6月底,公司持有约9477枚比特币。7月份,公司出售了价值约1.596亿美元的比特币相关证券,并用以增持比特币。至7月31日,其储备(包括已质押部分)已增至约14139枚比特币。 公司警告,从比特币赚取额外收益的策略伴随交易对手信用风险及资产损失可能性。部分合作方缺乏大型评级机构评级,可能在市场低迷时违约。若交易无担保,公司可能无法收回比特币。此外,在资产用于收益方案期间,公司的处置权受限。 分析指出,这并非公司首次因比特币投资亏损,显示其模式存在结构性弱点。从单纯持有转向收益策略,在资产波动性之上增加了交易对手风险。其他公司也曾经历类似路径,并在2022年因交易对手违约而受冲击。当前情况呈现一个悖论:公司在增持储备的同时承认使用风险。未来是坚持收益策略还是回归保守持有,有待后续财报揭示。

cryptonews.ru15 分鐘前

特朗普媒体公布2.38亿美元亏损,并重新审视其比特币策略

cryptonews.ru15 分鐘前

720次攻击0成功,Claude Code默认放权,AI替你点“同意”

Claude Code将于8月14日起,在Pro、Max和Team计划的新会话中,默认启用“Auto Mode”。该模式下,AI将自动审批有风险的行动,不再频繁弹窗请求用户确认。 此举的底气来自一项第三方安全测试:针对72种攻击场景的720次尝试,Claude的三款主力模型全部成功防御,攻击成功率降至0。内部数据显示,Auto Mode能拦截89%的危险命令,效率是人类用户(仅拦截13.6%)的6.5倍,并能有效避免因“弹窗疲劳”导致的人类判断力下降。 其安全架构依赖三层防御叠加: 1. **模型对齐**:训练模型区分用户指令与外部内容中的潜在恶意指令。 2. **输入探针**:在AI读取外部内容(网页、文件)前进行检测,标记可疑的提示注入尝试。 3. **动作分类器**:在AI执行动作前,由另一模型审核该操作是否获得用户真实授权、是否具有破坏性或数据泄露风险。 尽管测试数据亮眼,但行业专家指出,0/720的结果不等于绝对安全。该测试由Anthropic委托进行,且攻击场景针对旧模型优化,可能存在未覆盖的盲区。其他研究团队的压力测试曾发现,对于某些不经过动作分类器的操作(如项目内文件改动或第三方软件包指令),漏判率可能很高。 本质而言,默认开启Auto Mode是一次“风险交换”:用AI分类器替代人类进行高频、琐碎的审批,提升了效率和平均安全性,但同时也将具体的判断权从用户移交给了AI系统。用户节省了点击,但也让渡了部分控制。最终,责任的承担者仍然是用户本人。

marsbit36 分鐘前

720次攻击0成功,Claude Code默认放权,AI替你点“同意”

marsbit36 分鐘前

交易

現貨
活动图片